Sustaining the Commons

Authors John M. Anderies Marco A. Janssen

Plaintext

P UBLISHED BY C ENTER FOR THE S TUDY OF I NSTITUTIONAL D IVERSITY
A RIZONA S TATE U NIVERSITY, M ATTHEWS H ALL , 925 S. F OREST M ALL , T EMPE , AZ 85281
HTTP :// SUSTAININGTHECOMMONS . ASU . EDU

Except when indicated otherwise, this work and its contents are licensed under a Creative Commons
Attribution-NonCommercial-NoDerivs 3.0 Unported License (the “License”).

Under this license you are free to copy, distribute, and transmit this work under the following
conditions:
• You must attribute the work in the manner specified by the authors (but not in any way that
suggests that they endorse you or your use of the work). (Attribution)
• You may not use this work for commercial purposes. (Noncommercial)
• You may not alter, transform, or build upon this work. (No Derivative Works)

You may obtain a copy of the License at http://creativecommons.org/licenses/by-nc-nd/3.0.
In memory of Elinor Ostrom
v

Preface
In October 2009 the Nobel committee announced that political scientist Elinor Ostrom would receive
the Nobel Prize in Economic Sciences, together with economist Oliver Williamson, “for her analysis
of economic governance, especially the commons.” Many people had never heard of Elinor Ostrom
or her work. In fact, even many economists, including New York Times columnist and Nobel
Laureate Paul Krugman, were not familiar with her work. Some people noted that Ostrom’s winning
of the Nobel Prize was even more controversial than President Barack Obama’s winning of the Nobel
Peace Prize that year.
What was the controversy all about and what were Ostrom’s discoveries that warranted a
Nobel Prize? Ostrom’s work showed that a traditional approach in economics to the study of
the management of shared resources (e.g., public infrastructure, common-pool resources) was
incomplete. The conventional approach showed that when people share a resource, such as ground
water, fish or a forest, everyone acts in their self-interest leading to overharvesting of the shared
resource. The only way to avoid this so-called “tragedy of the commons”1 is to establish private
property rights or tax the use of the shared resource.
In a series of studies over several decades with many colleagues around the world, Ostrom
showed that people are able to self-organize and successfully govern their shared resources. Her
analysis provides insights into the conditions under which self-governance is possible. These findings
have major implications for policy and can help explain the ineffectiveness of many policies and
governance regimes.
Ostrom winning the prize was not only controversial because of the nature of her research. She
was also the first female to win the Nobel Prize in Economic Sciences. All her official degrees
were in Political Science, although she had minors in Economics. Her work was published in top
journals, but was less well known among the traditional economists. In fact, Ostrom worked across
many disciplines, using qualitative and quantitative methods to address questions relating to the
governance of resources that affect the most vulnerable people in the world.
The theoretical framework she developed over her career is applicable to the study of the
governance of shared resources in many different contexts including the “digital commons,” health
care, and education. Her Nobel Prize increased the attention this line of work has received and
earned her a listing among the Top 100 most influential people in the world by Time Magazine in
2012, just before her death on June 12, 2012.
This textbook will present the main intellectual framework, concepts, and applications of the
work of Elinor Ostrom and her colleagues. The book is geared for an undergraduate audience.
Because of our belief in the importance of Ostrom’s work and a desire to share it with a broader
audience, we started teaching a course on collective action and the commons in 2007 at Arizona
State University. Initially we made use of Ostrom’s classic book “Governing the Commons,” but this
book was not written for an undergraduate audience. Moreover, many new insights have emerged
since the publication of “Governing the Commons” in 1990. Therefore we decided to write our own
textbook, which we have been using since the Spring of 2012.
We worked with “Lin,” as she preferred to be called, from 2000 until her death. We collaborated
on various projects focused on the governance of the commons, especially on questions related to
1 Note, shared resources are often erroneously referred to as the “commons.” In fact the “commons” refers not to a
resource, but a type of governance regime employed in Medieval Europe. Unfortunately this erroneous usage has persisted
in the literature and now the term “commons” is used to refer to a shared resource of some type over which there is no
governance regime - i.e., the resource is “open access.” We use the term “commons” in this sense in this book
robustness. Lin Ostrom had been a professor at Indiana University for her entire career starting in
1965, and beginning in 2006 she held a part-time appointment as a research professor at Arizona
State University in order to collaborate in activities at the Center for the Study of Institutional
Diversity which we direct. Lin Ostrom also gave guest lectures to the class for which we developed
this book.
During a course based on this book, you will learn about institutions, the rules and norms that
guide the interactions among us and among us and the environment which sustains us. These rules
and norms range from traffic rules, rules in sports, and regulations on when and where alcohol can
be consumed to constitutional rules that define on what income sources taxes might be collected
and who can become president of the United States of America. Rules and norms can help guide
us to cooperative outcomes of so-called collective action problems. If we rely only on voluntary
contributions of time and resources to get something done, we will likely find that we will not
achieve the best results. But research also shows that coercive tactics to force people to comply to
strict rules does not necessarily lead to good outcomes. What combination of sticks and carrots
is needed to be successful in solving difficult collective action problems such as governing shared
resources or, as you might see it put in the literature, “sustaining the commons”?
There is no simple formula for designing the best set of rules. Unfortunately context matters and
when it comes to governance, the devil is often in the details. Fortunately, we are building increased
understandings about how context matters. In this book you will learn how to use a framework to
study the norms and rules that are used to solve challenging collective action problems. How do
we evaluate the success of the rules, for whom do the rules hold, and who can craft the rules? We
see that the ways in which people can create and adjust rules and norms affects the success of those
rules. We will discuss a range of examples from sustainability, the digital commons, sports, and
public health.
The book is based on the work of Elinor Ostrom, especially the 1990 book “Governing the Com-
mons” and the 2005 book “Understanding Institutional Diversity.” The way we present the material
is based on teaching this as an undergraduate course Arizona State University. We have included
many contemporary examples of institutions and commons that were not part of Ostrom’s original
publications, but help to explain the concepts more comprehensively. Providing many different
examples also illustrates the wide applicability of Ostrom’s work beyond her main applications in
the area of natural resource management.
In deciding how to publish this book we considered several different options. Finally, we decided
to give it away as a contribution to the knowledge commons. We would like to hear from you what
you think about the book. On the website sustainingthecommons.asu.edu we have a blog where we
post relevant examples on different commons we come across in our research. Let us know if you
come across interesting examples on governing the commons.
We would like to thank the many students of ASB 328 and teaching assistants who provided feed-
back on earlier versions of this book, Jennifer Fraser and Nathan Rollins for editing the manuscript,
and the National Science Foundation for financial support. When we started working on this project
Lin was part of it. She saw the first draft we wrote in April 2012, and she was very pleased with
it. Her passing made us even more motivated to continue the project. Therefore we would like to
dedicate this text book to the memory of our mentor and friend Lin Ostrom.
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

I T HEORETICAL BACKGROUND

1 Why Study the Commons? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1 What are the commons? 3
1.2 The tragedy 5
1.3 The common pasture of Hardin 7
1.4 The tragedy is not inevitable 8
1.5 Plan of the book 10
1.6 Critical reflections 11
1.7 Make yourself think 11
1.8 References 11

2 Defining Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1 Overview 13
2.2 Institutional diversity 14
2.3 How to analyze institutions? 16
2.4 Action arenas and institutional analysis 17
2.5 Context of the action arena 21
2.6 Critical reflections 23
2.7 Make yourself think 23
2.8 References 23
3 Action Arenas and Action Situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.1 Action arenas 25
3.2 The basic working parts of action situations 28
3.2.1 Participants 28
3.2.2 Positions 30
3.2.3 Potential outcomes 32
3.2.4 Actions 33
3.2.5 Control 33
3.2.6 Information about the action situation 33
3.2.7 Costs and benefits 34
3.2.8 Linking action situations 34
3.2.9 Outcomes 36
3.3 Critical reflections 36
3.4 Make yourself think 37
3.5 References 37

4 Social Dilemmas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.1 Introduction 39
4.2 Prisoner’s dilemma 40
4.3 Social dilemmas as an action situation 41
4.4 Coordination games 42
4.5 A typology of goods 43
4.6 Critical reflections 47
4.7 Make yourself think 47
4.8 References 47

II C ASE S TUDIES

5 Water Governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.1 Introduction 51
5.2 Water scarcity 54
5.3 Water abundance 56
5.4 Distribution of water: irrigation 58
5.5 Water pollution 60
5.6 Critical reflections 61
5.7 Make yourself think 61
5.8 References 62
6 Harvesting the Commons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
6.1 Introduction 63
6.2 Domesticated animals 64
6.3 Wild animals 66
6.4 Forests 69
6.5 Critical reflections 70
6.6 Make yourself think 70
6.7 References 70

7 Design Principles to Sustain the Commons . . . . . . . . . . . . . . . . . . . . . . 71
7.1 Introduction 71
7.2 Institutional design principles 72
7.3 Examples of design principles 72
7.3.1 Well-defined boundaries 72
7.3.2 Proportional equivalence between benefits and costs 73
7.3.3 Collective-choice arrangements 74
7.3.4 Monitoring 74
7.3.5 Graduated sanctions 74
7.3.6 Conflict-resolution mechanisms 75
7.3.7 Minimal recognition of rights 75
7.3.8 Nested enterprises 75
7.4 Using design principles in practice 76
7.5 Threats to sustainable use of the commons 77
7.5.1 Rapid exogenous changes 77
7.5.2 Translation failures 77
7.5.3 Blueprint thinking 77
7.5.4 Corruption and rent seeking 78
7.5.5 Lack of large-scale supportive institutions 78
7.6 Critical reflections 78
7.7 Make yourself think 78
7.8 References 78

III H UMAN B EHAVIOR

8 Social Dilemmas in the Laboratory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
8.1 Experiments in social science 81
8.2 The trust game 82
8.3 How people make decisions 84
8.4 Ultimatum and dictator games 86
8.5 Bringing the lab to the field 87
8.6 Critical reflections 89
8.7 Make yourself think 89
8.8 References 90

9 Self-governance in the laboratory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
9.1 Introduction 91
9.2 Common-pool resource action situations 91
9.3 Changing the CPR action situation 94
9.3.1 Communication 94
9.3.2 Costly sanctioning 96
9.4 CPR experiments in the field 97
9.5 More dynamic CPR experiments 98
9.6 Public good action situations 99
9.7 Communication and costly sanctioning 102
9.8 Critical reflections 102
9.9 Make yourself think 102
9.10 References 103

IV RULES OF THE G AMES

10 Classifying Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
10.1 Introduction 107
10.2 How to classify rules 108
10.3 The seven types of rules in detail 110
10.3.1 Position rules 110
10.3.2 Boundary rules 110
10.3.3 Choice rules 111
10.3.4 Aggregation rules 112
10.3.5 Information rules 112
10.3.6 Payoff rules 113
10.3.7 Scope rules 113
10.4 Default conditions 114
10.5 Critical reflections 115
10.6 Make yourself think 115
10.7 References 115
11 Rules, Norms and Shared Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
11.1 Introduction 118
11.2 The syntax components 121
11.2.1 ATTRIBUTES 121
11.2.2 DEONTIC 122
11.2.3 AIM 122
11.2.4 CONDITIONS 123
11.2.5 OR ELSE 123
11.3 How to use the grammar of institutions? 125
11.4 Critical reflections 125
11.5 Make yourself think 125
11.6 References 126

V A PPLYING THE F RAMEWORK

12 Public Health and Collective Action . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
12.1 Introduction 129
12.2 Vaccination 131
12.3 Organ donation 132
12.4 Sanitation 134
12.5 Critical reflections 135
12.6 Make yourself think 135
12.7 References 135

13 The Digital Commons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
13.1 Introduction 137
13.2 Wikipedia: a global public good 138
13.3 Social networking 141
13.4 Intellectual property 142
13.5 Critical reflections 143
13.6 Make yourself think 144
13.7 References 144

14 Rules of the Game: Applications to Sport . . . . . . . . . . . . . . . . . . . . . . 145
14.1 Introduction 145
14.2 Individual sports 146
14.3 Team sports 148
14.4 Sport leagues 151
14.5 Performance-enhancing drugs 151
14.6 Boundary rules 152
14.7 Rule changes 152
14.8 Critical reflections 154
14.9 Make yourself think 154
14.10 References 155

15 Challenges ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
15.1 What have we learned? 157
15.2 Why are there still so many problems in governance? 160
15.3 Closing 162
15.4 Critical reflections 162
15.5 Make yourself think 162
15.6 References 162

Credits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Part I

Theoretical background
Key Concepts
In this chapter we will:
– Define the meaning of the term “commons”
– Discuss the challenges to sustaining the commons
– Introduce the “tragedy of the commons” concept
– Provide examples and a critique of the “tragedy of the commons” concept

1 — Why Study the Commons?

1.1 What are the commons?
The original meaning of the term “commons”
comes from the way that communities managed US infrastructure gets a D+
shared land in Medieval Europe. This shared land
was not owned by any single individual but, rather, According to the prestigious American Society
was “held in common,” thus the term “commons.” of Civil Engineering (ASCE), our roads, dams,
Along with this shared land was a clear set of rules water systems, airports, and other infrastructure
developed by the community about how it was to be are in bad shape. In their 2013 assessment, the
used. Technically, the term “commons” thus refers ASCE gave U.S. infrastructure a D+. In espe-
to the land and the rules that go with it to govern cially bad shape are the nation’s levees and inland
waterways; they both received a rating of D-.
its use. Over time, the term commons has taken on
What would the U.S. need to do to get a better
several meanings. Most generally it can be used to grade? The U.S. would need to invest heavily in
refer to a broad set of resources, natural and cultural, its infrastructure. In fact, ASCE suggests that $3.6
that are shared by many people. Examples of re- trillion needs to be invested by 2020 to bring the
sources that are referred to as “commons” include U.S.’s infrastructure up to good standards.
forests, fisheries, or groundwater resources that are Recent tragedies caused by Hurricanes Katrina
accessible to members of the community. The key and Sandy illustrate the importance of investing
in the maintenance of infrastructure and the conse-
term here is “shared.” Forests, for example, need
quences of not doing so. It is the responsibility of
not be shared–there are many examples of private society at different levels (federal, state, and local)
forests. Thus, implicit in the term “commons” as it to make judgments about how much to invest in
is frequently used today is that there are no property the infrastructure that we all share (by increasing
rights established over the resource. That is, the re- taxes or redirecting spending) and whether to re-
source is “open access.” This departs somewhat from duce taxes, giving more resources to individuals
the original meaning and has, unfortunately, caused to expend as they see fit. Will individuals volun-
some confusion as we shall see later. Other examples tarily invest in building and maintaining critical
infrastructure?
of commons that the reader will encounter in every-
For a wealth of information regarding the state
day life include open source software, Wikipedia, of U.S. infrastructure and the challenges ahead,
public roads, and public education. Throughout this visit www.infrastructurereportcard.org.
book, we will use the term “commons” to refer to
a resource, or collection of resources over which
private property rights have not been established.
Regardless of how they are managed, these examples show that the types of resources that can
4 Why Study the Commons?

Figure 1.1: During the evening rush hour on August
1, 2007 the I-35W Mississippi River bridge collapsed
killing 13 persons. Federal inspection had ranked the
bridge structurally deficient, like about 70,000 other
bridges in the U.S.

be defined as “commons” are essential for our society. We share them, inherit them from previous
generations, and create them for future generations. The commons are therefore crucial for our
wealth and happiness.
Why would we care to study the commons? In this chapter we will explain that there is a
big challenge associated with sustaining the commons. Because of the lack of clear rules of use
and mechanisms to monitor and enforce those rules, some commons are overharvested. Examples
include fishers fishing the oceans in international waters, farmers pumping up groundwater, or
movie watchers using the limited bandwidth of the community internet connection, reducing data
availability for other users. Other types of commons are under-provided, such as high quality public
education and public health services. How can we make sure that the commons are used wisely
and fairly? Who should regulate the use of the commons? Who should make the rules? In the
original commons in Medieval Europe the answer to these questions was clear: the community who
held the land in common made the rules and enforced them to regulate the use of the commons. In
modern commons, where the resources in question are typically much more complex, answering
these questions is much more difficult.
Sourceforge is a repository of open source software projects and as such is a “software commons.”
It is a commons because there are no property rights governing access to the software. There are
more than 100,000 of such open source projects where volunteers contribute their time to create
software and make it available for free to whomever wants to use it. Not every project is as successful
as Linux or Firefox. In fact most projects do not have more than one contributor and most are not
updated. So one question that is of interest to people who study the commons is: “what makes some
projects successful and others unsuccessful?”
In this book, we will discuss many successes and failures regarding governing the commons. We
will introduce a framework that can be used to help us analyze the various types of commons that are
so important to our well being and illustrate how it can be used to provide a better understanding of
how to better govern our shared resources. There is no silver bullet solution that will always lead to
the outcomes we desire, but we can learn about mechanisms that increase the likelihood of desirable
outcomes.
How to govern the commons has been a long debate in academia. Over the last 40 years, the
traditional way of solving the commons problem through privatization or state regulation has been
challenged. The next section will introduce the basic elements of the debate, the controversy that has
arisen around it, and some alternative solutions.
1.2 The tragedy 5

1.2 The tragedy
In 1968 biologist Garrett Hardin (Figure 1.2) wrote a famous essay in the journal Science titled “The
Tragedy of the Commons.” Garrett Hardin was an American ecologist who warned of the dangers
that the increasing human population would impose on the environment. Hardin argued that when
people share a resource they will overharvest it because it is in their individual interest to take as
much as possible.
He used the metaphor of sheep herders sharing an open-access
pasture. Hardin erroneously referred to this open-access shared
resource as a “commons” (if it were really a commons the commu-
nity would use a common-property governance regime to regulate
access-more on this later). The title of his paper should have been
“The Tragedy of Open-Access.” Unfortunately, this use of the term
“commons” stuck and, in fact, has had unfortunate consequences, as
we will see shortly. Because there are no restrictions on the use of
the pasture, each herder can benefit as an individual by adding extra
sheep. Unfortunately, if all the herders add sheep, as a group they
will eventually bear the costs of the additional grazing, especially
when it creates a situation in which the total number of grazing
animals consumes grass faster than the pasture can regenerate new
grass. The effect of overgrazing is shared by all herders, but the
benefit of adding extra sheep goes to the sole owner of the sheep (as Figure 1.2: Garrett Hardin
long as other herders do not add too many sheep).
Based on the reasoning that people are rational selfish actors we can expect the commons will be
overgrazed. Hardin formulates this as follows:

Therein is the tragedy. Each man is locked into a system that compels him to increase
his herd without limit – in a world that is limited. Ruin is the destination toward which
all men rush, each pursuing his own best interest in a society that believes in the freedom
of the commons. Freedom in a commons brings ruin to all. (Hardin, 1968, p. 1245)

The observation that people cause a problem to the common good when they follow their self
interest is not new. The Greek philosopher Aristotle noted more than 2000 years ago that “what
is common to the greatest number has the least care bestowed upon it.” The reason that Hardin’s
argument got so much attention was due to his recognition that the concept can be applied to many
modern environmental problems. With the emerging interest in environmental conservation in the
1960s, he provided an explanation for why we were causing so much damage to the environment.
Hardin concluded that there were only two options to avoid the depletion of the commons. One
option was to give the herders private property rights. If each herder owned a piece of the common
land and the sheep caused overgrazing and erosion, the costs would be felt by the individual herder
only. For this reason, the herder will choose to put an appropriate number of sheep to graze the land
in order to maximize her long-term earnings. The other possible option is for a government body to
restrict the amount of grass that can be consumed. However, in this case, the government would have
to monitor the amount of grass consumed by each herder–a costly exercise. An alternative would be
for the government to require that herders pay a tax per head of sheep, which the government would
use to hire a guard to monitor whether the herders follow the rules.
6 Why Study the Commons?

The importance of Hardin’s argument is its conclusion that people are not able to self-govern
common resources. That is why he calls it a tragedy. The fact that Hardin focused on this inevitable
tragedy is perhaps related to his use of the term “commons.” In fact, in traditional contexts there was
no “freedom in a commons”—a commons always had a set of rules associated with its use, and these
rules did not necessarily include either of Hardin’s two options. Unfortunately, Hardin’s judgment
has been widely accepted due to its consistency with predictions from traditional economic sciences
and increasing numbers of examples of depletion of environmental resources. What this judgment
fails to take into account are the many cases of successfully managed commons in which the shared
resource is used sustainably. That is, there are many cases where a “tragedy of the commons” has
been averted without privatization or state control.
The consequences of this work were significant. Hardin and others did distinguish three types
of property rights: communal, private and state. However, they equate communal property with
the absence of exclusive and effective rights and thus with an inability to govern the commons.
Experience does not bear this definition out: communal property, or common-property governance
regimes do provide exclusive and effective rights, which are often used to govern the commons.
From Hardin’s perspective, which neglected this third governance regime, sustainable use of shared
resources without the state or private property was only possible when there was little demand or a
low population density.
Garrett Hardin provided a compelling explanation for the emerging environmental movement
in the 1960s. There was an increasing awareness of the decline of natural resources due to human
activities including the perceived scarcity of raw materials, deforestation, overfishing, as well as
increasing levels of water and air pollution, leading to smog and acid rain as well as health problems
for the human population.
A few years after the publication of Hardin’s article, the first oil crisis took place which led to a
rapid increase in oil prices. This shock generated the perception that oil was becoming scarce and
that we were overharvesting our common resources. Hardin’s paper provided a simple analysis and a
simple solution. Assuming people make rational decisions, the implications for policy were clear. To
avoid overharvesting of common resources it was critical for the state to either 1) establish, monitor,
and enforce private property rights or 2) directly regulate the use of the commons either by taxing or
directly restricting (e.g., licensing) its use.
Figure 1.3 shows the decline of the stock of predatory species in the various oceans around
the world (Myers & Worm, 2003). Since the 1968 essay, policies have changed, yet we haven’t
seen a reversal of the trends. The fish stocks in Figure 1.3 still have not started to recover despite
the existence of new fishing policies since the early 1970s. Moreover, we are now beginning to
experience new environmental commons problems like the loss of biodiversity and climate change
despite efforts by nations to draft international treaties to regulate these “global commons.”
As we have hinted above, we will show in this book why Hardin’s analysis was limited. Although
we see resource collapses around the world (tragedies of open access), we also see many success
stories of long-lasting governance of shared resources (triumphs of the commons). Open access
situations are not always tragedies. Many times common-property management regimes fail, as
do private property and state-centric regulatory governance regimes. There are no panaceas. The
goal of this book is to illustrate a set of tools that can be used to determine what conditions make
overharvesting more likely and what conditions will lead to the sustainable use of shared resources.
1.3 The common pasture of Hardin 7

Figure 1.3: Relative biomass estimates from the beginning of industrialized fishing (Myers & Worm, 2003)

1.3 The common pasture of Hardin
As we mentioned above, in his description of the “commons,” Garrett Hardin implicitly assumed
open access of the pasture. The example Hardin gave was grazing on common land in Medieval
Europe. Let’s look at the actual situation of the medieval open-field system in Europe, especially in
England, in more detail (Figure 1.4).
In the open-field system, peasants had private property
rights to the grain they grew on multiple small strips of
land that were scattered around a central village. However,
during particular seasons, peasants were obligated to throw
the land open to all the landowners in a particular village
so that they could all graze their sheep on the common land
under the supervision of one herdsman. The decision to
convert the strips of privately used land into shared land for
a period during each year was made by a village council.
This enabled people to take advantage of economies-of-
scale in grazing (as well as providing manure for their
land) and private incentives in grain growing (which lacks
important economics of scale and suffers from free-riding
[an example of a social dilemma, a topic we will discuss
in Chapter 4] when communal groups try to share labor
inputs).
The purpose for scattering small strips of land has
been debated among scholars, as the benefits of the two
scales could be achieved with or without the scattering of Figure 1.4: Open field system
the agricultural land. And the scattering appears to have
been an inefficient system, given that a single farmer had to divide his time between multiple, small
8 Why Study the Commons?

agricultural strips rather than being able to economize on his own time and focus on one piece of
land. Some scholars argue that the need to share risk due to different soil and precipitation patterns
may have been a contributing factor. Others argue that by not allowing any one farmer to gain a large
amount of contiguous land, the village avoided creating a situation of asymmetric bargaining power.
No farmer owned enough land to be able to “hold out” from the commons and graze his own animals
on his own land. Nor did an individual have a right to exclude others once the village decided the
land should be converted from agriculture to pasture. If all of the farmers had owned sizable chunks
of agricultural land in fee simple (a form of private ownership in England), rather than the village
being responsible for land-allocation decisions, transaction costs would be very high.
If the commons were managed effectively in the open-field system, why did it disappear? And
why did it take such a long time for it to disappear across most of Northern Europe? If private
property alone was a very efficient solution to the production of food, once a particular location
discovered this efficient solution, one would expect to see a change occur rapidly throughout
Europe. The explanation might relate to transportation costs. Due to high transportation costs, local
communities needed to produce both meat and grain in a small local area for their own consumption.
This was only feasible if they could convert agricultural land to a common pasture when the crops
had been harvested. When transportation networks improved and communities gained access to
markets in grain and meat, there was no longer a need to continue with this complicated adaptation.
Communities could specialize in meat or grain. Interestingly, this shift was facilitated by the
development of a new “commons”: the shared resource of the public transportation system.
Thus, as we mentioned above, the medieval commons used by Hardin as a metaphor were
in reality not open access. The commoners had crafted effective norms and rules to govern their
common pasture and to avoid overharvesting. Moreover, there are many implicit rules involved in
the use of the commons. The livestock is the private property of the farmers, but the grass does not
become private until a sheep swallows it. Could a farmer directly harvest the grass for her sheep?
The farmer who does so will likely get in trouble as there might be informal rules that grass can only
be harvested via the sheep.
As you may have noticed with this simple example of a common pasture with grazing sheep,
common-property governance typically involves many rules and norms. Often, the intentions of
these rules and norms and the way they function are not at all obvious from the start. We will see
that there are always many such norms and rules involved in the use of the commons.
In summary, at the time Hardin wrote his now-classic article, the work on collective action was
rooted in rational choice theory. A key assumption was that actors made rational decisions based
on selfish motives. The implications for policy were clear. To avoid overharvesting of common
resources it was critical to establish private property rights or tax the use of the commons. Much
work since has shown that this simply isn’t the case.

1.4 The tragedy is not inevitable
Since Hardin’s essay, an increasing awareness has emerged that tragedy is not the only possible
outcome when people share a common resource. There are many examples of long-lasting communi-
ties that have maintained their shared resources effectively. Since the 1980s there has been a steady
increase in interdisciplinary efforts to debunk the simplistic view of the tragedy of the commons.
Elinor Ostrom (Figure 1.5) and others showed through comparative analysis of many case studies
that humans can self-govern their shared resources.
Elinor Ostrom, was a political scientist who developed a theoretical framework to study the
1.4 The tragedy is not inevitable 9

ability of communities to overcome the tragedy of the commons. This research earned her the 2009
Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel (better known as the
Nobel Prize in Economics). Her PhD thesis, which she finished in 1965, focused on the management
of shared groundwater resources in Southern California. In her first fifteen years on the political
science faculty at Indiana University she studied police forces in U.S. cities, seeking to discover
which type of organization led to the most effective policing.
Because she worked on various types of projects related to the
governance of shared resources, she started to see commonalities.
Since the early 1980s, Ostrom developed a more theoretical under-
standing of the institutions, rules and norms that humans use to
organize themselves. This led her to create the Institutional Analysis
and Development (IAD) framework, which is a core framework in
this book.
During the mid-1980s Ostrom returned to the study of the en-
vironmental commons. An increasing number of scholars at the
time were realizing that reality clashed with the conventional view
that the use of a shared resource would end in environmental dis-
aster. Ostrom proved instrumental to this revolution in thinking by
leading an effort to compile hundreds of case studies—successes
and failures—from the lobster fisheries of Maine to the irrigation
systems of Nepal. Figure 1.5: Elinor Ostrom
All of the analyses of these case studies allowed her to distin-
guish features that were more common in successful cases. In her
1990 book Governing the Commons, she identified eight design principles that characterized suc-
cessful self-governance strategies, including having monitors who are accountable to the users of a
resource and cheap mechanisms of conflict resolution. Those principles are discussed in Chapter 7
and have held up to the test of time.
Since the early 1980s an increasing number of anthropologists, sociologists, political scientists,
ecologists, and many other scholars have been documenting examples of common resources that
have been managed sustainably for a long time without private property rights or governmental
interventions. This led to the development of a community of scholars who came together to create
the International Association for the Study of the Commons (www.iasc-commons.org) of which
Elinor Ostrom was the first president.
The work coming out of this community has provided an alternative framework to study the use
of shared resources, i.e. resources held in common. The material discussed in this book is largely
based on this alternative framework which has been widely recognized. Besides a Nobel Prize in
Economics (which was seen by Ostrom as a recognition of the whole research community in this
area, not an individual accomplishment) insights derived from this research are increasingly applied
to governance and policy issues.
Applications can be found in organizations that manage development projects in developing
countries, advance agricultural practices to improve food security, and protect biodiversity. Moreover,
the insights on how to sustain the commons are increasingly applied to non-traditional commons
such as knowledge, culture, education, and health. For example, the communication revolution
caused by the internet has generated all kinds of new challenges related to governing the digital
commons. Creations consisting of mainly information (movies, books, music) are so easy to copy,
that many get distributed without any payment to the owners of the intellectual property rights.
10 Why Study the Commons?

Strangers can post improper comments to websites. Emails are sent around in order to gain access to
your private information.
Although the framework on which this book is built has its roots in the study of traditional types
of commons, we will see that it can be used in many other action arenas ranging from urban open
space to health care. As such, the book does not focus on a specific application but, rather, provides
a series of concepts and frameworks that help us better understand the challenges of sustaining the
commons and design possible solutions.

1.5 Plan of the book
The book consists of 15 chapters. The first few chapters discuss some basic concepts and frameworks
such as institutions, action arenas and social dilemmas. These concepts will provide the key
theoretical foundation for analyzing problems related to the commons. In Chapter 2 we define
institutions, the rules and norms that structure human interactions. This is a very broad concept but
we will see that understanding the rules and the norms related to the use of the commons helps us
understand how to sustain them. We will use the general terminology of “institutions” rather than
of private property or markets, since those two examples are vague and imprecise definitions of
clusters of possible institutional arrangements. The institutional analysis and development (IAD)
framework that we will discuss in this book provides a more general and accurate way of studying
institutions and their performance.
In Chapter 3 we will focus on action arenas, a key component of the IAD framework which
connects participants to an action situation. We can use the concept of action arenas to dissect what
are the incentives, the possible actions, and the positions of people who are using the commons.
Social dilemmas, discussed in Chapter 4, are action arenas in which two or more persons can benefit
collectively from cooperation, but individually from freeriding or defecting. We discuss various
commonly studied social dilemmas such as the prisoner’s dilemma, public goods dilemmas and
common-pool resource dilemmas.
Chapters 5 and 6 discuss various empirical examples the governing the commons related to
natural resource use. The examples cover a range of different resources, from water management in
Arizona to lobster fisheries in Maine. Using the language and concepts introduced in Chapters 2-4
we will see that we can better understand the challenges people have to overcome in order to sustain
their commons. We will also see some remarkable examples of institutions that facilitate success.
The eight design principles defined by Elinor Ostrom, based on the analysis of many case studies,
are discussed in Chapter 7.
Finding regularities in case studies is not sufficient to understand the mechanisms that enable
communities to successfully sustain their commons and those that cause them to fail. Social scientists
use controlled experiments to test hypotheses about such mechanism. These experiments help us
better understand how humans make decisions in various types of action arenas. Moreover, we
can replicate the findings over and over again to build our confidence that we can generalize the
results. Chapter 8 discusses general findings from social dilemma experiments, and Chapter 9
focuses specifically on public goods and common pool resources. We see that humans do not always
act as the selfish rational actor model would predict. This calls into question a basic assumption
underlying the argument of Garrett Hardin. In contrast we find that humans have other-regarding
preferences and cooperate if they expect others will cooperate too.
In studying many case studies and experiments, we find many types of rules. Chapter 10 provides
a rule classification scheme based on linking the rules to the different attributes of the action arena.
1.6 Critical reflections 11

So far we have loosely talked about rules and norms, but in Chapter 11 we will be more precise and
introduce an approach to dissect and classify rules and norms from their linguistic statements.
Chapters 12, 13 and 14 will apply the concepts and frameworks covered in Chapters 2-11 to
different types of action arenas including public health, the digital commons and sports. These
examples illustrate how our tools help understand the challenges faced in different types of action
arenas. Chapter 15 closes the book with a discussion on the main lessons learned and which
challenges scholars in this field are now addressing.

1.6 Critical reflections
Commons are natural and cultural resources that are shared by many people. People can affect the
commons by harvesting from them and making contributions to their construction and/or preservation.
The core question this book attempts to address is how we can sustain the commons. Garret Hardin
introduced the notion of the tragedy of the commons which can occur if people share a resource.
The opportunistic behavior of individuals can lead to overharvesting of the shared resource. The only
way to avoid the tragedy, according to Hardin, is to establish private property rights or tax the use of
the commons. Elinor Ostrom and her colleagues show from case study analysis that overharvesting
is not inevitable and that successful self-governance of the commons is possible.

1.7 Make yourself think
1. Come up with commons you experience yourself.
2. Are these commons functioning well?
3. Did your grandparents use different commons than you do?
4. Now that you know about the commons, can you relate the idea of the commons to the budget
discussions in Washington D.C.?

1.8 References
Hardin, G. (1968). Tragedy of the commons. Science, 162, 1243–1248.
Myers, R., & Worm, B. (2003). Rapid worldwide depletion of predatory fish communities. Nature,
423, 280–283.
Ostrom, E. (1990). Governing the commons: the evolution of institutions for collective action.
Cambridge University Press.
Ostrom, E. (2005). Understanding institutional diversity. Princeton, NJ: Princeton University Press.
Key concepts
In this chapter we will:
– Learn how to define institutions
– Begin recognizing institutions in everyday situations
– Become familiar with the Institutional Analysis and Development Framework (IAD)
– Understand how incentives that impact decision making
can be studied using rigorous scientific methods
– Recognize the wide diversity of institutions in use around the world

2 — Defining Institutions

Cow jails
In the hills of Nepal there are some remarkable ways people manage irrigation systems. When Elinor
Ostrom performed field work in this area in the 1990s she noticed an enclosed field with a domesticated
cow in the center of a village. Her Nepali colleagues explained that this was a kind of “cow jail.” If
three adult members of the local irrigation system agreed that a village member had not followed the
irrigation rules, they could confiscate a cow from the offender. These rules were related to how much
water a farmer was allowed to take from the irrigation system and how much labor was to be contributed
to maintaining the canals of the irrigation system. In these small communities everybody will recognize
the cow. Thus, if your cow is grazing in the center of the village, everybody knows you are cheating
the community by either taking more than your fair share of water or by not contributing sufficient
labor to maintaining the irrigation infrastructure. Moreover, people in the community could milk the
cow—a sort of payment for the water or labor that they lost to the guilty farmer. Once the farmer has
paid a fee (on top of the milk he lost from his jailed cow) the cow would be released and returned to
the owner. Needless to say, most members of the irrigation system prefer to follow the rules rather than
being embarrassed in this way.
The use of cow jails to get people to comply with the rules looks very exotic. Cow jails are unlikely
to work in the U.S. However, the principle of shaming as punishment is also used in the U.S. For example,
many states publish the names of families and firms who are delinquent on their taxes. The top 250 tax
delinquents in the State of California can be found at https://www.ftb.ca.gov/aboutFTB/Delinquent_
Taxpayers.shtml. This list might be less controversial than a cow in the middle of a village, but being on
that list will affect your reputation. As a result, such families and companies may correct the situation
and pay their taxes in order to get their name off the list (i.e., get their “cow out of jail”) and gain back
their reputation.

2.1 Overview
The example of the cow jail illustrates that there are very diverse ways to induce people to contribute
to the commons. In the case of the cow jail the “commons” consists of the shared infrastructure made
up of canals and water diversion mechanisms and the common-pool resource of irrigation water (we
will define these different types of goods that appear in the “commons” more precisely in Chapter 4).
When farmers do not contribute to the maintenance of the shared infrastructure, they will experience
some shaming. The solution in villages in Nepal cannot be blindly copied and be successful in other
places. Most families in urban areas have no cows, a fact that will limit the applicability of this
solution. We will see that there are many different possible mechanisms that groups use to solve
problems related to sustaining the commons. These different mechanisms, however, all rely on the
same principles. Therefore, in this book, we will try to understand how people solve these problems
14 Defining Institutions

by studying the basic principles contained in the institutions they use.
Broadly defined, institutions are the prescriptions that humans use to organize all forms of
repetitive and structured interactions. This includes prescriptions used in households, schools,
hospitals, companies, courts of law, etc. These prescriptions can function at different scales, from
households to international treaties. These prescriptions can be one of two broad types: rules or
norms. Because rules and norms are essentially human constructs, agreed-upon or recognized by a
group of people, they are not immutable. That is, individuals can make choices whether or not to
follow the rules or norms. Importantly, their choices and actions have consequences for themselves
and for others.
In the following chapters we will see that rules and norms are everywhere and define - sometimes
literally, sometimes indirectly—how we live our lives. For example, rules and norms can affect who
we marry, which schools we go to, which countries we enter, where we may sit on a bus, where we
may park, who leads a discussion in a group, etc. Where do these rules and norms come from, and
why do they differ in different countries and contexts? In this book, we are especially interested in
answering this question for different types of commons.
We will see that all of us can play a big role in defining rules and norms if we take the initiative
to do so. Crafting rules and norms is not something that is undertaken exclusively by those in
business suits in Washington, D.C. We ourselves create rules and norms too. For example, when you
undertake a group project during a course, you will have to rely on some rules and norms. Some
rules might come from the syllabus, others are created by you and the members of your group during
your meetings.
In more abstract terms, the rules (or the absence of rules) in a particular situation affect who gets
what benefits, who bares what costs, who is allowed to participate, and who gets what information.
Further, the rules affecting one situation are themselves crafted by individuals interacting at higher
levels. For example, the rules we use when playing basketball at lunch time were themselves crafted
by officials who have to follow such rules and norms to structure their deliberations and decisions.
This chapter provides a brief overview of the framework we will use in this book to study
institutions. In the following sections we will discuss these core questions:
• Why are there so many different types of institutions?
• How do we analyze institutions?
• What is the appropriate unit of analysis for studying institutions in general, and the commons
in particular?
• How do we use one choice of an analytical unit, the action arena, to study institutions?
• What are the core components of an action arena?

2.2 Institutional diversity
During a typical day we experience many situations in which we interact with others in a structured
way where rules and norms may apply. This can be at work, in the classroom, at the sport field, in
the supermarket, during commuting, when we bring our kids to daycare, when we watch a movie
online, when we go to church, at the dinner table, etc. In all these different settings different types of
norms and rules hold. At work you may have a formal contract regarding the duties that are expected
from you and the compensation you are given for undertaking those duties. At the dinner table you
may adopt some manners (which are the equivalent of norms) taught to you by your parents. In
traffic you follow the norms and rules of the road. For example, a rule of the road is a speed limit. A
norm is you don’t cut other drivers off when changing lanes. Can you tell the difference between a
2.2 Institutional diversity 15

rule and a norm just based on these examples? Finally, we interact with many strangers every day
whom we expect to follow the same rules.
When you start realizing the number of rules and norms we implicitly deal with on a daily
basis, it might become overwhelming. But most of us are easily able to participate in all these
diverse sets of situations without thinking too much about the rules and norms that structure
them or specific decisions we make in those situations. Several scholars have explored the ques-
tion of what enables us to do this. Not only are we faced with many different situations each
day, the situations we can experience change over the generations. It is likely that today we
experience more different types of situations at different levels of social organization as com-
pared to previous generations. People living in a small village in Europe in 1200 were not
thinking about the implications that political developments in China might have on their lives.
We now expect to communicate with our relatives or be able to check
on the latest news wherever we are in the world. Our meals are not
restricted by the seasonal availability of foods produced by local
farmers. We transport the ingredients for our meals from all over the
world (e.g., tropical fruits and vegetables in the winter in New Eng-
land) at considerable environmental costs. Such changes are not just
caused by technological developments, but also through changes in
institutions. To make sure fruits and vegetables are transported reli-
ably from location A to B, we have to create institutions to structure
repetitive interactions between all the individuals involved. Without
institutions, the transaction costs for exchanges between farmers,
transporters, and retailers would make long distance transport of
food extremely costly.
It is obvious to us what to do when we are shopping in a super-
market. We take the goods we prefer from the shelves. We then
“arrange a meeting with the cashier,” which is made easy by check
out lines and the norms for standing in lines (standing in lines is not Figure 2.1: Differences in tech-
a norm everywhere). The cashier knows we wish to have a meeting nology affect the type of institu-
by virtue of the fact that we are standing in line. We then engage in tions that are used.
an exchange with the cashier. What exactly do we exchange with the
cashier? Do we exchange food? No—the cashier does not own the food. We exchange information.
We may give a piece of plastic with information on it (a credit card) to the cashier or we may use
cash—which is also a form of information about value and obligation. But this strategy does not
work everywhere. When we are shopping in an open bazaar in Asia or Africa, we may bargain over
the price of the fruit that is left on the stand at the end of the day. Such bargaining to get a lower
price is also happening for other goods in a bazaar. In fact, not bargaining (i.e., not adopting a local
norm) for a lower price will be a clear indication that you are a stranger and you do not know what
to do in this situation. This may drastically affect the price of the good. Finally, in this case, the
seller may actually own the fruit and thus will be exchanging goods with you. What will you be
exchanging for the goods? Probably not a number and an expiration date on a piece of plastic. Can
you use U.S. dollars, say in Africa? Maybe, maybe not. These examples illustrate that there are
many (subtle) changes from one situation to another even though many variables are the same. These
subtle changes can have major consequences for the interactions between people.
The types of institutional and cultural factors we have been discussing affect our expecta-
tions regarding the behavior of others and their expectations regarding our behavior. For example,
16 Defining Institutions

once we learn the technical skills associated with driving a car, driving in Phoenix (Arizona) or
Bloomington (Indiana)—where everyone drives fast but generally follows traffic rules—is quite a
different experience from driving in Rome, Rio, and even in Washington, D.C., where drivers
appear to be playing a bluffing game with one another at intersections rather than following
traffic rules. When playing racquetball with a colleague, it is usually okay to be aggressive
to try to win by using all of one’s skills. On the other hand,
when teaching a young family member how to play rac-
quetball, the challenge is how to help them have fun while
they learn a new skill. Being too aggressive in this setting—
or in many other seemingly competitive situations—may
be counterproductive. A “well-adjusted and productive”
adult adjusts their expectations and ways of interacting
with others to “fit” a wide range of different situations.
Such adjustments are often second nature.
Although we may not explicitly realize it, we have a
lot of implicit knowledge of expected do’s and don’ts in Figure 2.2: If you want to buy a ticket for a
a variety of situations. Frequently, we are not even con- concert and you see people standing in line,
scious of all of the rules, norms, and strategies we follow. you would automatically join the back of the
Nor have the social sciences developed adequate tools to line. What do you suppose would happen
help us translate our implicit knowledge into a consistent if you bypassed the line to buy your tick-
explicit theory of human behavior. In most university ets? Although there might not be any formal
signs that say you need to wait for your turn,
courses students learn the language of a particular disci- it is generally assumed you understand you
pline, from anthropology to economics, from psychology have to do this.
to political science, etc. This disciplinary narrowing of
language may hinder our understanding of how to analyze
the diverse sets of situations we encounter is social life. The framework we discuss in this book may
provide a common language to study these different situations.

2.3 How to analyze institutions?
If the situations in which people experience dif-
ferent norms and rules are so diverse, how can we The genes of institutions
study them? How can we make sense out of such
complexity? Given that there is such a large variety There are millions of different species on our
of regularized social interactions in markets, hier- planet that interact in complex ways at different
archies, families, legislatures, elections, and other spatial and temporal scales. How does one study
situations, is it even possible to find common termi- such complexity? One of the breakthroughs in
nology to study them? If so, what framework could biology is the concept of genes and the discovery
we use to analyze these different situations across of DNA, the building blocks of the diversity of
lifeforms on Earth.
different cultures? Can we learn from one type of
Can we develop an equivalent set of concepts
institutional arrangement and apply the lessons to for building blocks that create institutions? This
another one? will help in the study of the large institutional di-
Can we identify attributes of the context in which versity we observe around us. In this book we
people carry out their repeated interactions in order will discuss the initial steps of a genomics of in-
to find communalities that distinguish success stories stitutions that enable us to decode and compare
from failures? If we are successful with this, we may institutions.
2.4 Action arenas and institutional analysis 17

be able to explain behavior in a diversity of situations
varying from markets and universities to religious groups and urban governance. This analysis of
interactions among people may take place at a range of levels from the local to the global, and we
may analyze whether processes occurring at the local level may explain some of the challenges at
the global level.
These are all very ambitious goals. However as you will see from
the material in this book and associated coursework, the framework
that we will discuss will help to provide us with a much better
understanding of key features that appear throughout a diverse set of
situations. The framework is an outcome of many studies conducted
at the Vincent and Elinor Ostrom Workshop of Political Theory and
Policy Analysis at Indiana University, which was created in 1973 by
Vincent and Elinor Ostrom (Figure 2.3). Many of their colleagues
all over the world have contributed to this framework by testing it
on diverse sets of problems. In teaching this framework over the
years in undergraduate classes at Arizona State University, we have
developed this book to communicate the rather complex framework Figure 2.3: Elinor and Vincent
to a broader audience. Ostrom
In the rest of this chapter we will provide a brief overview of
the basics of the framework. The framework is called the Institutional Analysis and Development
(IAD) framework. One of the aspects of social systems that makes the IAD framework complex
is the existence of different types of regularized social behaviors that occur at multiple levels of
organization. There is no simple theory that predicts everything, and therefore we need to understand
what kind of behavior is to be expected in each type of context.

2.4 Action arenas and institutional analysis
When two people exchange a product on eBay,
they are in an action arena. This is an example of An action arena occurs whenever individu-
the focal level of analysis we use throughout this als interact, exchange goods, or solve prob-
book. In an action arena participants, rules and lems. Some examples are teaching a class,
norms, and attributes of the physical world come to- playing a baseball game, having dinner.
gether. The latter two elements, the rules and norms
and physical world are said to define an action situ-
ation. Action situations remain stable over time relative to the participants who may take part. For
example, the eBay action situation does not change over the course of a day during which millions
of participants can enter the action situation and generate an action arena. As participants interact
in the action arena, they are affected by exogenous variables and produce outcomes that, in turn,
affect the participants and the action situation. Action situations exist in homes, neighborhoods,
regional councils, national congress, community forests, city parks, international assemblies, and
in firms and markets as well as in the interactions among all of these situations. The simplest and
most aggregated way of representing any of these arenas when they are the focal level of analysis is
shown in Figure 2.4, where exogenous variables affect the structure of an action arena, generating
interactions that produce outcomes. Evaluative criteria are used to judge the performance of the
system by examining the patterns of interactions and outcomes.
Let’s discuss some examples. Consider two participants: John and Alice. When John and Alice
18 Defining Institutions

play a game of chess, the action situation is composed of (a) the physical game of chess including
the board with 64 squares and the pieces: 2 black and 2 white rooks, 2 black and 2 white knights,
2 black and 2 white bishops, 1 white and 1 black queen and 1 white and 1 black king; and (b) the
rules of chess—how each piece can be moved, how pieces interact, and what constitutes a victory.
When John and Alice sit down at the chessboard to play, this forms an action arena. The interactions
between the players may lead to either John or Alice winning the game or a tie. Hence the outcome
is whether game is won by one of them or whether it was a tie. The same persons may also be in an
action arena involving money lending. In this action arena, the action situation may be less structured
than the chess game. Consider the action arena in which Alice lends money to John. Suppose Alice
and John are good friends and the amount of money is small. Alice gives the money to John who
agrees to return the money at some specified date (often rather vague in such situations). In this
case, the action situation is simple: it is defined by the shared norms of informal money lending in
Alice’s and John’s culture. Suppose, on the other hand, that this exchange is performed in a formal
way. Another participant enters the action arena, a notary public, who formulates a contract that is
signed by Alice and John. In this case, the action situation is slightly more complex as it involves
a formal contract legitimized by the notary’s presence and the signatures of Alice and John. Now
the formal rules of contract law, the testimony of a third party recognized by the state (notary) who
will testify to the identity of the signatories of a contract, and an entity that will archive the contract
form the action situation. The outcome of this transaction is that John receives the money and pays it
back according to the conditions as stated in the contract. A third possible action arena would be
an election. Alice and John are both candidates for president of the student association. Within the
action arena participants include all the students of the association who are allowed to vote for one
of the candidates. The interactions include debates, a campaign, and finally the election day in which
a winner is decided. The evaluation criteria stipulates that the winner is determined based on which
candidate has a simple majority (i.e. more than 50%) of the votes. In the last example, Alice and
John are neighbors who have a conflict about the barking of Alice’s dog. The action situation is a
conflict. Within the action arena we have Alice, the dog, John, and the local authorities whom are
called by John to intervene. Alice and John may both hire lawyers to represent themselves when the
action situation (conflict) is played out in court. The interactions include the daily occurrences of the
dog barking, the initial friendly requests of John to silence the dog, and the escalation of the conflict
into a court case. There are various possible outcomes: John or Alice moves out of the neighborhood,
the dog gets training to stop barking, the dog is sold, John gets a financial compensation for the
inconvenience, etc. Each outcome is evaluated differently by each of the participants, including the
lawyers. For example, if John’s lawyer gets a certain percentage of the financial compensation, she
may focus on winning a case to get financial compensation, although this may lead to long-term bad
relations between Alice and John.
Outcomes feedback into the participants in the action arena (the dashed arrow from outcomes
to the action arena in Figure 2.4). For example, the fact that a player loses a chess game affects
her next decision regarding the action situation of playing chess (play another game or not). The
dog continuing to bark after one interaction (John asks Alice to quiet the dog) will undoubtedly
affect John’s next decision. This changed view by one or several participants may induce the action
situation to transform over time as well. Over time, outcomes may also slowly affect some of the
exogenous variables. For example, decisions people make regarding energy use creates outcomes
including emissions of CO2 which in the long term affect the climate system. In a world with a
changed climate the costs and benefits of various human activities are affected, which will affect
action arenas. In undertaking an analysis, however, one treats the exogenous variables as fixed—at
2.4 Action arenas and institutional analysis 19

Exogenous
Action Arena Interactions
Variables

Evaluative
Criteria

Outcomes

Figure 2.4: The focal level of analysis—an action arena (adapted from Ostrom (2005).

least for the purpose of the analysis.
When the interactions yielding outcomes are productive for those involved, the participants may
increase their commitment to maintaining the structure of the situation as it is, so as to continue to
experience positive outcomes. For example, wealthy people who may have benefited from low taxes
in the past may support tax cuts that the Bush administration introduced. However, if participants
view interactions as unfair or otherwise inappropriate, they may change their strategies even when
they are receiving positive outcomes from the situation. For example, a group of millionaires
requested that President Obama raise taxes for wealthy people (http://politicalticker.blogs.cnn.com/
2011/11/16/millionaires-ask-congress-to-raise-their-taxes/).
When current outcomes are perceived by those involved (or others) as less desirable than other
outcomes that might be obtained, some participants will raise questions about particular action
situations and attempt to change them. But rather than trying to change the structure of those
action situations directly, they may move to a different level and attempt to change the exogenous
variables. The Occupy Wallstreet movement of 2011 was a protest against the perceived unfairness
in society due to a culture of greed by bankers and other participants who control the financial
system. The protesters requested a change of the financial system (the exogenous variable) in order
to move toward a more equitable society, in which they may also succeed (a different outcome)
(http://www.huffingtonpost.com/2011/11/08/wall-street-bonuses_n_1081902.html). But they didn’t
try to change the banking system directly. They tried to affect the exogenous variables, for example,
the perception of the general public toward the actions of banks.
Figure 2.4 is the simplest schematic representation of an action arena. As you see from the
example, there are many important layers to each action arena. We unpack this simple representation
in Figure 2.5 in order to make these layers more apparent. An action arena refers to the social space
where participants with diverse preferences interact, exchange goods and services, play a game, solve
problems, have an argument, receive and deliver health care, etc. We make a distinction between an
action situation and an action arena to emphasize that the same participants can fill different roles in
different action arenas as we saw with John and Alice. The action situation refers to the positions,
actions, outcomes, information and control that provide the structure by which participants interact.
Thus the action situation provides the institutional context with which the participants in an action
arena are confronted. In Chapter 3, we will zoom in and unpack the action arena. Let’s look at a
broader overview of the IAD conceptual map.
Let’s apply the framework to a concrete example, namely the course you are taking for which
you are reading this book (Figure 2.5). The action situation is defined by the general rules about
20 Defining Institutions

Figure 2.5: A framework for institutional analysis (adapted from Ostrom, 2005)

taking a course at your university (grades, credits, conduct) further specified by the syllabus for this
particular course and the characteristics of the space in which the participants meet. Your taking this
course (along with all the other students) then constitutes an action arena. In the action arena there
are a number of different participants, namely the students, the professor and the teaching assistant.
The participants interact via lectures, taking exams and writing essays. The syllabus of the course
specifies what is needed to receive a good grade in the course. It specifies the weight of the different
types of interactions, from participation in class, giving a talk, writing an essay and taking an exam.
For each of these activities there are more detailed evaluation criteria on how to receive a good exam
grade or writing the essay. The final outcome of the course is a grade.
The exogenous variables in which these interactions take place are the facilities of the university
campus (the quality of the classrooms, computer commons, etc.), the attributes of the students
(what criteria is required to be admitted to the university, quality of other courses, etc.), and the
university regulations. These are specific examples of the general categories of exogenous variables
in Figure 2.5: The biophysical conditions, the attributes of the community, and the rules in use,
respectively.
Although the final grade is mentioned here as the outcome of the course, this can be debated. If
this were truly the only outcome we cared about, the participants could agree (e.g., all vote to give
each other an A) that the students could all get an A without putting in the effort of taking the course.
Obviously, this is not the purpose of a course and is a violation of university regulations. Although
the focus of many participants in the action situation might be on the grade, there are other outcomes
that we may include. Does the course material lead to new insights and useful experiences for the
students? Do the students comprehend the material and can the students apply this to other topics or
problems they may encounter in life? Is the atmosphere in the classroom pleasant and productive?
These kinds of outcomes are more difficult and costly to quantify, but are nonetheless very important.
However, the difficulty with measuring such outcomes might be a reason that officials may choose to
focus on grades to measure course outcomes.
2.5 Context of the action arena 21

Figure 2.6: Framework applied to taking a course.

2.5 Context of the action arena
The action arena does not occur in a vacuum. Participants are interacting in an action situation
which is affected by a broader context. As mentioned above, this broader context is defined by three
clusters: (1) the rules used by participants to order their relationships, (2) the biophysical world that
are acted upon in these arenas, and (3) the structure of the more general community within which
any particular arena is placed.
Different scholarly disciplines focus on different clusters. Anthropologists and sociologists may
focus more on the role of the community and culture while economists focus more on how rules
affect the incentives of the participants. Environmental scientists may focus more on the biophysical
attributes of the action arena. In this book we focus on the rules, but take into account the role of the
community and the biophysical environment.

Rules

Many of the readers of this book are used to an open and democratic governance system where
there are many ways in which rules are created. Under these conditions, it is not illegal or improper
for individuals to self-organize and craft their own rules for many activities. This may be in stark
contrast to more dictatorial states in the world. At work, in a family, or in a community organization
there are many ways we experience the crafting of rules to improve the outcomes we can expect in
the future. Some of these rules are written down on paper, others are verbal and may be confirmed
by a handshake.
In our analysis of case studies in this book, we make a distinction between rules-on-paper (de
jure) and rules-in-use (de facto). It is not uncommon that in practice, somewhat different rules are
used at the work floor, in the classroom, or on the sports field than those officially written down
on paper. For example, a referee in a soccer match may not stop the game for each possible rule
infraction, but judge whether the infraction is severe enough to stop the flow of the game and enforce
penalties.
Human behavior, including the tendency of humans to comply with rules, is not as predictable
as biological or physical phenomena. Humans are reflexive and have opinions and moral values.
22 Defining Institutions

They may not necessarily obey instructions from others. All rules are formulated in human language.
As such, rules might not always be crisp and clear, and there is a potential for misunderstanding
that typify any language-based phenomenon. Words are always simpler than the phenomenon to
which they refer. In many office jobs, for example, the rules require an employee to work a specified
number of hours per week. How accurately do we need to specify what the employee will be doing?
If the employee is physically at her desk for the required number of hours, is daydreaming about
a future vacation or preparing a grocery list for a shopping trip on the way home within the rules?
Written rules are always incomplete and therefore the very act of interpreting the rules may lead to
different outcomes. Monitoring rule compliance is a challenging activity if rules are not always clear
and fully understood. Thus, when we study an action arena, we will look not only at the official
rules on paper, but also the rules in use. Misinterpretations may lead to differences between the
two. For good performance of institutional arrangements, it is important that the rules are mutually
understood.
The effectiveness of a set of rules depends on the shared meaning assigned to words used to
formulate them. If no shared meaning exists when a rule is formulated, confusion will exist about
what actions are required, permitted, or forbidden. The effectiveness of rules is also dependent upon
enforcement. If rules are perfectly enforced then rules simply say what individuals, must, must not,
or may do. Participants in an action arena always have the option to break rules, but there is a risk
of being caught and penalized. Has the reader ever driven faster than the official speed limit? If
the risk is low, rule breaking might be common. Further, because of the feedbacks in action arenas,
the likelihood of rule breaking can grow over time. If one person cheated without being caught,
others may follow and the level of cheating will increase. This will increase the detection of cheating
behavior and more rigorous rule enforcement might be implemented. If the risk of exposure and
sanctioning is high, participants can expect that others will make choices from within the set of
permitted and required actions.
One of the main benefits that accrue to participants when the majority of people follow the rules
is the increased predictability of interactions. Virtually all drivers in the U.S. use the right side of the
road to drive almost all the time. If such a rule were not obeyed frequently, imagine how difficult
it would be to drive and how ineffective it would be to use the road. Knowing what to expect in
interactions with others vastly improves the performance of many social systems.

Biophysical conditions

As we will see throughout the book, the rules affect all the different aspects of the action arena.
The biophysical world also has an important impact on the action arena. What actions are physically
possible, what outcomes can be produced, how actions are linked to outcomes, and what is contained
in the actors’ information sets are all strongly affected by the environment around any given action
situation. For example, water can’t run up hill. Once you say something, you can’t retract it. The
same set of rules may yield entirely different types of action arenas depending upon the context. We
will discuss many case studies in different application domains in this book that will help recognize
how context affects decision making and the effectiveness of rule configurations.

Attributes of the community

A third set of variables that affects the structure of an action arena relate to the attributes of the
community of which the participants are members. Examples of attributes that might be important
2.6 Critical reflections 23

are the shared values within the community, the common understanding and mental models that the
community members hold about the world in which they live, the heterogeneity of positions within
the community such as class and caste systems, the size of the community, and the distribution of
basic assets within of the community.
The term culture is frequently applied to the values shared within a community. Culture affects
the mental models and understanding that participants in an action arena may share. Differences in
mental models affect the capacity of groups to solve problems. For example, when all participants
share a common set of values and interact with one another frequently, it is more likely that the
participants will be able to craft adequate rules and norms for an action arena. If the participants
have different mental models, come from different cultures, speak different languages, have different
religions, it will become much harder to craft effective institutional arrangements.

2.6 Critical reflections
Institutions are rules and norms that structure human interactions. They are complex and difficult to
study. The Institutional Analysis and Development (IAD) framework helps us organize our thoughts
and direct our questions. The focal element of the IAD framework is the action arena in which
participants interact in an action situation. These interactions lead to outcomes which affect
decisions made in the next iteration. The interactions are affected by the social and biophysical
context in which the action situation takes place.

2.7 Make yourself think
1. Come up with institutions you deal with every day. Some you don’t like, and some that you
do like.
2. Do you think banks should be regulated in their lending practices? What are the key elements
necessary to address this question?
3. What is the most important outcome for you in taking this class?
4. What can explain the fact that people solve problems differently in India as compared to the
U.S.?

2.8 References
Ostrom, E. (2005). Understanding institutional diversity. Princeton, NJ: Princeton University Press.
Key concepts
In this chapter we will:
– Learn how action situations define the structure of interactions
– See that adding individuals to an action situation leads to an action arena
– Dissect the structure of an action situation

3 — Action Arenas and Action Situations

3.1 Action arenas
Whenever two or more individuals are faced with
a set of potential actions that jointly produce out- Terms and conditions
comes, these individuals can be said to be “in” an
action situation. Within an action situation a partic- Do you ever read the terms and conditions of
ipant occupies a certain position. The same partici- products you buy, web services you sign up for, or
pants can interact in another action situation where software you install on your computer? These
they occupy different positions. terms and conditions specify the roles and re-
An action arena combines the action situation, sponsibilities of the various people involved in
which focuses on the rules and norms, with the par- the exchange. Let’s have a look at the terms of
the free service Twitter, which can be found at
ticipants who bring with them their individual prefer-
https://twitter.com/tos. In order to make use of
ences, skills and mental models. The need to distin- Twitter you have to agree with the terms of ser-
guish between action arenas and action situations is a vices. Content is the responsibility of the user.
result of the the fact that when different participants The user agrees that content is stored for a maxi-
occupy positions in the same action situation, this mum of 18 months. Furthermore, the user accepts
may lead to very different outcomes. Put simply, the the services of Twitter as it is delivered and Twit-
action situation remains the same for a given period, ter is not responsible for any damage caused by
the use of Twitter.
but a new action arena is generated every time a new
What these terms define are the roles of the dif-
set of participants enters the action situation. For ex- ferent participants involved, the actions they can
ample, an action situation might be the market place take, the information that they have or do not have,
on eBay. The same product offered by different sell- and the potential outcomes. For example, you are
ers might not lead to the same price since it depends responsible for keeping your password secret (in-
on the preferences and actions of the different par- formation), and any tweet will immediately be
ticipants who enter the action situation and generate available to the whole world (outcome).
a new action arena. Other examples of action situa- The terms and conditions may change over time.
When accounts get hacked, when people claim
tions include resource users who can extract resource
identities of others (like celebrities), or when Twit-
units (such as fish, water, or timber) from a shared ter is used to coordinate riots, terms and condi-
resource, politicians in congress crafting new laws, tions may be evaluated and adjusted to improve
and schools with educators and students. performance of the Twitter experience.
Likewise, the same participants can have very
different types of interactions in different action situ-
ations. This could be the result of the simple fact that the participants in different action situations
26 Action Arenas and Action Situations

Figure 3.1: The internal structure of an action situation (adapted from Ostrom, 2005).

occupy different positions. This could also be due to different rules on the information available
in different action situations. A boss and his employee in one action situation might become two
squash players in another arena. The boss and employee interact very differently in terms of their
power relationship—they leave their professional relationship at the squash court door.
The structure of all action situations can be described and analyzed by using a common set
of variables. These are: (1) the set of participants, (2) the positions to be filled by participants,
(3) the potential outcomes, (4) the set of allowable actions and the function that maps actions into
realized outcomes, (5) the control that an individual has in regard to this function, (6) the information
available to participants about actions, outcomes, and the linkages between them, and (7) the costs
and benefits—which serve as incentives and deterrents—assigned to actions and outcomes. The
internal structure of an action situation can be represented as shown in Figure 3.1. In addition to
the internal structure, whether a situation will occur once, a known and finite number of times, or
indefinitely, affects the strategies individuals adopt. And again, with the same action situation but
different individuals participating, we have a different action arena.
Within a college course, participants have different positions for which different actions are
assigned. Students have different responsibilities compared to the professor and teaching assistant.
For example, a professor has information regarding the scores of all the students and the authority to
give the grades. Students do not have full information about the scores of individual students in the
classroom. They may, however, have aggregate information about the all the student scores (i.e., the
average). A teaching assistant can grade essays based upon an agreed upon evaluation criteria, but it
3.1 Action arenas 27

is the responsibility of the professor to give the grades. Some of the costs and benefits for a professor
include the amount of time spent in preparing the class content and lectures and grading and the
wage she receives for doing so. The consequences of different allocations of time invested can be
seen in the grades the students receive and the evaluations the professor receives. Also the student
has to balance the investments of time in taking the course and other activities and this choice will
be materialized in the grade received.
An individual can take a class one year (be in the position of student), and become a teaching
assistant the next year. That is, the same participant can occupy many different positions. The student
could attend a course in the morning as a student, then act as a teaching assistant for a different
course in the afternoon. In the morning course, the student has no information about other students’
grades. In the afternoon course, she will have more information about the individual students, but
now also bears more responsibility for the performance of the students in the class.
The number of participants and positions in an action situation may vary, but there must be at
least two participants in an action situation. Participants need to be able to make choices about the
actions they take. The collection of available actions represents the spectrum of possibilities by
which participants can produce particular outcomes in that situation. Information about the situation
may vary, but all participants must have access to some common information about the situation
otherwise we cannot say that the participants are in the same situation. The costs and benefits
assigned to actions and outcomes create incentives for the different possible actions. How these
affect the choice of participants depends on the preferences, resources and skills participants have.
Who has power? Not all participants may have the same level of control, allowing some to have
substantial power over others and the relative benefits they can achieve.
There is inequality in wealth between countries and within coun-
tries. There are differences in information access and access to
decision makers between the haves and the have-nots. Poor people
have fewer possible actions available to them than do rich people.
With wealth comes access. The rich man in the Figure 3.2 can polish
his own shoes but can also pay somebody else to do this. The poor
man does not even have shoes that need to be polished let alone
resources to pay somebody else to polish the shoes. This example
shows that not every person can occupy each possible position in
an action situation. Both men can occupy the positions of citizen,
or legal adult, but only the wealthy man can be both polisher and
“polishee.” The fact that the poor man cannot be the “polishee” is Figure 3.2: A poor man polishing
due to one of two factors (1) formal rules or norms about social roles the shoes of a wealthy man.
and occupations allowed for different social roles such as the cast
system in India or (2) because the poor man lacks the resources or capacity to be the “polishee.” In
the first case, it is the action situation that limits the actions of the poor man. In the second case,
it is an outcome of the action arena that limits his choices. Whether an individual can occupy a
certain position may be affected by wealth, education, elections, inheritance, passing a test, age,
gender, and many other criteria. As we will see in later chapters, the rules that affect positions play
an important role in how communities can sustain their commons.
When we study an action situation, we analyze the situation as given. We assume the structure
of the action situation is fixed for at least the short run. Then we can analyze the action arena by
exploring assumptions of the likely human behavior of the individuals leading to particular outcomes.
Within a particular situation, individuals can make choices about their own actions. However, in
28 Action Arenas and Action Situations

the longer term, individuals may—at least those who are living in an open society—take actions that
may eventually affect the structure of action situations (i.e., the choices others can make). This is
possible when one is able to change the rules affecting the action situation. For example, the rules
regarding the marketplace at eBay have changed over time because participants have learned what
works and what does not work. If action situations do not lead to good outcomes, one may attempt
to change the rules. To do so, they must move to action situations at a higher level of decision
making such as collective-choice or constitutional-choice action situations, where the outcomes
generated are changes in the rules that structure other action situations such as who can participate,
what actions are available to them, what payoffs are associated with actions, etc. In a closed society,
individuals at an operational level may have little opportunity to change rules at any level and
may find themselves in highly exploitative situations. Democratic countries are examples of open
societies, while dictatorships are examples of closed societies (see Figure 3.3). We discuss the
process of shifting to higher-level action situations in the last half of this chapter.

3.2 The basic working parts of action situations
Let us now discuss the elements of an action situation so
we can begin to understand what is common to all of the
interactive situations we may observe or experience in our
lives.

3.2.1 Participants
Participants in an action situation are assigned to a po-
sition and capable of making a choice between different
possible actions. The participants in action situations can
be individuals but also corporate actors such as nations,
states in a federal system, private corporations, NGOs, Figure 3.3: North Korea is an example of a
and so forth. Whenever participants are organizations, one closed society. Citizens have no control over
treats them in the situation as if they were a single individ- the decisions they can make, have no control
over the information they receive, and have
ual but one that is linked to a series of additional situations
no ability to change regulations.
within their own organization. When one is interested in
the outcome of an action situation for the organization,
we may ignore the linked situation and just focus on the
strategy of the organization as an individual actor. However, if we notice that there are problems
with the functioning of an organization within an action situation, we may look at the functioning of
the organization itself, and study the action arena of that organization. As such, action arenas can be
composed of action arenas of lower level actors. For example, the United Nations consists of many
countries. To understand the functioning of the action arena of the United Nations, we may look
at the ambassadors as participants, or we may look into the action arena of a country in which the
ambassador also participates to understand the decisions made by that ambassador.
Several attributes of participants are relevant when representing and analyzing specific situations.
These include (1) the number of participants, (2) their status as individuals or as a team or composite
actor, (3) and various individual attributes, such as age, education, gender, and experience.
3.2 The basic working parts of action situations 29

Figure 3.4: Different numbers of players in the action situation: a) two sumo wrestlers, b) eleven players per
soccer team c) 200 cyclists in a stage of the Tour de France, and d) more than 10,000 runners in a marathon.

The number of participants

The focus of this book is the action situations associated with the commons. Therefore, action
situations that are of interest to us require at least two participants where the actions of each affect
the outcomes for both. This could be two farmers sharing a water source.
The specific number of participants is often specified in detail by formal regulations, such as
for legislation (number of seats in the Senate and Congress), juries (number of jury members), and
most sports (number of members on a team). Some descriptions of a situation, however, specify
the number of participants in a looser fashion such as a small or a large group, or face-to-face
relationships versus impersonal relationships. Since many other components of an action situation
are affected by the number of participants, this is a particularly important attribute in the analysis
of any action situation. Figure 3.4 shows some examples from sports illustrating the number of
participants in action situations.

The individual or team status of participants

Participants in many action situations may be individual persons or they may represent a team or
composite actors, such as households. A group of individuals may be considered as one participant
(a team or organization) in a particular action situation. What might be the conditions in which it
makes sense to treat a group of individuals as a participant?
To consider a group of individuals to be a participant, one must assume that the individuals intend
to participate in collective action. One needs to assume that the individuals who are being treated as
a single actor intend to achieve a common purpose. Sometimes there are groups of individuals who
share many similar characteristics, such as “veterans,” “urban voters,” or “legal immigrants,” but
they have different individual preferences and do not act as a cohesive team. Corporate actors, such
30 Action Arenas and Action Situations

as firms, are not so dependent on the preferences of their members and beneficiaries, because they
are legally defined as an individual entity. The activities in firms and organizations are carried out by
staff members whose own private preferences are supposed to be neutralized by formal employment
contracts.
A fully organized market with well-defined property rights, for example, may include buyers and
sellers who are organized as firms as well as individual participants. Firms are composed of many
individuals. Each firm in a market is often treated as if it were a single participant.
So when do we consider a group of individuals as a collective rather than as a bunch of individu-
als? This depends on the questions we have. The action arena of a basketball game, for example,
may be represented as having either ten participants or two teams composed of five individuals. If
we are interested in studying a league or a tournament, we will include more teams and focus on the
teams as participants rather than as individual players. If we are interested in the performance of a
single team within the league we would look at individual players, coaches, trainers, and owners.

Attributes of participants

Participants differ in their characteristics such as their skills, ethnic background, education,
gender, values, etc. These characteristics may influence their actions in some situations, but not in
others. The educational level of participants is not likely to affect the actions of drivers passing one
another on a busy highway. But when the participants meet each other in an emergency room in
the positions of patient and physician, education becomes an important attribute. Whether gender
or ethnic background is important varies between cultures and countries. In some cultures, female
patients are not allowed to be examined and treated by male physicians. During the Apartheid regime
in South Africa, Black patients did not receive the same treatment as White patients did.
The outcomes of many situations depend on the knowledge and skills of the parties. Experienced
drivers will have on average a different driving style compared to younger drivers. This fact is
born out in the differences between insurance policies for the two participants. Drivers who have a
reputation with an insurance company for getting involved in accidents will have to pay a higher
insurance premium compared to those who are accident free.

3.2.2 Positions
Participants occupy positions in action situations. Examples of positions include students, professors,
players, referees, voters, candidates, suspects, judges, buyers, sellers, legislators, guards, licensed
drivers, physicians, and so forth. It is very important to understand that “positions” do not refer to
people, but rather to roles that participants can play in an action situation. For example, in a market
situation (in your local mall), the same person may be a “seller” when she is at work at the Apple
Store helping customers choose their latest iPhone, and a “buyer” when she goes for lunch in the
food court. Thus, positions and participants are separate elements in a situation even though they
may not be clearly so identified in practice.
In practice, the number of positions is frequently significantly less than the number of participants.
In a class there are typically only two positions—student and professor—while there may be hundreds
of participants. Hunters who have a valid license all occupy the position of a licensed hunter; and
while there are more than a billion participants at Facebook, there are only a limited number of
positions (such as the person featured on the webpage, or the administrator of a page representing an
organization).
3.2 The basic working parts of action situations 31

Figure 3.5: a) Barack H. Obama in the position of President of the United States of America, b) the U.S.
Suppreme Court, c) judges on the television show American Idol, and d) a police officer.

Depending on the structure of the situation, a participant may simultaneously occupy more than
one position. All participants will occupy whatever is the most inclusive position in a situation —
member, citizen, employee, and the like. In a private firm, additional positions such as foreman,
division manager, or president will be occupied by some participants while they continue to occupy
the most inclusive position—that of employee. Some examples of positions are given in Figure 3.5,
where some positions are filled by election, others are filled by selection after an interview.
Positions connect participants with potential actions that they may take in an action situation.
Not all positions have the same potential actions. A surgeon can do surgery on a patient. It is likely
not advisable to allow the patient this potential action, unless in the unlikely event that the patient
happens to be a surgeon too. Other positions are less restrictive. Every person with a driver’s license
shares a large set of potential actions with every other driver. Some drivers have a special positions
and additional potential actions, such as drivers of ambulances or large trucks.
The president of the U.S. can sign a bill into new legislation, which confirms that the new
legislation will be implemented. The president can only sign such a document under particular
conditions (agreement in the Senate and Congress), but a signature of a regular citizen does not have
the same effect. A U.S. citizen who has registered as a voter can vote, but a permanent resident
(Green Card holder) cannot register as a voter, even if such a permanent resident is a professor at a
prominent university.
The nature of a position assigned to participants in an action situation both defines the set of
authorized actions and sets limits on those. For example, licensed drivers may operate a motor
vehicle on a road or highway, but this action is also restricted by speed limits. Those who hold the
position of a member of a legislative committee are authorized to debate issues and vote on them.
The member who holds the position of chair can usually develop the agenda for the order of how
32 Action Arenas and Action Situations

issues will be brought before the committee or even whether a proposal will even be discussed. The
order of events on this agenda may affect how the votes turn out.
If you do a group project how do you organize a group? Will different group members have
different roles? Is one of the members leading the discussion?
Participants may occupy different positions, but which position one can be hold is not always
something a participant can choose. A defendant in a criminal trial does not control her movement
into or out of this position. A candidate for the U.S. Congress can certainly influence her chances of
winning an election and securing the position, but does not have full control. In the end, this decision
is in the hands of voters. Holding the position of a pedestrian in traffic is available without much
limitation to most people. Individuals have to compete vigorously for getting a tenured professorship
at a university, but once obtained, they may hold their positions for life, subject only to legal actions.
This might be true for universities in the U.S., but in most European countries professors are required
to retire at the age of 65 and are removed from their position as a tenured professor.

3.2.3 Potential outcomes
In the case of health care reform, there are different poten-
tial outcomes that can be discussed: total costs of health
care, access to health care, distribution of costs and bene-
fits, quality of health care, etc. Which outcome will weigh
most in the design of policies is a political decision.
When we want to understand how rules, attributes of
the environment, or attributes of the community change
an action situation, careful attention must be given to how
participants value certain outcomes. If there is a market
where goods are exchanged at known prices, one could
assign a monetary value to the goods. If there are taxes
imposed on the exchange of goods (a sales tax), one could Figure 3.6: The BP Deepwater Horizon oil
represent the outcomes in a monetary unit representing rig ablaze.
the market prices minus the tax. If one wanted to examine
the profitability of growing rice as contrasted to tomatoes or other cash crops, one would represent
the outcomes in terms of the monetary value of the realized sales value minus the monetary value of
the inputs (land, labor, energy, fertilizer, and other variable inputs).
To examine the effect of rules, one needs to distinguish the effect of material rewards from
financial values. For example, the physical amount of goods produced during a particular time period
is different than the financial rewards to workers and owners for that time period. If no goods are
sold, the financial rewards for the owner might be negative, but the worker may still receive a reward
in exchange for the hours worked to produce the goods. Besides monetary values and physical
quantities of goods, participants also have internal values, such as moral judgment, that they use to
examine potential outcomes. Gun ownership can be evaluated based on the numbers of different
types of guns owned, the monetary value of the gun collection and the moral value placed on gun
ownership.
Frequently the outcomes are assumed to be the consequence of self-conscious decisions, but
there can also be “unintended outcomes.” For example, oil spills in the Gulf of Mexico are not an
intended outcome of operations of oil companies (Figure 3.6).
3.2 The basic working parts of action situations 33

3.2.4 Actions
Participants assigned to a position in an action situation
must choose from a set of actions at any particular stage
in a decision process. An action can be thought of as
a selection of a setting or a value on a control variable
(e.g., a dial or switch) that a participant hopes will affect
the outcomes. The specific action selected is called a
choice. A complete specification of the actions, taking in
all possible variations of the action situation is called a
strategy. It is important to note that it might not always
be clear to participants what all the valid actions are in
an action situation. A switch may clearly indicate two
different positions, but sometimes participants are much Figure 3.7: Possible actions.
more innovative in the use of possible actions in an action
situation. Calling somebody with a mobile phone can be very expensive in some places. As a result,
people may use the technology in a different way than was intended by the manufacturers. For
example, in some communities, signaling systems evolved where the receiver of a mobile phone call
can understand the message from simply counting the number of rings of the call and the information
about who the caller is from the display. Users in such situations may seldom use their mobile phone
for an actual voice call. How could mobile phone service providers change this outcome?

3.2.5 Control
The extent to which participants have control over aspects of the action situation vary widely.
Obviously, the position that a participant occupies affects the power of this participant (her ability
to affect the actions of other participants and outcomes). The level of control a participant has can
therefore change over time, for example if she changes her position. Barack Obama acquired a new
repertoire of actions and control when he assumed office on January 20, 2009. And this repertoire
has changed over the years from being a lecturer and giving grades at the University of Chicago,
a community organizer, a member of the Illinois Senate, and a member of the U.S. Senate. Each
position held certain duties and rights. Leaving a position also means losing the duties and rights
that hold to the specific position.

3.2.6 Information about the action situation
What is the information participants have in an action situation? In an extreme case they have
complete information and know the number of participants, the positions, the outcomes, the actions
available, how the actions are linked to outcomes, the information available to other players and the
payoffs available. If they know what other participants will do, participants are said to have perfect
information. Of course, perfect information is an extreme case especially when people make their
decisions privately. Often there is no perfect understanding of how actions will lead to outcomes, or
what others plan to do. Even if people communicate and negotiate what everyone will do, the actual
actions may turn out different since people make mistakes or cheat.
In many situations there is asymmetric access to the available information. For example, in
work situations, a boss cannot know exactly what employees are doing. That is why providing
an incentive to increase productivity is a challenge. The same holds for insurance companies.
Your insurance company does not have perfect information about your driving abilities and health
34 Action Arenas and Action Situations

conditions, but makes an informed guess based on statistics of historical events. Would you like your
insurance company to have access to your genetic profile? What about your driving behavior? The
car insurance company Progressive allows customers to join a voluntary program where a device
is installed in your car to track your driving style. One can save a significant amount on their car
insurance with proper driving style.

3.2.7 Costs and benefits
To evaluate the outcomes of the actions taken in the action situation
we have to look at the costs and benefits. These costs and bene-
fits accumulate over time. Not all participants will experience the
same costs and benefits. Sometimes the positions that participants
hold affect their cost and benefits since it affects the compensation,
penalties, fees, rewards and opportunities. A physician receives
a monetary benefit from doing a treatment while the patient will
pay to receive an improvement in their health condition. Even if
participants hold the same position, like players on a sports team,
their rewards vary as defined by their individual contracts.
If we study action arenas we need to make a distinction between
the physical outcome and the valuation that a participant assigns to
that outcome. In economics, the value assigned by participants is
often referred to as utility. Individual utility is a summary measure Figure 3.8: The cost of speeding.
of all the net values to an individual of all the benefits and costs of
the outcome of a particular action situation. Utility might increase with an increase in profit, but
depending on the study at hand, it may also include elements like joy, shame, regret and guilt.
For example, driving above the speed limit can save you time. However, if you are caught you
will have to pay a traffic fine (Figure 3.8). You may challenge a ticket by appearing in court, yet
this will take time and may have other costs associated with it. Paying the fine (accepting guilt for
the traffic violation) could also result in higher car insurance and accumulate points on your driving
record. If you accumulate too many points on your driving record, your license may be suspended.
Not paying a ticket in time will lead to additional penalties.

3.2.8 Linking action situations
In reality people make decisions in different action situations that are often linked together. Rarely
do action situations exist entirely independently of other situations. For example, new laws in the
U.S. need to be approved by the Congress and the Senate before the president may sign it. Signing a
bill is meaningless unless the bill has successfully passed through the Congress and Senate action
arenas.
Given the importance of repeated interactions to the development of a reputation for reciprocity
and the importance of reciprocity for achieving higher levels of cooperation and better outcomes
over time, individuals have a strong motivation to link situations.
Action situations can be linked through organizational connections. Within larger organizations,
what happens in the purchasing department affects what happens in the production and sales
department and vice versa. Sometimes action situations are structured over time. For example,
a tournament or sport competition is a description of how players (e.g., tennis) or teams (e.g.,
basketball) will proceed through a sequence of action situations. In other examples, action situations
3.2 The basic working parts of action situations 35

Figure 3.9: Different levels of rules related to a university class (clockwise): a) day-to-day assignments; b)
syllabi are approved by committee; c) university senate approve new degrees; and d) upper administration
make decisions that have university-wide impact, such as tuition rates.

are not formally linked. Farmers who have successful innovative practices in deriving better profits
are frequently copied by others.
Another way in which action situations can be linked is through different levels of activities. We
can distinguish three levels of rules that cumulatively affect the actions taken and outcomes obtained:

• Operational rules directly affect day-to-day decisions made by the participants in any setting.
These can change relatively rapidly—from day to day.
• Collective-choice rules affect operational activities and results through their effects in de-
termining who is eligible to be a participant and the specific rules to be used in changing
operational rules. These change at a much slower pace.
• Constitutional-choice rules first affect collective-choice activities by determining who is
eligible to be a participant and the rules to be used in crafting the set of collective-choice
rules that, in turn, affect the set of operational rules. Constitutional-choice rules change at the
slowest pace.
An example of an operational-level situation is a group of fishers who decide where and when to
fish. At the collective-choice level the group of fishers may decide on which seasons or locations
to implement bans on fishing. At the constitutional-choice level decisions are made regarding the
conditions required in order to be eligible for membership in the group of fishers.
Figure 3.9 illustrates the different levels of rules related to a class at a university. Within a
classroom, decisions are made based on the rules set in the syllabus. Day-to-day decisions include
what the assignments are for next week, who will give a talk, and when students can come to office
hours. In order for a regular course to be approved, a committee (upper right) will review the
proposed syllabus and make a recommendation to approve or not approve the course. The committee
also solicits comments of departments that provide similar courses to avoid potential conflicts. The
university senate will come into play when new degrees are proposed (middle left). Finally, the upper
administration of the university will be involved in decisions that have university-wide impact, such
36 Action Arenas and Action Situations

as changing tuition rates. Such a tuition raise will have to be approved, at least for public universities,
by a state level committee.

3.2.9 Outcomes
It is difficult to predict the outcomes of rule changes made in action situations. Changing the rules in
one action situation may have consequences in other action situations. The difficulty of predicting
the consequences of changes shows that we have to closely observe what is happening before rules
are changed and after rules are changed. This suggests that we should view policies experiments,
and closely observe these experiments in order to learn and have a better understanding of what will
happen in a similar case in the future.
Besides the difficulty of predicting outcomes, how to evaluate outcomes is also often not
immediately evident. There are different criteria that one can use to evaluate the outcomes:
• Economic efficiency—what are the costs relative to the benefits?
• Equity—how are costs and benefits distributed among the participants?
• Accountability—are participants in leadership positions accountable for the consequences of
their decisions?
• Conformance to general morality—are the procedures fair, is cheating detected, and are
promises kept?
• Sustainability—how do the outcomes evolve over time? And what are the consequences of
decisions on the underlying system?
In order to evaluate the outcomes one needs to evaluate trade-offs associated with the different
criteria. If some groups are affected differently than others, it will be important to define procedures
in the collective-choice or constitutional-choice rules to address such differences. For example, the
outcome of changing the criteria for student-loans will not have the same consequence for each
individual student. It would be important to consider the different types of outcomes for different
types of participants and develop agreements regarding how to evaluate such outcomes.

3.3 Critical reflections
The concept of action arenas was the main topic of this chapter. An action arena consists of an
action situation that defines the structure of interactions, actions and outcomes, and the individuals,
organizations or nations who may participate in the action situation. When two or more participants
interact, there is an action arena where participants hold positions, and can make decisions. Not
everybody in an action situation can take the same actions, or has the same level of information.
The consequences of the actions are the outcomes of the action situation, which can be evaluated
differently by each participant in the action situation.
3.4 Make yourself think 37

3.4 Make yourself think
1. What positions do you hold in different action situations? Provide some examples.
2. What is an action situation you experience regularly? What are the possible outcomes in this
situation? What actions can you take? Be sure to distinguish between actions you may take
and choices you do make.
3. Do you have an example from your own personal experience where you have experienced the
same action situation but with different participants that led to a different action arena and a
different outcome?

3.5 References
Ostrom, E. (2005). Understanding institutional diversity. Princeton, NJ: Princeton University Press.
Key Concepts
In this chapter we will:
– Learn the concept of social dilemmas
– See what selfish and rational individuals do in social dilemmas
– Be introduced to coordination games
– Learn to recognize different economic goods and resources
– See how the Institutional Analysis and Development framework
can be used to study social dilemmas

4 — Social Dilemmas

Buying online

In the early 1990s the New Yorker published a cartoon
with the saying “On the Internet, nobody knows you’re a
dog” (Figure 4.1). The cartoon illustrates the fact that it
is not always possible to verify someone’s identity on the
internet. When you buy a product online, how does the seller
know who you are??
Making a purchase through the internet service eBay
can lead to challenging dilemmas. You put a bid on an iPad
as depicted in the image shown on the eBay website, but
how can you trust the product is working properly? If you
are the highest bidder, how does the seller know you will
send the money?
eBay tries to solve this problem by using a reputation
system. Each time you have had a transaction you can earn
points for your reputation if the other participant will give
feedback on your performance. Since a good reputation
will make it more likely that people will want to buy from
you or sell to you, it is important to keep a good reputation. Figure 4.1
Professional sellers will remind you to submit feedback so
that they can improve their reputation.

4.1 Introduction
In many action situations we experience every day, there often exists a dilemma between what is
best for the individual and what is best for the group. For example, it would be beneficial for an
individual to be able to use a siren so that all the other cars on the road will pull over, allowing the
one with the siren to get from point A to point B as fast as possible. It would also be beneficial to an
individual not to pay taxes. Nevertheless, society would not function if everybody used a siren when
they drove or did not pay their taxes. In general, society will not function if individuals do only what
is best for themselves alone. This is called a social dilemma.
The study of social dilemmas is a very prominent topic in the social sciences. In this chapter we
will discuss different types of social dilemmas and describe them from the perspective of the action
situation we covered in the previous chapter.
40 Social Dilemmas

We will see that social dilemmas lead to the prediction that people will not contribute to the
common good or will overharvest from shared resources. But as we will see in later chapters, these
predictions are based on a very narrow notion of human behavior, namely that everyone behaves
as selfish rational beings. In later chapters we will discuss behavioral studies that show that this
is not the case. Nevertheless, it helps to think about social dilemmas using the naïve model of
decision-making. It will point to some potential problems related to action situations.
The very existence of Wikipedia demonstrates that we do not
all behave as selfish rational beings. (Figure 4.2). Many people
now use Wikipedia because it is a very powerful resource with high
quality information on many topics. There are fewer people who
write articles for Wikipedia than there are people reading Wikipedia
articles. Those people who only consume Wikipedia “free ride” on
the contributions of others. Those who write articles spend time
(experience a cost) to do so. Luckily there are enough people willing
to make a contribution voluntarily in order to have a very useful
product. If people are fundamentally selfish, why do you think these
individuals contribute to Wikipedia?
Have you ridden a bicycle to school lately? Decided not to go
on vacation? Or switched to a vegetarian diet? Some people have
elected to adopt these types of changes in their lives in order to
Figure 4.2: Wikipedia logo.
reduce greenhouse gas emissions and reduce the possible impacts
of climate change. However, many more people will need to change
their behavior in order to solve this social dilemma than are currently doing so. Although policy
makers have discussed climate change for more than 20 years, global emissions continue to increase
almost every year. This is an example of a social dilemma (and there are many like it) that cannot be
solved by voluntary contributions alone, indicating that regulations to stimulate action (e.g., lower
energy use) are needed. That is why we will talk about rules and norms in later chapters. We may
need to use rules and norms to overcome social dilemmas.

4.2 Prisoner’s dilemma
A classic description of a social dilemma is captured in the prisoner’s dilemma. This dilemma was
originally developed in 1950 by Merrill Flood and Melvin Dresher working at RAND. The basic
idea is as follows: There is a riot leading to plundering in the city by youngsters and two suspects are
arrested by the police. The two suspects did not know each other before the riots and the police do
not have sufficient evidence for a conviction. Having separated the prisoners, the police visit them
alone offering each one a deal for testifying against the other. If one testifies for the prosecution
against the other (defection) and the other suspect remains silent (cooperation), the defector goes free
and the silent accomplice receives the full one-year sentence. If both remain silent, both suspects
will receive one month in jail for a minor charge of, say, disturbing the peace. If each betrays the
other, each receives a three-month sentence. Each prisoner must choose to betray the other or to
remain silent. Each one is assured that the other does not know the decision before the end of the
investigation. How should the prisoners act?
We can apply some formal methods to reason through what the best strategy is for the prisoners.
But in doing so, we first make some assumptions. Both prisoners want to minimize the amount of
time in jail for themselves and they both reason rationally. Both prisoners also assume that the other
4.3 Social dilemmas as an action situation 41

prisoner reasons like she does. If the first prisoner remains silent, the best decision for the second
prisoner is to betray so she can go free. And if the first prisoner testifies against the second, it will,
again, be best for the second to betray. So, whatever the first prisoner does, the best decision for
the second is to testify and betray the other. Since both prisoners reason in that way, the expected
outcome will be that both betray and both will spend three months in jail.
Obviously, the best solution for both of them would be to remain silent so that both only receive
one month in jail. Hence when each reasons what is best for them individually and acts on it, we
arrive at an outcome that is not the best for the both of them. The assumptions we have had to make
to arrive at this conclusion, however, are pretty restrictive. What have we left out regarding the
behavior of the prisoners that might change the outcome?
In Table 4.1 we present the different possible outcomes of this action situation in a so-called
payoff table. Depending on what the two individuals decide, there are four possible outcomes. In
each cell of the payoff table the payoffs for each outcome is defined.

Prisoner A
Remain Silent Betray
Prisoner B Remain silent One month in prison for both Prisoner A is free and Prisoner
B gets one year
Betray Prisoner B is free and Prisoner Three months in prison for both
A gets one year

Table 4.1: Contextual payoff table for the prisoner’s dilemma

This payoff table is often represented in a more abstract way. The story about the prisoners is
only one illustration. This kind of social dilemma can be found in many other examples. We often
represent the payoff table of this type of action situation, often called a two-player game, by the
numerical outcomes for player A and B. In Table 4.2 the outcome (-1, -1) means that both players
lose one unit. In this case, currency for the payoff is time. In the original prisoner’s dilemma story
both prisoners loose one month by going to jail. In other situations, the payoff could be in the
currency of water, money, or other valuable resources.

Player A
Cooperate Defect
Player B Cooperate (-1,-1) (0,-12)
Defect (-12,0) (-3,-3)

Table 4.2: Payoff table for the prisoner’s dilemma

4.3 Social dilemmas as an action situation
As alluded to earlier, we can view the social dilemmas described above from the perspective of an
action situation. Let’s use the ideas we have discussed so far to dissect the prisoner’s dilemma. There
are two participants who have equal positions. Both participants have a limited set of actions—they
can choose between actions A or B. They have limited information about the other participant since
communication is not possible. Both players know that the other player is in the same situation she
42 Social Dilemmas

Figure 4.3: The countries in red indicate where cars drive on the right-hand side of the road, while blue
indicates where cars drive on the left-hand side of the road.

is in, but do not know what the other player will do. The players only have control over their own
actions. The outcomes are defined clearly in the payoff table, and the evaluation of the outcomes is
based on the material benefits for the players.
If you were in a prisoner’s dilemma what would you do? Would it make a difference if you
knew the other person, even if you cannot communicate? What does this say about information in an
actual action arena? Indeed you can be in the same action situation with two different individuals
and have different expectations what the other will do.

4.4 Coordination games
Although the focus of this chapter is on social dilemmas, a related phenomenon is coordination,
which can be expressed in terms of a “coordination game.” Suppose that while driving a car on a
dirty road you decide to drive in the middle of the road since that provides the cleanest surface. Then
you see another car coming from the other direction. To avoid a head-on collision, what will you do?
You have two options. You can go left or right. If both drivers go right (i.e., from their perspective;
from an onlookers perspective, one goes right, one goes left), they will avoid a collision. The same
is true if both drivers go left.
Most countries have an established rule for driving on a particular side of the road (Figure 4.3).
If you are accustomed to driving on the right side of the road on the mainland of Europe, it can
be confusing to drive in the United Kingdom where people drive on the left. Crossing a road as a
pedestrian in London can be dangerous if you are not used to cars driving on the left side of the road!
Not all countries have historically used the same side of the road for driving. Sweden, for
example, switched from left to right on the 3rd of September 1967 at 5:00 am. The accident rate
dropped in the years immediately after the switch, but has since bounced back to levels similar to
before the switch. It may be that this pattern is a result of the fact that in the time just after the switch,
everybody was more careful in driving since they were used to driving on the left side.
Table 4.3 shows the payoff table for a coordination game. Unlike a social dilemma, individuals
cannot free ride on the cooperation of others. If both players drive on the same side of the road in
opposite directions, both players receive 10 points. But if they drive on different sides, both will
receive 0 points. If both players cannot communicate, it is very difficult to coordinate. In many
4.5 A typology of goods 43

societies there are social norms that facilitate the coordination. Think about the ways people greet
each other when they meet. If you meet people from different cultures, there might be some initial
confusion; shaking with the right hand, giving one kiss on the cheek, or two, or three, giving a hug,
or rubbing noses.

Player A
Left Right
Player B Left (10,10) (0,0)
Right (0,0) (10,10)

Table 4.3: Payoff table for the coordination game

The next example we discuss is called the “battle of the sexes” (Table 4.4) game. It illustrates
the dilemma faced by two players who want to go out at night, but have different preferences. Player
A wants to see a movie, but player B wants to go to the opera. Assuming that player A and B, for
whatever reason, cannot communicate and have to make a decision where to go, the possible set of
actions is movie or opera. If both players do the same, both will receive points, but one will be more
satisfied than the other. However, if they end up in different locations, neither will have an enjoyable
evening, and both receive zero points.

Player A
Movie Opera
Player B Movie (5,10) (0,0)
Opera (0,0) (10,5)

Table 4.4: Payoff table for the battle of the sexes game.

A practical example of the battle of the sexes would be a negotiation between two stakeholder
groups. For example, a farmer would like to have wind mills on her land to stimulate clean energy.
The energy would be available at a low cost to the residents of the nearby town. However, the
homeowners do not want their views polluted by wind mills and would prefer to buy the land from
the farmer at a price that is higher than the market value, as long as they don’t have to look at the
wind mills. Whatever option is agreed upon, one of the parties does not receive the best possible
outcome. But worse would be that they do not come to any agreement and both parties loose.

4.5 A typology of goods
Many social dilemmas are related to the production or consumption of goods. In this section we
describe different types of goods and how rules relate to them. We can use two attributes, exclusion
and subtractability, to distinguish four basic goods and services (Table 4.5).
Exclusion relates to the difficulty of restricting those who benefit from the resource or service.
Subtractability refers to the extent to which one individual’s use subtracts from the availability of a
good or service for consumption by others. Different levels of exclusion and subtractability define
different types of goods. So what do these dimensions actually mean for goods and services? Let’s
discuss the different dimensions in more detail.
44 Social Dilemmas

• Subtractable: the means that the use of a good or service by one participant in an action
arena reduces the availability of the good or service to another participant.
An example is the use of a fossil fuel like oil. This non-renewable resource is used for many
energy sources. The gallons of gasoline that you put in your car cannot be used by somebody
else after you have burned the gasoline during your trip. A more direct example is a cake
you have made for your friends. Every piece of cake eaten by one person is not available for
somebody else. This is related to the old adage “You can’t have your cake and eat it too.”
• Non-subtractable: The consumption or use of the good or service by one participant in
the action arena does NOT reduce availability or utility of the good or service to another
participant.
Examples: Reading an article on Wikipedia does not reduce availability of the article for
somebody else. Many information goods like movies, photos, books, and scientific knowledge
have this property.
• Excludable: Any excludable good or service is one that a participant can be prevented from
accessing if they do not pay for it, or have passed another form of entry barrier.
Examples. Going to the movies requires you to buy a ticket to get access. Some websites
require you to sign up, like Facebook. You can be excluded if you are too young or have
misbehaved in the past.
• Non-excludable: Any good or service that a participant cannot be prevented from accessing
or it is extremely expensive to exclude.
Examples. Public roads are available to all cars, even though not all participants pay taxes
for the maintenance. It is very costly to deny access of fishing boats on the oceans, especially
outside the areas of control zones of countries. This makes an ocean fishery very difficult to
regulate.

When we combine these two dimensions we can define four different types of goods (Table 4.5):
private goods, club goods, public goods and common-pool resources. We will discuss below
examples of each of these categories.

Excludable Non-Excludable
Subtractable Private goods Common-pool Resources
Non-Subtractable Club Goods Public Goods

Table 4.5: Four basic types of goods.

Let’s start with private goods. One can restrict the use of the good easily and when that good is in
use, someone else cannot use it. For example, a mobile phone or a car is a private good (Figure 4.4).
You can restrict the use of these goods by having password protection or a key. If somebody else
takes the private good you can go to the police to report a theft. If you are driving your car, no one
else can drive it. If you are making a call on your mobile phone, others cannot use it during your call.
Hence private goods are the typical products we own as an individual. Can you think of examples of
private services?
4.5 A typology of goods 45

The second type is club goods. Compared to private goods,
the use of the club good by one person does not affect the use
by others. Examples of these so-called club goods are satellite
television, cinemas, and toll roads (Figure 4.5). Access is restricted
to “members of the club,” but those members can consume the same
product. For toll roads it means that you pay a fee to use the road,
and moviegoers pay a price for a ticket to watch a movie during a
showing.
The provision of goods is more difficult with those for which it
is challenging to exclude consumers. Public goods—open source
software, Wikipedia, national defense—can be used by everybody,
and use by some does not reduce the ability of others to use it.
The challenge associated with providing public goods is having
a sufficient number of people to invest in their provision. There
is a temptation to “free-ride” on the contributions of others. Due
to the potential for free-riding, there might be an underinvestment Figure 4.4: Smart phone.
in public goods. Ask yourself: are the roads in your city in good
enough condition and are there enough of them? Are there enough parks and open spaces?
A public road is open to all, but not everybody has
provided a contribution to its construction. On a toll road
each user must pay to get access to the road, but a public
road can be used by those who have not made any contri-
bution (by paying local taxes). The same comparison is
true in pay-per-view versus public television. The same
physical product can be offered as different types of goods,
by changing the rules governing who has access to the
good. A public library is a public good for those who want
to read a book or newspaper in the library. If you want
to take a book home for a limited amount of time, you
need to become a member of the library. But consider for Figure 4.5: Access to toll roads is restricted
a moment: if you are reading a book in the library, can to those who are willing to pay a fee.
another patron use that book? No. So how then is the
library a public good? In fact, the story is a bit more complicated as we shall see.
The final example is a common pool resource, such as a lake, an ocean, an irrigation system, a
fishing ground, a forest, the Internet, or the stratosphere. These are natural or constructed resources
from which it is difficult to exclude or limit users once the resource is provided by nature or produced
by humans. One person’s consumption of resource units, such as water, fish, or trees, removes
those units from what is available to others. Thus, the trees or fish harvested by one user are no
longer available for others. The Internet has a limited capacity to move all the information around.
Bandwidth that is used by some to watch movies may cause delays in the sending of email by others.
When the resource units produced by a common-pool resource have a high value and institutional
rules do not restrict the way resource units are appropriated (an open-access situation), individuals
face strong incentives to appropriate more and more resource units eventually leading to congestion,
overuse, and even the destruction of the resource itself. Because of the difficulty of excluding
beneficiaries, the free-rider problem is a potential threat to efforts to reduce appropriation and
improve the long-term outcomes achieved from the use of a common-pool resource. If some
46 Social Dilemmas

individuals reduce their appropriation levels, the benefits they generate are shared with others
whether the others also cut back on their appropriation or not. Some individuals may free-ride on the
costly actions of others unless ways are found to reduce free-riding as an attractive strategy. When
free-riding is a major problem, those who would be willing to reduce their own appropriations for
the benefit of all, provided others would reduce as well, become unwilling to make such a sacrifice
for the benefit of a large number of free riders.
Space is a resource too (Figure 4.6). You
have, no doubt, experienced the heavy appropri-
ation of space on the road during rush hour. If
enough people drive to work earlier or later than
rush hour, this would free up space for others
motorists. But who wants to arrive an hour early
to work? Or take the bus—an option that might
not be convenient for everybody as many people
prefer the privacy and control of driving their
own car. Hence, the next time you experience
a rush hour traffic jam, think about the options
everybody has. Due to the free riding behavior,
all of the drivers experience lower performance
of the road. Figure 4.6: Rush hour in traffic.
Let us return to the question of how (or
whether) the library is a public good. We have also said that roads are a public good. But we
have also mentioned the fact that roads and libraries can become congested. Doesn’t that violate
the non-subtractability of the resource? It turns out that there are very few examples of pure public
goods or any of the other goods for that matter. Typically public goods provide multiple streams of
services with different characteristics. Further, they must typically be combined with other types
of goods to produce a final service. For example, a road isn’t much good without a private car,
public bus, private scooter, etc. Thus, it is better to think of a road or a library as public infras-
tructure that makes available common-pool resources. In the case of the road, the common-pool
resource is transportation space (or capacity). In the case of the library, the common-pool resource is
book-contact-time. The reasonable use of these common-pool resources typically involves solving
coordination dilemmas in space and time. Sometimes, three or more types of infrastructure (goods)
are combined. Take, for example, the iPhone.
The iPhone itself is a private good. It is useless, however, without a network. The contract
you have with the telephone company enables you to use their infrastructure (a club good) to make
calls. Many apps can be downloaded for free and used on an iPhone. The ecology of free apps
really are pure public goods since they are available without restriction, at least if you have the
proper technology to run it. Finally, if many people use their iPhone to watch movies, the wireless
bandwidth gets crowded. This is exactly like cars crowding a road: information bits crowd fiber optic
and copper cables. The so-called “bandwidth” is a common-pool resource that is made available by
telecommunication infrastructure and if too many people start downloading movies the available
bandwidth is completely consumed, and the wireless signal is not available for anyone.
As we conclude this chapter, we encourage the reader to reflect on how different types of
goods (which we can also refer to as infrastructure) must be combined to produce any final service.
Yosemite is a “public good.” But unless it is combined with roads, cars, or planes, what services
can it provide? For some, it does provide a truly pure public good simply in the idea that such a
4.6 Critical reflections 47

beautiful place exists. For most, Yosemite is, in fact, a piece of public infrastructure that provides a
common-pool resource consisting of “nature viewing time” exactly as the library is a piece of public
infrastructure that provides a common-pool resource consisting of “book viewing time.” Understand-
ing the subtle nature of how almost all final services are provided by a complex combination of the
different types of goods we have discussed in this chapter is a critical prerequisite for designing
institutions to wisely govern their use.

4.6 Critical reflections
Social dilemmas are situations in which two or more participants can benefit collectively from
cooperation, but an individual who is selfish and rational can also benefit from free-riding on the
cooperation of others. Four types of goods can be distinguished based on the extent to which it is
possible to exclude others and the subtractability of the resource: private goods, toll (or club) goods,
public goods, and common-pool resources. In this book we will mainly look at problems of public
goods (underinvestment) and common-pool resources (overuse).

4.7 Make yourself think
1. Have you ever been in a social dilemma? Describe the dilemma and what you decided to do.
2. What are examples of public goods and common-pool resources you have experienced lately?
3. Why is access to a resource a key component in defining the four types of goods?

4.8 References
Ostrom, E. (1990). Governing the commons: the evolution of institutions for collective action.
Cambridge University Press.
Part II

Case Studies
Key Concepts
In this chapter we will:
– Be introduced to social dilemmas related to water
– Learn about long-lasting successful water governance systems
– Learn that many successful systems developed effective institutions
without central governmental control
– See that water is often governed at multiple levels
– Notice that access inequality is a core problem for many institutions

5 — Water Governance

5.1 Introduction
Water Governance is the first of many examples that we will present in the remainder of this
book on “sustaining the commons” or, unfortunately “not sustaining the commons” in many cases.
We are finally in a position to give a precise definition of the commons. There are two kinds of
commons: commons provided by nature, and commons provided by humans. Often these two types
of commons interact (e.g., a community irrigation system). The most important distinguishing
feature of a “commons” is that it involves one or several common-pool resources over which there
are no established private property rights. The second distinguishing feature of a “commons” is that
it involves actions on the part of the individuals who use it to maintain its productivity. These actions
are one of two sorts: (1) making a contribution of real resources (money, time, or physical capital),
or (2) restricting one’s activities with respect to the use of the common-pool resource associated
with the commons (e.g., voluntarily fishing less). Actions of type 2 are typically associated with
renewable resources and in this case, are functionally equivalent to actions of type 1: by not fishing,
fishers allow the fishery to regenerate itself. This is the same as investing time and effort into
increasing the productivity of the fishery. In general, reducing appropriation of a renewable resource
is equivalent to investing in natural infrastructure (e.g., a marine ecosystem) that makes the common-
pool resource of interest available (fish). The subject of this chapter, water, is a combination of both
kinds of commons, the natural infrastructure of watersheds and climate systems and the human-made
infrastructure of dams, canals, pumps, and treatment plants. Governing the water commons is
arguably the most important issue we face today.
We use water every day to drink, to cook, to shower, to wash, to flush, to clean, to irrigate, to
swim in, etc. The importance of water makes it a central object of many social dilemmas: too little
clean water, too much polluted water, unequal distributions of water, and natural disasters.
This chapter focuses on water governance. We will discuss some main themes in this field and
discuss some in-depth cases to illustrate examples of success and failure when groups attempt to
solve the social dilemmas associated with water.
The first theme is water scarcity. Large areas around the world suffer from lack of water to
meet the demand for human activities (Figure 5.1). There are, of course, different types of scarcity.
The areas denoted with economic water scarcity do not have sufficient investment in infrastructure
(delivery pipes, pumps, etc.) to satisfy the demand for water even though there may be plenty of
water available. In this case, the human-made public infrastructure (a type of public good described
in the previous chapter) is the limiting factor. Those denoted with physical water scarcity, on the
52 Water Governance

Figure 5.1: Water scarcity.

other hand, simply don’t have enough water available. In this case, it is the common-pool resource
(another type of good described in the previous chapter) that is the limiting factor.
To overcome water scarcity in a particular region, water is often imported from other regions
and often groundwater resources are tapped in the region with the water deficit. This generates
dependencies on other regions and leads to the depletion of groundwater resources. For example,
the main urban areas in Arizona are dependent on the Colorado River for their water needs. Due to
recent drought conditions, there is less water available in the Colorado River. Right now, Arizona
has plenty of water for urban areas as the bulk of Arizona’s water, now being used in agriculture, can
be transferred to urban uses. However, if the region suffers a prolonged drought, eventually conflict
could erupt between the states that use the water in the Colorado River, and Arizona could have
difficulty meeting its water needs. In such a situation, the Phoenix Metropolitan Area would have to
rely heavily on its groundwater resources, which could, under extreme circumstances, be depleted.
One drop of water can be used multiple times for different uses, so technically water itself
is not subtractable in the sense that the same drop of water used by a city resident to wash their
dishes can be used later to water a golf course. This is exactly the same as the situation with a
book in our library. One person can read a book now, then someone else can read it later. The key
limit is that the resource cannot be used for two different purposes at the same time. So, if clean
fresh water is being used for washing, flushing, drinking, etc., that same quality of water is not
available for other activities at that time. Even though we can use creative coordination mechanisms
and technology to recycle shower water to flush toilets, nevertheless, useable water is frequently
discarded as waste water to the sewer system or the natural environment. At that point, such water
needs to be recollected and treated for reuse. Hence inefficient use of water by individuals leads
to a shortage of clean fresh water for the group. Thus, water use is a social dilemma, or as social
dilemmas associated with common-pool resources are sometimes called a “commons dilemma.”
The second theme is water overabundance, which leads to flooding. Extreme rain events are
5.1 Introduction 53

Figure 5.2: Global flood total economic loss risk distribution.

not the only cause of floods. Another factor is land use change, deforestation and canalization of
rivers. These processes reduce the ability of ecosystems to absorb surplus water, causing flood
events. Figure 5.2 shows the distribution of damage due to floods over the past 20 years. Note that
most damage occurs in the large river basins, such as the Mississippi, the Rhine, the Ganges and the
Mekong.
A third theme is water distribution. An increasing share of food production is dependent on
irrigation where water is distributed from sources (rivers, ground water, lakes, etc), to individual
plots of land. About 70% of global freshwater supply is used to irrigate small plots (over 90% of
plots worldwide are less than 2 hectares or about 4 football fields). In order to get available water to
the right location at the right time, considerable infrastructure is required in the form of irrigation
systems and the institutions required to run them.
Figure 5.3 shows the distribution of irrigation areas around the world. Note that the proportion
of irrigated land is highest in south Asia. A large proportion of this land is used to grow rice.
Finally, there is a problem with water pollution. Waterways can become polluted since rivers
are used as sewage systems and to dispose of industrial waste in many countries around the world.
Even countries that have developed extensive public infrastructure for waste management still face
problems associated with the pollution of waterways. Some of the most common pollutants in
rivers in the U.S., for example, are nitrates and phosphates picked up when rain water flows off of
agricultural fields into rivers. Such pollution is a classic case of a negative externality: those who
benefit from activities that cause the pollution are not the same people who experience the negative
effects. Someone pollutes the water upstream causing damage downstream. We will now present a
number of cases in more detail to analyze the social dilemmas involved and the various ways people
have dealt with these dilemmas.
54 Water Governance

Figure 5.3: Global map of irrigation areas: Percentage of 5-minute grid cell area that was equipped for
irrigation around the year 2000.

5.2 Water scarcity
The city of Phoenix was founded in 1861. Since then, the population
in the Phoenix metropolitan area has grown to about 4.2 million
people (2010). The city has a subtropical desert climate with tem-
peratures up to 118◦ F (48◦ C) in the summer, and average annual
precipitation of 8.3 inches (210 mm). As places to live go, Phoenix
is a harsh climate to be sure (Figure 5.4). This raises the question of
why do so many people live in the desert? Where do they get their
water from?
When the Mexican-American war ended in 1848, Americans
started to explore the west in search of riches. One such explorer,
Jack Swilling, while on an outing near the White Tank Mountains in Figure 5.4: Phoenix valley.
1857, noted an abandoned river valley. This valley, where modern-
day Phoenix is located, in fact, has excellent terrain, fertile soils,
and an excellent climate for farming. All that was required was water.
Swilling was not the first to recognize the farming potential of the valley. The Hohokam people
lived in the valley for more than 1000 years and created 135 miles (217 kilometers) of irrigation
canals. The Hohokam were very successful farmers and engaged in an extensive trade network that
covered a significant portion of what is now the state of Arizona. There are many archaeological
features that attest to the scale of Hohokam irrigation society. After about 1070 C.E. Hohokam
society began to change and by 1450, the Hohokam abandoned the valley. The reasons for the
abandonment are not known, but might relate to a period of major droughts and severe floods,
destroying important infrastructure. It is the remains of Hohokam irrigation systems that sparked
5.2 Water scarcity 55

Swilling’s imagination for what the potential of the valley might be. Swilling built a series of canals
following the Hohokam system, thus founding Phoenix. The name Phoenix was chosen to reflect
the fact that it is a city born from the ruins of a former civilization. Even today, canals providing
Phoenix with water follow the ancient canal systems of the Hohokam.
Initially, Phoenix was a productive agricultural area for cotton and citrus with year-round sun
and plentiful irrigation water from the Salt and Gila Rivers. In 1911, the Roosevelt dam was created
east of the valley, which provided a more predictable source of water and with it, the opportunity to
grow to a population of 150,000 people. By damming the Salt River, water could be accumulated in
the mountains near Phoenix and distributed via irrigation canals. As a result, the Salt River bed that
runs through Phoenix is dry most of the time.
In 1922, the seven U.S. states that are part of the basin of the Colorado River created an agreement,
called the Colorado River Compact. This agreement defines the allocation of water rights among
the states of Colorado, New Mexico, Utah, Wyoming, Nevada, Arizona and California. Based on
historical rainfall patterns, the flow of the Colorado was equally divided between the upper division
states (Colorado, New Mexico, Utah and Wyoming), and the lower division states (Nevada, Arizona
and California).
Defining the rights to Colorado River water created opportunities for irrigation, and led to projects
like the creation of the Hoover dam to harvest Colorado River water. Arizona was dissatisfied with
the agreement and did not ratify it until 1944. Negotiations dragged on for almost 20 more years
until specific disagreements with California were settled in the Supreme Court in 1963. The verdict
specifies that California can use 50% of the river flow up to a maximum of 5.4 cubic kilometres
annually, Nevada 0.4 cubic kilometres and Arizona the remainder of the lower Colorado river flow.
After solving the allocation problem, the way was cleared for
the construction of the largest aqueduct system within the U.S.,
the Central Arizona Project. Three hundred and thirty six miles of
canals bring the water from the Colorado River to the urban areas
in the central and southern regions of Arizona (Figure 5.5). The
project started in 1973 and took 20 years to complete. Since the
1950s the population in the Phoenix metropolitan area has rapidly
increased. This was made possible by the availability of affordable
air conditioning. With the rapid growth of the city, agricultural
land was transferred into urban use. Although urban use includes
golf courses, swimming pools, and the domestic use of water, water
use per ha in an urban setting is still considerably less than for
agricultural land, which uses a lot of water for irrigation.
Nevertheless, since the 1990s there has been a drought that has
led to lower water flows in the Colorado River. As water demand Figure 5.5: Central Arizona
Project canal.
continues to grow, the city will either be forced to increase the use
of groundwater or to direct ever more water away from agriculture
to meet the demand. In order to evaluate the water use challenges for the future of Arizona, Arizona
State University has developed a simulation model that enables observers to explore the consequences
of droughts, population increase, and water policies. You can explore different scenarios of water
use projections at http://watersim.asu.edu/.
The story of the development of Phoenix is the story of how water scarcity was overcome with
public goods (infrastructure). There is no way that a small group of individuals could build the
Roosevelt Dam. Previous efforts by smaller groups to build canal systems often failed. Getting water
56 Water Governance

to the arid west required a monumental effort. In this case, a central government was essential to
provide public infrastructure due to the scale of the problem. This is not always the case. Sometimes
smaller groups can solve such large-scale problems, as we shall soon see.

5.3 Water abundance
The Netherlands is a small country in Western
Europe. The average income per person is one
of the highest in the world. However, 25% of
the Netherlands is actually below sea level (Fig-
ure 5.6). Furthermore, some of the biggest rivers
in Europe cross the Netherlands on the way to
their final destination in the North Sea.
In contrast to Phoenix, the Netherlands is a
place with an abundance of water. Over time the
Dutch have had to solve various social dilemmas
to keep their feet dry.
Before 800 C.E. the inhabitants of the pre-
cursor of the Netherlands used non-structural
measures to keep their feet dry. Such mea-
sures like man-made hills or abandoning ar-
eas in times of danger were the result of de-
cisions made by individual households. Due
to increased population pressure, technological
know-how, and finance, there was a rapid in-
crease in the development of structural water
control measures after 800 C.E. Such measures
included dikes and sluices. Construction and
maintenance of these structures required coop-
eration within communities.
Farmers whose lands directly bordered the
Figure 5.6: Map of the Netherlands with portion below
dikes agreed to commit themselves to the neces-
sea level.
sary construction work and maintenance activ-
ities. Coincident with the construction of dikes,
drainage activities began to be developed as well. To make the lowland area inhabitable, it was
necessary to get rid of the extra water. Small dams and sluices were built and maintained, based on
similar agreements as for the flood protection systems between direct beneficiaries. A noticeable
difference with regard to input for dike maintenance and small dams and sluices was the fact that in
the case of the latter, all beneficiaries had to pay for the benefits received. These dikes, dams, and
sluices are all quintessential examples of public infrastructure.
Originally, the local communities in the countryside were in charge of all general collective
interests and took responsibility for water management as well. Around 1100 C.E., however, a new
adaptation occurred as water management tasks gradually began to become separated from general
public tasks. The reason is likely due to the increase in the number and severity of flooding events
as well as a growing interdependence and complexity of the hydraulic works that began to stretch
beyond the local scale.
5.3 Water abundance 57

Starting at the end of the 11th century and the beginning of the 12th century, the first public
bodies charged with governing local and regional water management appeared on the scene and the
phenomenon of the water boards was born. The purpose of the water boards was to construct and
maintain the necessary hydraulic structures, providing safety through dikes and dry feet through
drainage (Figure 5.7). Their establishment was recognized by the higher, regional authorities who still
held themselves responsible for good water management but who resigned from their administrative
duties.
Each of the water boards differed in their design and
implementation of physical structures as well as rules.
They were also confronted with different problems. They
were not always successful in preventing floods or drain-
ing areas effectively. During the period known as the “Re-
public of the United Provinces of the Netherlands” from
1581 to 1795, there were severe floods and extensive peat-
digging (for fuel), which caused unintended artificial lakes
and diverse management problems. Still, the water boards
survived this period. One of the main reasons for the
long-term adaptation and survival of the water boards is
the institutional arrangements upon which they are based. Figure 5.7: Windmills at the Kinderdijk,
The design of rules was based on the shared norms and near the hometown of one of the authors.
values of the population. Although the water boards were
not always successful in maintaining safety and dry feet, they were maintained anyway because
changing them would be costly. The benefits associated with switching to new and unfamiliar
institutional arrangements in an effort to improve performance may have been outweighed by the
costs of operating these new institutions. Thus, the water boards maintained the familiar institutional
arrangements, which they knew how to operate and that they could adapt. Perhaps the roots of the
shared norms in contemporary Dutch society goes back to those people who found ways to make the
land liveable by developing institutions based on reciprocity.
Since the Napoleanic occupation of the Netherlands in the early 1800s, there has been an
increased centralization of water governance over time. Although water boards are still independent
organizations, a ministry of water management was created in order to coordinate water management
over the entire country. In 1953, a major flood in the south of the Netherlands killed 1800 people.
This event led to an increased effort to protect the increasingly urbanized Netherlands from potential
floods. As a result, there has been a huge amount of investment in infrastructure made to reduce the
risks of flooding over the past 60 years.
Interestingly, major challenges for the future of the Netherlands do not come from the sea. The
canalization of the river Rhine has made the river more suitable for the transport of goods to Germany
via ships, but it has also reduced the natural buffering capacity of the river. Removal of swamps
that naturally would be areas to buffer excess water, now leads to rapid transport of water down the
river during rain events. As a consequence, floodings now happen more frequently in the river delta.
Upstream countries need to increase their buffer areas to reduce the flooding risks in downstream
countries. This situation raises an important point: there are often inherent trade-offs when choosing
among performance, robustness (the capacity to cope with change) and robustness to different types
of shocks. So the Netherlands has become fairly robust to weather shocks from the sea but, in
so doing (occupying more and more low-lying land) it has become more vulnerable to weather
shocks from continental Europe (flooding of the Rhine due to rain). Can you run the same mental
58 Water Governance

experiment with the situation in Arizona?
The history of the water boards shows a continuous tinkering with rules at different levels
of organization and spatial scales. Disturbances like floods and the unintended consequences of
peat digging have triggered the development of new rules and structures. The Dutch water boards
illustrate how local-level governance structures may evolve into a resilient collaboration of multi-level
governance structures when national institutions recognize the importance of smaller governance
units and work with them rather than destroying them.

5.4 Distribution of water: irrigation
Irrigation requires coordination and cooperation. One has to build infrastructure in order to move
the water around. This infrastructure needs to be maintained in order to function properly. With
maintenance we mean cleaning of the canals and repairing damage to levies and diversion structures.
Once the infrastructure is in place, water needs to be shared. In most cases, farmers who have
their plots of land near the source of the water have preferential access to the water. In order for
downstream farmers to get the water they need, the upstream farmers need to restrict their use of
water. But given our earlier discussion of social dilemmas, why would they do this?
The earliest records of irrigation that have been found date to about 8000 years ago in the Middle
East (contemporary Iraq and Iran) and independent development of irrigation systems are found all
over the rest of the world. There is a wide variety of ways in which infrastructure is built ranging
from mud-based canals and simple wooden diversion structures, to concrete canals and computer
operated diversion gates.
In 1957 Karl Wittfogel published the book Oriental Despotism: A Comparative Study of Total
Power in which he argued that the problems associated with complex irrigation systems are the
source of complex societies. In his view, large bureaucratic systems are needed to coordinate labor
and supplies required to operate large irrigation systems. These bureaucratic systems, he argues,
were then later extended to the rest of society. Although this is an interesting and plausible argument,
it rests on a basic assumption: top-down intervention is needed for the operation of large, complex
irrigation systems. From many studies of irrigation systems in the last few decades, we see a different
story emerging that suggests that this is not necessarily the case. Let’s discuss a concrete example.
Bali is one of the islands of Indonesia that has had a
complex and very productive irrigation society for about
1000 years. Hundreds of irrigation communities called
subaks are connected via waterways that begin at a lake
near the top of on an old volcano (Figure 5.8). Canals
connect this water to subaks downstream. This complex
irrigation system has been studied in detail by anthropolo-
gist Stephen Lansing and has been made famous in a book
titled Priests and Programmers.
The irrigators have to solve a complex coordination
problem involving water distribution and pest control (they Figure 5.8: Subak irrigation.
didn’t have pesticides 1000 years ago). On the one hand,
control of pests is most effective when all rice fields in a
particular subregion are on the same schedule for planting rice. This is due to the fact that the pests
(insects called plant hoppers) are limited in their ability to move (or disperse) on the landscape. If
large enough areas are kept fallow (areas without plants) between planted areas, the plant hoppers
5.4 Distribution of water: irrigation 59

can’t cross them because there isn’t any food to keep them alive while they cross (i.e., the fallow
areas are “food deserts”). This keeps pest outbreaks localized if they occur at all. On the other hand,
the terraces (see Figure 5.8) are hydrologically interdependent, with long and fragile systems of
weirs, tunnels, canals, and aqueducts used to control where water goes, making it challenging to get
the water to all the fields while maintaining large enough fallow areas between planted areas, all in a
limited growing season.
To balance the need for coordinated fallow periods and the use of water, a complex calendar
system has been developed that determines what actions should be carried out on each specific
date. These actions are related to the spiritual practice of making offerings to “water temples” at
several levels: little temples at the rice terrace level, the temple at the village level, and the temple
at the regional level, associated with the Pura Ulun Swi, “the Head of the Rice Terraces” (this is
the temple of the high priest Jero Gde, the human representative of the Goddess of the Temple of
the Crater Lake, the main source of water for irrigation). These offerings of water and other items
were collected as a counter performance for the use of water that belonged to the gods. These ritual
practices trigger the calendar actions (i.e., people make offerings at particular times after which they
can plant, etc.).
Balinese society consisted of many kingdoms before the conquest of territory of Bali around
1900 by the Dutch. The Dutch saw these offerings made to the various temples in a different light,
namely as a royal irrigation tax. The fact that during the nineteenth century there were quite a
number of kingdoms in Bali was a sign that the institution of kingship had weakened over time from
one powerful kingdom to a number of smaller kingdoms. Therefore, the Dutch wanted to restore
centralized government; in particular they wanted to use a revived royal irrigation tax to improve the
irrigation system. The Dutch administrative reorganization failed, partly due to lack of funding, but
also because historical analysis conducted during the 1930s demonstrated that there was no evidence
that Bali had ever had a centralized government. Although Indonesia became independent from the
Netherlands after World War II, many aspects of the colonial bureaucratic system were adopted by
the new independent government.
During the late 1960s the Indonesian government made self-sufficiency in rice production a
major goal for national development. In the same period the Green Revolution began in Asia. The
Green Revolution involved the spread of new rice-growing technologies that promised a dramatic
increase in rice production. Bali was one of the first targets of the Green Revolution. In contrast to
the earlier Dutch attempts to modernize rice production in Bali, this time the engineers were well
funded.
The function and power of the water temples were invisible to the planners involved in promoting
the Green Revolution. They regarded agriculture as a purely technical process. Farmers were forced
to switch to the miracle rice varieties which would lead to three harvests a year instead of the two that
could be achieved with traditional varieties. Farmers were stimulated by governmental programs that
subsidized the use of fertilizers and pesticides. The farmers continued performing their rituals, but
now they no longer coincided with the timing of rice farming activities. Soon after the introduction
of the miracle rice, a plague of plant hoppers caused a huge amount of damage to the rice crops. A
new rice variety was introduced, but it was followed by another pest plague. Furthermore, water
shortages began to occur because there was nothing to replace the rituals (which were now out of
step with plantings) which had been the basis for the efficient allocation of water.
During the 1980s, an increasing number of farmers wanted to switch back to their old ritual-based
system, but the engineers interpreted this as religious conservatism and resistance to change. Steve
Lansing quotes a frustrated American irrigation engineer “These people don’t need a high priest,
60 Water Governance

they need a hydrologist!” (Lansing, 1991, p. 115). It was Lansing who unravelled the function of the
water temples, and was able to convince the financers of the Green Revolution project on Bali that
irrigation and rice cultivation was best coordinated at the level of the water temples. Lansing built a
computer model of the artificial ecosystem, and showed that for different levels of coordination, from
farmer level up to central control, the temple level was where decisions could be made to maximize
the production of rice.
As this story suggests, the complex irrigation system on Bali and the role of the temples in
operating it has evolved over a long history of local adaptations at different levels of organization
and different spatial scales. The water temples played a significant role in the coordination of the
use of water, but also in providing technical advice and mediating water use conflicts between
different subaks. By making offerings to different temples, the farmers were made aware of the
interconnections between the water flows at different scales. Due to Lansing’s insight and analysis,
some of these systems have evolved still further and avoided the fate of many self-organized systems
of this kind when experts declared them defunct and constructed new infrastructure without paying
much attention to local property rights, ecology, culture, and traditions.

5.5 Water pollution
One of the challenges of water pollution is that the polluter often doesn’t experience the consequences
of the pollution because the water carries it away (usually toward someone else). This is a classic
form of a “negative externality.” The polluter gets the benefits of the activities leading to pollution,
while the population living where the pollution ends up experiences the costs of the pollution. To
illustrate this point, we examine the widespread problem of eutrophication.
Farmers use artificial fertilizers on their land to increase production. Not all the nutrients that
farmers apply are taken up by the plants they grow. The surplus nutrients are picked up by surface
water runoff and end up in waterways. Unfortunately, often farmers develop a perception that with
fertilizers, more is always better. Even if farmers become aware that some fertilizer is good but after
a point it may become detrimental, it is often difficult to determine where this point is in practice. In
either case, the basic pattern is that farmers increasingly use more fertilizers.
Eventually, the surplus nutrients (phospho-
rous, nitrogen, etc.) end up in lakes, seas, and
oceans. As a result, many lakes face the chal-
lenge of eutrophication where clean, clear water
flips into a green soup of algae. Once this hap-
pens, the lake cannot be used for recreation and,
over time, the algae may kill the fish. Obviously,
eutrophication increases the treatment costs if
the body of water in question is used as a source
of drinking water. Note that the benefits gener-
ated through the use of fertilizers are not shared
proportionally by the people who experience the
costs of eutrophication. Yes, they experience the
same reduced cost of food that everyone else en-
joys, but bear much more of the cost of that Figure 5.9: Dead Zone in the Gulf of Mexico.
“cheap” food.
5.6 Critical reflections 61

One of the most dramatic consequences of eutroph-
ication is the dead zone in the Gulf Mexico. A dead
zone is a low-oxygen area in large water bodies such as
oceans. Most marine life cannot flourish in such areas.
The dead zone in the Gulf of Mexico consists of more
than 22,000 square kilometers of water surface near the
coast of Texas and Louisiana (Figure 5.9 and 5.10). The
Mississippi River carries nutrients that run off with surface
water from the agricultural areas in the midwestern U.S.
These nutrients accumulate at the coast near the mouth of
the Mississippi. The coastal area is a productive fishing Figure 5.10: Mississippi River sediment
ground that is heavily impacted by the dead zone. Hence, plume entering the Gulf of Mexico near
Southwest Pass.
the livelihoods of fishers in the Gulf of Mexico are affected
by the actions of farmers in the Midwest.
This example illustrates a fundamental challenge with different levels of social organization.
Specifically, this problem cannot be addressed by individual farmers talking with individual fishers.
Higher level organizations must be part of the solution. How can farmers get incentives to reduce the
use of fertilizers? Who will compensate fishers for the loss of productivity they suffer?
Dead zones can be reversed. The North Sea had a dead zone in the 1980s and 1990s. A special
organization was created that included all countries in the river Rhine basin. Due to this higher level
of coordination, policies were implemented and enforced to reduce sewage and industrial emissions.
The nutrient inputs were reduced and the dead zone disappeared. Unfortunately, this is not always
the case. If the dead zone is an entire lake, for example, it may be the case that it cannot be restored
to a clear, blue state.

5.6 Critical reflections
Water governance has many dimensions. In this chapter we have provided a simple categorization of
some of the most important problems we face: water scarcity, water abundance, water distribution,
and water pollution. A common challenge in solving social dilemmas related to water is power
inequality. Generally, upstream participants have easier access to water and pollution is felt by
people downstream. In order to solve these problems, higher level authorities are often needed to
coordinate and mediate.
Water is involved with many social dilemmas. People share common water resources, build
public infrastructure to protect them from flooding, coordinate when and where to use common-pool
water resources and avoid pollution that affects participants in downstream positions.

5.7 Make yourself think
1. Where does the water you use in your household come from?
2. How much water does your household use, and what do you pay for it?
3. What are the main sources of water use in your household, and what can you do to reduce
water use?
4. Can you think of how to apply the action arena concept and the Institutional Analysis and
Development framework to the water you are using?
62 Water Governance

5.8 References
Hirt, P., Gustafson, A., & Larson, K. (2008). The mirage in the valley of the sun. Environmental
History, 13(3), 482–514.
Kaijser, A. (2002). System building from below: institutional change in dutch water control systems.
Technology and Culture, 43(3), 521–48.
Lansing, J. (1991). Priests and programmers: technologies of power in the engineered landscape of
Bali. Princeton, NJ: Princeton University Press.
Wittfogel, K. A. (1957). Oriental despotism: a comparative study of total power. New Haven: Yale
University Press.
Key Concepts
In this chapter we will:
– Be introduced to successful examples of governing the commons
– Become aware of the importance of monitoring and enforcement
– Learn about paper parks

6 — Harvesting the Commons

6.1 Introduction
In his 1968 essay “Tragedy of the Commons,”
Garrett Hardin argued that people are trapped in Rotating fishing spots
an unsolvable problem because they are unable to
self-organize equitable and sustainable institutional Alanya, an inshore fishery in Turkey, experienced
arrangements. As a result, he argued, outside inter- overfishing in the early 1970s and violence fre-
vention is required to overcome the overharvesting quently erupted over competition for the good fish-
of what Hardin referred to as “the commons” (what ing spots. In response, the fishers experimented
should actually be called open access resources). Al- with new institutional arrangements designed to
though there are many examples of overharvesting of enable them to recover the fishing stock and solve
problems over excessive competition. The insti-
the “commons,” this chapter will discuss success sto-
tutions that the fishers developed are as follows:
ries of self-governance of common-pool resources. Each year in September, a list of eligible fishers
These success stories provide insights into which is prepared, consisting of all licensed fishers in
factors may increase the success of self-governance Alanya. All usable fishing locations are named
efforts. and listed. The stock is such that fishing in one
We will see that solutions to overcome the site will not directly reduce the available fish in
tragedy of the commons make use of the physical adjacent sites.
The eligible fishers draw lots and are assigned to
characteristics of the action situation, and so we dis-
one of the fishing locations. Each day each fisher
tinguish three types of situations in this chapter. In moves to the next location. Under this rotational
the first type there are domesticated animals. Re- scheme, all fishing boats have equal chances to
source users in this situation must move these ani- fish at the best spots. Fishers are now themselves
mals around the landscape (consistent with the orig- the monitors. Each fisher knows which spot is
inal example used by Hardin concerning sheep shar- available for their turn for each day of the fishing
ing a meadow in his now famous paper). The second season. If a fisher wants to fish on a productive
situation involves the harvesting of wild animals such spot while it is not his turn, the fisher whose turn it
is will defend their right to the good spot. Hence,
as fish, lobster or deer (that move around on their
due to the self-monitoring of fishers there is no
own). Finally, the third type of situation involves incentive to cheat.
resources that do not move, such as forests.
64 Harvesting the Commons

6.2 Domesticated animals
Törbel, Switzerland, is a village of about 600 people located in the Vispertal trench of the upper
Valais canton. For centuries, Törbel peasants have planted their privately-owned plots with bread
grains, garden vegetables, fruit trees, and hay for winter fodder. Cheese produced by a small group
of herdsmen, who tend village cattle pastured on the communally owned alpine meadows during the
summer months, has been an important part of the local economy.
The earliest known written legal documents are from 1224, and provide information regarding
the types of land tenure and transfers that have occurred in the village and the rules used by the
villagers to regulate the five types of communally owned properties. On February 1, 1483, Törbel
residents signed articles formally establishing an association to improve the regulation of the use of
the alp, the forests, and the wastelands.
The law specifically forbade a foreigner (Fremde) who bought or otherwise occupied
land in Törbel from acquiring any right in the communal alp, common lands, or grazing
places, or permission to fell timber. Ownership of a piece of land did not automatically
confer any communal right (genossenschaftliches Recht). The inhabitants currently
possessing land and water rights reserved the power to decide whether an outsider
should be admitted to community membership (Netting, 1976, p. 139).
The boundaries of the communally-owned lands were firmly established long ago, as indicated in a
1507 inventory document.
Access to this well-defined common property was
limited to citizens, to whom communal rights were specif-
ically extended. Here it is important to underscore why
Hardin’s use of the term “commons” is incorrect. The
alpine meadows of Törbel are “commons” (Figure 6.1)
in the sense that they consist of a common-pool resource
over which there are no private property rights. It is prop-
erty held in common with communal rights. Thus, as we
discussed previously, “commons” is not equal to “open
access,” which refers to property with no rights attached.
These pastures in Törbel are examples that not all “com-
mons” end in tragedy as Hardin suggested. As far as the Figure 6.1: Cow in Törbel.
summer grazing pastures were concerned (the common-
pool resource), regulations written in 1517 stated “no
citizen could send more cows to the alp than he could feed during the winter” (Netting, 1976, p. 139).
This regulation is still enforced today and provides for the imposition of substantial fines for any
attempt by villagers to appropriate a larger share of grazing rights. Adherence to this “wintering” rule
was administered by a local official who was authorized to levy fines on those who exceeded their
quotas and to keep one-half of the fines for himself. Many other Swiss villages use the wintering
rule as a means for allocating appropriation rights (frequently referred to as “cow rights”) to the
commons. This and other forms of cow rights are relatively easy to monitor and enforce. The cows
are all sent to the mountain to be cared for by the herdsmen. They must be counted immediately, as
the number of cows each family sends is the basis for determining the amount of cheese the family
will receive at the annual distribution.
The village statutes are voted on by all citizens and provide the general legal authority for an
alp association to manage the alp. This association, which includes all local citizens owning cattle,
6.2 Domesticated animals 65

holds annual meetings to discuss general rules and policies and elect officials. The officials hire the
alp staff, impose fines for misuse of the common property, arrange for distribution of manure on
the summer pastures, and organize the annual maintenance work, such as building and maintaining
roads and paths to and on the alp and rebuilding avalanche-damaged huts. Labor contributions or
fees related to the use of the meadows are usually set in proportion to the number of cattle sent by
each owner. Trees that will provide timber for construction and wood for heating are marked by
village officials and assigned by lot to groups of households, whose members are then authorized to
enter the forests and harvest the marked trees.
Private rights to land are well developed in Törbel and other Swiss villages. Most of the
meadows, gardens, grain fields, and vineyards are owned by various individuals, and complex
condominium-type agreements are devised for the fractional ownership among siblings and other
relatives of barns, granaries, and multistory housing units. The inheritance system in Törbel ensures
that all legitimate offspring share equally in the division of the private holdings of their parents and
consequently in access to the commons, but family property is not divided until surviving siblings
are relatively mature. Prior to a period of population growth in the nineteenth century, and hence
severe population pressure on the limited land, the level of resource use was held in check by various
population-control measures such as late marriages, high rates of celibacy, long birth spacing, and
considerable emigration.
The Swiss villagers have experienced the advantages and disadvantages of both private and
communal tenure systems for at least five centuries, and they continue to use the communal tenure
system. Although the yields are low, the land in Törbel has maintained its productivity for many
centuries. Netting (1976) associates five attributes to land-use patterns with the differences between
communal and individual land tenure. He argues that communal forms of land tenure are better
suited to the problems that appropriators face when (1) the value of production per unit of land is
low, (2) the frequency of dependability of use or yield is low, (3) the possibility of improvement or
intensification is low, (4) a large territory is needed for effective use, and (5) relatively large groups
are required for capital-intensive activities.
Not all owners of livestock own land, private or com-
munal. Nomadic herders, or pastoralists, lead their live-
stock to graze around a large spatial landscape in order to
be at the right place at the right time (Figure 6.2). In such
cases, institutions have been developed to gain access to
the land of various landowners. Our next example, the
complex dynamics of Gaddi shepherds and their landscape
in the Himachal Pradesh in India, demonstrates just such
a situation. No particular place in this landscape is ideal
for the maintenance of goats and sheep throughout the
entire year. The only way that these animals can be cared
for is to move them across a very large area with highly
Figure 6.2: Gaddi shepherds with flock.
variable terrain. These pastoralists originally adapted their
institutions to the harsh ecological conditions they faced
in order to survive. They move their animals, goats, and sheep across a vast mountainous landscape
within Himachal Pradesh. During the winter, they descend from the mountains and graze in the
valleys and the lower elevation forests. The shepherds have made arrangements with agriculturalists
(who own private plots of land) to graze on the stubble left after a harvest from private fields in
return for the highly valued manure of the goats and sheep. In the summertime, it is too hot at lower
66 Harvesting the Commons

elevations, so the pastoralists move into the mountains around the tree line. Lyall writes:
Snow and frost, in the high ranges, and heavy rain and heat in the low, make it impossible to
carry sheep farming on a tolerably large scale with success in any part of the country. The only way
is to change ground with the seasons, spending the winter in the forests in the low hills, retreating in
the spring before the heat, up the sides of the snowy range, and crossing and going behind it to avoid
the heavy rains in the summer (Lyall, 1872, p. 46; cited in Chakravarty-Kaul, 1998).
These seasonal movements are based on reciprocal relationships. The Gaddis shepherds invest a
lot of time in social networking among themselves and with outsiders to provide access to grazing
areas in return for manure and other goods and services.
The informally-evolved rights of the Gaddis shepherds have never been formally recognized by
the national government. In 1947, the Indian government adopted policies that reduced the shepherds’
access to the usual grazing grounds by building dams to generate hydropower and by providing
strictly private property rights to farming communities. This has resulted in more concentrated areas
where livestock can graze, and may have contributed to erosion in the forested hilly regions. The
government has accused the Gaddis of free-riding within this commons dilemma. However, the
government had not recognized the efficient system that the participants in this action situation had
already worked out; in fact, the shepherds and agriculturalists had developed an effective bargaining
solution by trading manure for grazing rights.
The shepherds adapted to temporal and spatial variability in their system by moving around the
landscape in a particular, well-ordered pattern. Activities that hinder this movement pattern on the
landscape hit the vulnerable point of this transhumance system (the seasonal movement of people
with their livestock between fixed summer and winter pastures). When these movement patterns are
affected, the shepherds are forced to use a smaller area which may, in turn, lead to overgrazing. Thus,
the transhumance system is highly tolerant to seasonal variation through very specific institutional
arrangements, but is extremely vulnerable to changes in access by social or physical barriers.

6.3 Wild animals
The Maine lobster fishery is a remarkable story of self-governance
within the contemporary United States. The lobster fishery of Maine
is organized into territories along most of the coast. Day-to-day
fishing regulations are organized by harbor gangs. These harbor
gangs are informal groups that enforce local customs. In order to
fish for lobster you need to become a member of a harbor gang,
which is the group of fishers who go lobstering from a single harbor.
Members of this group can only set traps in the traditional territory
of the harbor group. There are various fishing practices that each
member is expected to obey (Figure 6.3). These practices vary from
harbor to harbor .
To become a member of a harbor community (a formal position)
requires participation by family members in that community for
several generations. People who are not born and raised in these
harbor communities are considered outsiders and will have difficulty
Figure 6.3: Lobster fisher.
gaining the level of acceptance by the rest of the community that is
required before the right to fish for lobster would be granted.
In all harbor communities a person who gains a reputation for damaging others’ gear or for
6.3 Wild animals 67

violating conservation laws will be severely sanctioned. For example, if a fisher goes out to collect
his traps, and discovers that somebody else has put traps in the same location, he may signal that
this norm violation has been noticed by taking a lobster from the trap and leaving the trap open.
If violations of accepted locations of traps continue, more severe measures can be taken, such as
damaging gear and cutting the traps loose.
If a fisher put traps in another gang’s territory, similar types of enforcement can be expected.
The damages eventually make it unprofitable to continue breaking the informal rules, but are not so
severe as to initiate a legal dispute.
There are also formal laws in the state of Maine to protect the breeding stock and increase the
likelihood that the regeneration rate remains high. The most important conservation laws are the
minimum and maximum size measures, a prohibition against catching lobsters with eggs, and a
law to prohibit the taking of lobsters which once had eggs and were marked (i.e., the V-notch law)
(Figure 6.4).
When a fisher collects the catch from his trap, he measures each
lobster. If it is too small or too big, it will be thrown back. Such
a rule avoids catching young lobsters, allowing them to mature to
an age at which they will start generating offspring. If a lobster is
caught that is carrying eggs, it is a productive female. The lobster
will get a V-notch in the tail, and will be thrown back. This is also
the case if a lobster is caught with a visible V-notch. This policy
ensures that fertile female lobsters are kept alive.
Why would fishers not cheat? How could other fishers find out
if a V-notched female were caught? How would they find out if a
lobster that is too small was caught? It turns out that the coopera-
tion of the middlemen (the lobster buyers) in the lobster industry is
required to prevent cheating. These middlemen do not accept inel- Figure 6.4: V-notch in tail of lob-
igible lobsters because their long-term financial viability depends ster. Notice the eggs.
on a productive lobster population. This gives them an incentive to
help enforce the rules (by simply not buying such lobsters).
Interestingly, neither the state nor any of the lobster gangs has tried to limit the quantity of lobster
captured. Further, the state does not try to limit the number of lobster fishers, since this is already
done at a local level. However, the state has been willing to intercede when issues exceed the scope
of control of local groups.
In the late 1920s, when lobster stocks were at very low levels and many local areas appear to
have had compliance problems, the state took a number of steps (including threatening to close the
fishery) that supported informal local enforcement efforts. By the late 1930s, compliance problems
were largely resolved and stocks rebounded. Note however, that there are still too many unknowns
about lobster biology to make the claim that change in management was the cause of the recovery.
Figure 6.5 shows the decline of lobster catches from the 1880s until the 1930s. The absolute
numbers dropped, as well as the catch per trap and per fisher. Since the 1940s the catch numbers
have increased, with a rapid increase after 1990. Since the catch numbers per trap and per fisher are
also increasing, this suggests that the lobster population is in good condition. If the rapid increase in
total catches were caused by overharvesting we would expect a reduction in the catch per trap over
time, since a decrease in the total number of available lobsters would cause an increase in the time
needed to catch each lobster. Thus, each fisher would be catching fewer lobsters in a given season.
But if more fishers are fishing, the total catch could remain the same, or even increase, while the
68 Harvesting the Commons

Figure 6.5: From top to bottom: (a) Lobster catch (in lbs) since 1880. (b) Lobster catch per trap and license
based on data from the State of Maine historic data.
6.4 Forests 69

stock is being depleted. These are the classic symptoms of overharvesting. Figure 6.5 shows, in fact,
that the catch per trap and catch per license are actually increasing. This suggests that the lobster
fishery is not being overharvested, at least at the present moment.
Recently, in response to changes that were breaking down the informal harbor gang system, the
state has formalized the system by dividing the state into zones with democratically elected councils.
Each council has been given authority over rules that have principally local impacts—trap limits,
days and times fished, and so forth. Interestingly, the formalization of local zones was followed
almost immediately by the creation of an informal council of councils to address problems at a
greater than local scale.

6.4 Forests
The final examples we will discuss are forests. More specifically we look at the most effective ways
to protect forests. Due to the importance of forests for biodiversity, there has been an increasing focus
on creating parks and protected areas. One key concern is whether these areas are best protected by
putting a fence around them or allowing human populations to continue to occupy them and help
with conservation efforts. A related concern is whether designated areas become “paper parks,” i.e.,
areas set aside for protection on paper, but in practice, the lack of enforcement allows for a lot of
poaching and illegal logging.
Ostrom and Nagendra (2006) discuss long-term studies of land use change to test which type
of management is most effective. They compared governmental, community, and private forests
and found that the particular form of ownership is not important for the condition of the forests as
measured by the quality and size of the trees. More important is whether boundaries have been
well established in the field and are considered legitimate and whether regular monitoring and
enforcement of rules related to entry and use exist (Figure 6.6).
Whether the boundaries are considered legit-
imate depends on whether people have lived in
the protected area before or if surrounding popu-
lations have used the resources over an extended
period of time. If the boundaries are considered
legitimate, how can they be monitored? For ex-
ample, if indigenous populations are taken out
of the protected areas (as has been the approach
in several cases), one also loses potential moni-
toring capacity and it is not uncommon to see an
increase in poaching in protected areas. Rather
than relying in indigenous populations, perhaps Figure 6.6: Bicycles and trucks confiscated from people
caught illegally removing large logs from the forests.
paid guards are the answer? What are the incen- Note the circular modification in the cycle frame (Inset)
tives facing paid guards to monitor and enforce made to hold large logs of teak wood.
the rules? If guards are not paid well, they might
be willing to accept bribes and not bother those
who harvest illegally. This has also been observed frequently.
Community forests (as opposed to those run by the state with paid monitoring and sanctioning)
can be effective since the population who benefits most from protecting the forests also monitors
their use. In this case, because the cost of monitoring is aligned with its benefits, there is an incentive
for high level monitoring effort, and those who are caught breaking the rules will experience social
70 Harvesting the Commons

sanctions. On the other hand, community forests might be less effective in dealing with intrusion
from outsiders due to lack of capacity and legal abilities to limit access (i.e., social sanctions won’t
be effective on outsiders). In this case, community forest managers need the assistance of the state.
This is a clear example illustrating the importance of multilevel governance.
Governmental and privately owned forests can be effective if sufficient effort is made in enforce-
ment. But they might be more prone to corrupt guards. Further, when local people do not feel a sense
of ownership and participation in the process, they are less willing to assist with monitoring (hiring
enough guards is prohibitively expensive). This will lead to a lack of sufficient eyes to monitor the
use of the forests.

6.5 Critical reflections
There are many examples of self-organized governance of common-pool resources that have been
successful over a long time. In the successful examples we see that local communities play an
important role, often crafting the rules and monitoring and enforcing those rules.

6.6 Make yourself think
1. Do you make use of public parks? What is the state of the park? What are the rules, and how
are they enforced?

6.7 References
Acheson, J. (2003). Capturing the commons: devising institutions to manage the maine lobster
industry. University Press of New England.
Chakravarty-Kaul, M. (1998). Transhumance and customary pastoral right in Himachal Pradesh:
claiming the high pastures for Gaddis. Mountain Research and Development, 18, 5–17.
Lyall, F. (1872). Final report of the regular settlement of Kangra, 1868–72. Lahore, India. Govern-
ment of the Punjab.
Netting, R. (1976). What alpine peasants have in common: observations on communal tenure in a
swiss village. Human Ecology, 4, 135–146.
Ostrom, E. (1990). Governing the commons: the evolution of institutions for collective action.
Cambridge University Press.
Ostrom, E., & Nagendra, H. (2006). Insights on linking forests, trees, and people from the air, on
the ground, and in the laboratory. Proceedings of the National Academy of Sciences, 103,
19224–19231.
Key concepts
In this chapter we will:
– Be introduced to design principles for effective institutions
– Learn about boundary rules, monitoring, graduated sanctions
and input of local participants to collective-choice arrangements
– See how violations of design principles may lead to corruption and rent seeking

7 — Design Principles to Sustain the Commons

7.1 Introduction
Paper parks
If so many hectares of land and ocean are designated
as protected areas, why is there still so much overfishing,
deforestation, species extinction and plundering of many
natural resources? One of the problems is that areas might be
protected on paper, in official legislation signed by officials
in suits, but there are no effective institutional arrangements
on the ground to support this legislation.
It is not uncommon that areas are officially protected
but some high ranking officials have received access to har-
vest timber (such as is happening with the national parks in
Sumatra). During his dictatorship in Indonesia, Suharto gave
timber rights to powerful military leaders to “buy” their sup- Figure 7.1: Illegal logging in West Kali-
port for his regime. As a consequence the official national mantan’s Gunung Palung National Park.
parks are now heavily deforested. Hardwoods are being trafficked by a tim-
How can we establish parks in practice instead of simply ber baron in the town of Ketapang.
on paper?

What is the best set of rules to govern the commons? This is the ultimate question asked by
policy makers involved with managing common-pool resources and providing public infrastructure.
Unfortunately, there is no such optimal set of rules. What we have learned over the years is that there
are some “design principles” that explain why some communities are successful while others are not.
These design principles are based on a systematic study of many case studies of fisheries,
irrigation, groundwater, and forestry systems. Some of these cases were discussed in the previous
two chapters. Information was collected from each case study regarding the size and composition of
the community, the formal rules and norms in use, how the system was monitored and by whom, the
conflicts that arose, and how the resource system evolved over time.
Elinor Ostrom studied hundreds of these case studies in the 1980s and proposed the design
principles in her classic 1990 book Governing the Commons. She initially focused on determining
which rules were best but was unsuccessful in identifying a particular set of rules that were “best” in
all circumstances. Instead she turned her efforts toward identifying eight underlying design principles
that characterized case studies of long-lasting common-pool resource systems. The design principles
she identified were mostly met in these long-lasting systems, but were absent in those that collapsed.
72 Design Principles to Sustain the Commons

The concept of eight design principles was an initial proposal in 1990. Twenty years later, an
analysis of about 100 case-studies by Cox et al. (2010) provides evidence that the design principles
hold up when challenged with data.

7.2 Institutional design principles
The design principles derived from case studies of long-lasting systems of common-pool resource
governance:
1. Clearly defined boundaries. The boundaries of the resource system (e.g., irrigation system
or fishery) and the individuals or households with rights to harvest resource units are clearly
defined.
2. Proportional equivalence between benefits and costs. Rules specifying the amount of
resource products that a user is allocated are related to local conditions and to rules requiring
labor, materials, and/or money inputs.
3. Collective-choice arrangements. Many of the individuals affected by harvesting and protec-
tion rules are included in the group that can modify these rules.
4. Monitoring. Monitors, who actively audit biophysical conditions and user behavior, are at
least partially accountable to the users and/or are the users themselves.
5. Graduated sanctions. Users who violate rules-in-use are likely to receive graduated sanctions
(depending on the seriousness and context of the offense) from other users, from officials
accountable to these users, or from both.
6. Conflict-resolution mechanisms. Users and their officials have rapid access to low-cost,
local action situations to resolve conflict among users or between users and officials.
7. Minimal recognition of rights to organize. The rights of users to devise their own institutions
are not challenged by external governmental authorities, and users have long-term tenure
rights to the resource
For resources that are parts of larger systems:
8. Nested enterprises. Appropriation, provision, monitoring, enforcement, conflict resolution,
and governance activities are organized in multiple layers of nested enterprises.
We will now look at each design principle in more detail. Although the design principles are
intended for natural resources, we will illustrate their use with other kind of systems too.

7.3 Examples of design principles
7.3.1 Well-defined boundaries
The first principle relates to two types of boundaries: the boundaries of the resource system, and
the boundaries that determine which individuals or households have rights to harvest the resource.
Having clear boundaries is critical to addressing the problem of free-riding. If there are clear
boundaries indicating who can appropriate which resources, it will be clear who is following the
rules or not when harvesting activities are observed.
These two types of boundaries are demonstrated in our roadway systems. Not having a valid
drivers’ license is a clear violation of a boundary rule of driving a car on the road. Roads themselves
benefit from clear demarcations on the road to define the lanes.
7.3 Examples of design principles 73

Such clear demarcations also help in sports to define the boundaries of the common space for
competition. What would John McEnroe do if there were no clear lines on the tennis court? In
many sports, players wear a uniform and a number that indicates that they are allowed to participate.
You cannot just run in the New York marathon. You have
to sign up and wear your start number and start in your
assigned position. Also many internet services require
that you register in order to be able to use those services.
Watching a streaming Netflix movie in the U.S. is possible
if you have paid your monthly fee, but it is not possible
to do this on the same account on your laptop in Europe.
The reason for this is that the rights for consuming the
Netflix content are clearly defined for certain countries for
which Netflix owns copyrights. When your laptop uses an Figure 7.2: Nowadays technology helps to
IP address outside the U.S. it will not provide access to determine whether a tennis ball is in or out.
the content.
When resource users create boundaries, they can make use of the ecological context. For
example, defining specific areas where one can harvest from the resource makes it easier to monitor
and enforce.
Clearly defined boundaries are not enough for a successful institutional arrangement. One also
needs to be able to defend boundaries against potential intruders. But the research shows that clearly
defined boundaries are a prerequisite for success.

7.3.2 Proportional equivalence between benefits and costs
Do the rules allocate the benefits from the resource in proportion to the effort people put into
harvesting and maintaining the resource? If some users get all the benefits and pay a small proportion
of the costs, other users might not be willing to follow the rules over the long term. This design
principle relates to fairness. Yet some inequality in the benefits people derive from the resource are
acceptable provided the cost they bear in relation to the benefit is proportional. For instance, users
may accept that the individuals who put a lot of effort into building an irrigation system get a better
plot of land to grow his/her crops.
Salary levels are a typical topic of debate, especially in periods of economically challenging
times. For the sustainability of an organization, it is not wise to lay off employers while the boss
keeps her seven-figure salary. Similar levels of sacrifices are expected over the whole organization.
There are important differences between countries in how costs and benefits are defined. For
example, because the labor market in the U.S. is more flexible it is easier to lose a job and there
is only a limited safety net. Within Europe, social security provides a minimum level of income
for people who have no job. How a society allocates its costs and benefits—as exemplified in their
rules—is largely defined by prevailing social norms. Fairness within the U.S. relates to receiving
benefits based on effort, while in Europe fairness relates to equality among people independent of
effort. This is a gross simplification, but as the authors can attest—since both of us have lived on
both continents—there are important differences in social norms that affect the rules on allocation of
costs and benefits.
74 Design Principles to Sustain the Commons

7.3.3 Collective-choice arrangements
People who are affected by institutional arrangements should have a way to participate in making
and modifying the rules. Action situations where local resource users are able to devise rules are able
to better tune those rules to local circumstances. Further, when participants make the rules together,
they often receive more careful consideration by participants. When rules are viewed as imposed by
an elite, participants are less likely to comply with them.
Universities typically include students in their governance systems, and sport federations have
athletes on their boards. These are all ways to include relevant knowledge in the decision making
process—the perspective of students and athletes—and increase the likelihood that the rules are
accepted by the broader community.
In larger organizations, or even states, countries and the global community, the ability of people
to participate in rule crafting seems difficult. Sometimes we may think the only thing we can do is
to vote for who represents us in the decision-making process. One of the big challenges we face in
modern times is the large scale of the communities we participate in.

7.3.4 Monitoring
How is monitoring of the rules organized? Effective monitoring is not only a matter of counting
the number of guards, referees and policemen and making sure this number is high enough. The
subtleties of what these monitors actually do to monitor and enforce the rules and the incentives
they face are critical. If guards in a national park are not paid well, it is not surprising that they
accept bribes to look the other way when poaching is happening. It is often more effective to have
local people as monitors. This may assure conditionally-cooperative resource users that somebody is
genuinely checking on the conformance of others to local rules. In urban areas neighbors sometimes
organize themselves in a neighborhood watch instead of hiring security personel.
To understand whether monitoring will be effective, we need to understand what the incentives
are for a monitor. Will a monitor be paid independent of whether rules are broken or not? Will a
monitor be affected by rule breaking (are they a local resource user themselves). Will somebody
notice whether the monitor is doing his or her job or not? If there is a lot of variation in harvest
levels, people may not notice in the short run whether rules are broken or not.
Sometimes monitoring is so important for the functioning of the system that high investments
are made in monitoring. In some highly productive fisheries in the Artic, each boat has an official of
the federal agency NOAA (National Oceanic and Atmospheric Administration) on board. Does this
sound outrageous? This is like having a police officer riding in every car to make sure drivers don’t
break the rules. What about all the official and unofficial referees (general public) during a football
match?

7.3.5 Graduated sanctions
People make errors. When you make a mistake you get a warning. If you keep bullying a player on
the other team you may receive a yellow card. And if you continue ignoring the rules you may get
sent off the field with a red card or even be expelled from the league.
For many action situations there is a graduated sanctioning system. One reason is that rules are
not always commonly understood or known, and getting a warning when a rule is broken may remind
people of the actual rules in use. Another reason is the potentially high costs of strict enforcement.
What would happen to a sport if there was strict enforcement of the rules? There are norms of fair
play, and a tolerance of players exploring the boundaries of the formal rules.
7.3 Examples of design principles 75

7.3.6 Conflict-resolution mechanisms
The goal of conflict resolution mechanisms is to provide access to rapid, low-cost, local opportunities
to resolve conflict among users or between users and officials. Rules, unlike laws of nature, have
to be understood in order to be effective (the laws of nature function whether or not we understand
them—right?). There are always situations in which participants can interpret a rule that they
have jointly made in different ways. By devising simple, local mechanisms to get conflicts aired
immediately and resolutions that are generally known and accepted in the community, the number of
conflicts that reduce trust and cost time and money can be minimized. If individuals are going to
follow rules over a long period of time, some mechanism for discussing and resolving what is or is
not a rule infraction is quite necessary to the continuance of rule conformance itself. For example,
within Wikipedia there can be eruptions of editing wars and designated editors can mediate between
the different parties to resolve the conflict.
Not all disputes within the U.S. appear in front of the Supreme Court. Most conflicts can be
resolved informally by having a good discussion in a neutral environment. By having a drink with
your colleague or neighbor, you may discover that the conflict is mainly caused by a misunderstanding.
When conflicts are not resolved in informal ways, more formal procedures are possible, such as
through your home owners’ association, your company, the court system, etc. The importance of cost
effective conflict resolution cannot be understated. Consider the resources expended on litigation in
the U.S. Litigating minor conflicts in hope of financial gain is enormously costly to society.

7.3.7 Minimal recognition of rights
When local users can organize themselves to craft their own rules, do national and local government
entities recognize and respect these arrangements? The lobster fishers in Maine organized themselves
in the 1920s and 1930s after the lobster population almost collapsed. The rules devised by these
organizations were informal arrangements among fishers. In the 1990s the federal government
wanted to reorganize fishery regulations along the east coast of the U.S. and have all fishers in all
states comply with the same regulations. The well-functioning lobster fishery system would have
been negatively affected by this, and significant efforts by fishers and scientists resulted in their
informal arrangements receiving legal recognition, permitting them continue as they had within the
framework of modern regulations.

7.3.8 Nested enterprises
When systems are larger, it may be necessary to have systems of governance at different levels.
What might be needed is a “polycentric” (many centers of governance and authority) system. Every
neighborhood may need to have policemen to patrol the streets, but not each neighborhood needs to
have a crime lab. One crime lab for the whole city might be a better solution in terms of having the
specific expertise available at a reasonable cost.
Polycentric systems emphasize approaching problems at the right level and ensuring that all
parties with some control over outcomes (centers of power) are involved. This may mean that some
collective action problems are addressed locally, while others are addressed at a regional or national
level. For example, in resolving disputes on the use of water from the Colorado River, it is not
productive to have Los Angeles and Phoenix debate this topic alone. Several states and the Bureau
of Reclamation need to be involved in order to develop a meaningful plan for allocating the available
water.
There are various challenges with polycentric systems. What is the right level of governance
76 Design Principles to Sustain the Commons

for each problem? Some lower level communities might not be able to self-organize, or they may
be dominated by local elites. How do higher-level authorities facilitate the local level governance
systems to succeed?
In a polycentric system, some units are general-purpose governments while others may be highly
specialized. Self-organized resource governance systems in such a system may be special districts,
private associations, or parts of a local government. These highly specialized governance units are
nested in several levels of general-purpose governments that provide civil, equity, as well as criminal
courts.
A university is often organized as a polycentric system. Each department has a certain level
of autonomy in offering courses. New courses can be offered, but to get them permanently on the
books and count for college-wide credits, they have to be approved at the college level. Some basic
requirements hold for all the majors, while the rest of the course work can be tailored to the specifics
of the major.
Besides departments, there are specialized organizations and services such as fraternities and
sororities, research institutes, financial aid offices, and libraries. Although there is a university
president who oversees the whole university, most units have a lot of flexibility within the general
constraints set by the higher levels of authority. If there is a lack of autonomy such that each decision
at the local level requires approval from “above,” a large organization will grind nearly to a halt
because of the transaction costs associated with decision making and many decisions will be made
by higher level officials without proper knowledge of the detailed practical problems associated with
the decision.

7.4 Using design principles in practice
The design principles were originally proposed as hypotheses based on analysis of several case
studies. In the more than 20 years since they were developed, they have held up to scrutiny. Although
some people may interpret the design principles as blue prints for designing robust institutional
arrangements, they are not. They are observed regularities derived by looking at cases after the fact.
So how can we use the design principles in practice?
One way to use design principles is to translate them into questions concerning how to improve
institutional arrangements for governing the commons. For example, for local resource users we can
ask:
• How can we better define the boundaries of this resource and of the individuals who are using
it so as to make clear who is authorized to harvest and where harvesting is authorized?
• How can we clarify the relationship between the benefits received and the contributions to the
costs of sustaining this system?
• How can we enhance the participation of those involved in making key decisions about this
system?
• Who is monitoring this system and do they face appropriate incentives given the challenge of
monitoring?
• What are the sanctions we are authorizing and can they be adjusted so that someone who
makes an error or a small rule infraction is sufficiently warned so as to ensure longer-term
compliance without our trying to impose unrealistic sanctions?
• What local and regional mechanisms exist to resolve conflicts arising over the use of this
resource?
7.5 Threats to sustainable use of the commons 77

For design principles seven and eight, questions need to be addressed at a higher level of
governance.
• Are there functional and creative efforts by local appropriators to craft effective stewardship
mechanisms for local resources that should be recognized?
• How do we create a multiple-layer, polycentric system that can be dynamic, adaptive, and
effective over time?
These are not, of course, the only questions local resource users and officials should ask in
an effective process of crafting institutional arrangements, but they can be thought of as a good
beginning.

7.5 Threats to sustainable use of the commons
Even as we look carefully to the design principles for guidance, there are various threats to the
sustainable use of common-pool resources. We list a number of them below based on what we have
learned from many case studies.

7.5.1 Rapid exogenous changes
Rapid changes in technology or population numbers can become a challenge to effective governance
of the commons. A new technology that enables fishers to catch more fish with the same amount of
effort, may render rules on where and when to fish useless unless restrictions on gear are implemented.
The provision of music, movies and books has experienced many challenges during the last few
decades with the emergence of new physical devices to digital files. These challenges have caused
the need for change in the original business structure of creative activities in order to remain viable.
A musician cannot depend on the sales of music records anymore, but must find other ways to earn
revenues from his or her creations.

7.5.2 Translation failures
Informal arrangements can be translated into official rules. For example, when writing the bylaws of
a homeowners association the rules are often based on informal practices. Over time the reasons why
some formal rules are written in the books may be lost, which may lead to problems.
For example, when a simple majority rule is used to make decisions, one can push forward
with important decisions when the minimum number required is reached, but this might not be best
for the viability of the community. A slight majority means that almost as many members of the
community oppose the decision. Leaders who rely on minimal majorities for too many decisions
may find themselves having to use coercion and/or corruption, rather than general agreement, to
keep themselves in power.

7.5.3 Blueprint thinking
The fact that some rules work out well in one action situation does not mean that those same rules
will work well for other situations. The rules and regulations for urban planning in Boston might
not work well in Phoenix given the many differences in the biophysical and social context. Rules
of the game of ice hockey may not work well for field hockey. Nonetheless, we see blueprint
thinking frequently, especially when large organizations implement many projects, as is the case
in development agencies. The World Bank may implement projects on community development
78 Design Principles to Sustain the Commons

in many places in the world, but requires each project to use the same blueprint policies to receive
funding.

7.5.4 Corruption and rent seeking
When individuals in power have the opportunity to allocate resources, there is always the possibility
of corruption. In these situations various actors may collaborate to harvest subsidies or large
infrastructure investments. When a bridge needs to be built, will the money be spent to build it
according to the specifications, or can the inspector be bought off to save on expensive, high quality
construction materials?

7.5.5 Lack of large-scale supportive institutions
Small-scale communities can be very effective in self-organizing and sustaining their shared resources
but will eventually experience challenges in the long term if they do not have the support of larger-
scale institutions. For example, when efforts are coordinated at a large scale, scientific information
can be collected and analyzed with expertise and resources that are not possible at the small-scale
(it is difficult to build a particle accelerator by yourself). Farmers, for example, may receive help
from highly trained professionals on new technologies and methods. Agricultural extension agencies
provide a supportive role in disseminating knowledge, information and experience that farmers
would not have the time to gather as individuals working alone.

7.6 Critical reflections
Based on the analysis of many case studies of fisheries, forestry, irrigation systems and other long-
lasting social-ecological systems, design principles can be identified. Those design principles are
not blueprints for design but are guiding principles to analyze institutional arrangements and help
researchers and practitioners to ask appropriate questions to improve the governance of the commons
as needed.

7.7 Make yourself think
1. What are examples of boundaries that are not clearly defined?
2. Neighborhood watch programs are an example of self-monitoring. What are the pros and cons
of such programs? How about neighborhood Home Owners Associations (HOA’s)? Have you
ever heard someone say about the new house they bought “Well, the good thing is that there is
NO HOA”.
3. When economic times are tough, politicians often recommend centralization of government
function to save money. Why is centralization of governance structures in cities not necessarily
an effective way to save monetary resources?

7.8 References
Cox, M., Arnold, G., & Villamayor Tomás, S. (2010). A review of design principles for community-
based natural resource management. Ecology and Society, 15(4), 38.
Ostrom, E. (1990). Governing the commons: the evolution of institutions for collective action.
Cambridge University Press.
Ostrom, E. (2005). Understanding institutional diversity. Princeton, NJ: Princeton University Press.
Part III

Human Behavior
Key Concepts
In this chapter we will:
– Learn what laboratory experiments are in the social sciences
– See how people make decisions in controlled social dilemmas
– Explore trust, ultimatum and dictator games
– Discover that experiments in jungles in the Amazon lead to
different results

8 — Social Dilemmas in the Laboratory

8.1 Experiments in social science
In previous chapters you have learned about the structure of action situations. In action arenas
individuals interact in action situations. Recall our example of the chess game as an action situation
that becomes an action arena when two people sit down to play. Every chess game is played the
same way—i.e., the structure of the action situation does not change over time. But the outcome of
every chess match is different. The decisions the players make in the action arena animate the chess
action situation. But how do people make decisions in such situations? In Chapter 4 we assumed
that decisions were made as if people behave selfishly and rationally. This chapter will show that
this is not the case in practice.
One of the ways to study how people make decisions in action situations is through the use of
experiments. There are many types of experiments but in this chapter we focus mainly on controlled
experiments in laboratory situations. Experiments are an excellent way to connect action situations
with social dilemmas. When we design an experiment, we have to be explicit about the structure
of the action situation. We can construct specific action situations that mimic social dilemmas that
people encounter in the real world. We can then systematically vary particular characteristics of the
action situation to determine how changing such characteristics affects peoples’ decisions. Indeed,
as we discuss various experiments below, you will see how changes in the action situation affect the
way people make decisions. Finally, data from experiments can be compared to observations from
case studies.
We begin by clarifying what we mean by laboratory
experiments in the social sciences. In a typical experiment,
the experimenter creates a situation where a number of
participants make decisions in a controlled setting. Often
experiments are performed at universities using undergrad-
uate students as participants. Maybe you have participated
in experiments by this point in your academic career?
Participants voluntarily consent to take part in an ex-
periment. They receive instructions on the actions that
can be taken and the consequences of these actions that Figure 8.1: Set up of an experimental labo-
ratory.
eventually result in monetary rewards. Decisions are made
in private. In a typical lab, participants sit at a desk with
82 Social Dilemmas in the Laboratory

a computer, surrounded by partitions that prohibit them from talking with others or viewing other
participants’ computer screens (see Figure 8.1).
The experiments are designed and performed to test scientific hypotheses. As you will see in this
chapter, the participants in the experiments make decisions in an artificial setting (controlled action
situation). The goal of an experiment is not to recreate reality, but to focus on a specific situation A
and test the decisions people make compared to situation B.
For example, in Chapter 4 we discussed the prisoner’s dilemma. We considered what the
outcomes would be in theory for selfish rational individuals (they will not cooperate). But what
will real people actually do? What would you do? Since the 1950s, many experiments have been
performed with the prisoner’s dilemma action situation. Typically we see that 50% of the people
cooperate in the prisoner’s dilemma experiment (Sally, 1995). This is considerably different to the
theoretical prediction that 0% would cooperate.
In the remainder of this chapter we will explore some classic social dilemma experiments. For
each of these experiments we will first analyze the action situation to determine what outcomes
theory would predict for selfish rational actors and then discuss actual empirical outcomes. We end
the chapter by discussing examples of these experiments that have been replicated by anthropologists
with tribes in the jungle.

8.2 The trust game
We begin with the so-called “trust game” (Figure 8.2).
This two-person game was developed by Joyce Berg, John
Dickhaut and Kevin McCabe in 1995 to focus on the
factors that affect the likelihood that an individual will take
an action whose outcome may be potentially costly when
that outcome depends on actions of a second individual.
In simple terms, this action requires that the individual
places trust in a second individual. Think about giving
somebody your credit card information. Can you trust
that the other person will not misuse this information?
The more people trust each other, the more cooperation is
possible. Of course, if people easily trust each other, they
are also easy victims of selfish rational individuals who Figure 8.2: A physical trust game. Will they
catch him?
“free ride” on the trust of others.
Using the terminology of the action situation, the sim-
plest baseline trust game is composed of the following elements:

1. Participants. Two persons play the game.
2. Positions. The two positions are player 1 (investor) and player 2 (trustee).
3. Actions. Player 1 is given 10 tokens at the beginning of the game. Player 1 then has the choice
of how many of those 10 tokens to keep and how many to “send” to player 2. Choosing to
“send” tokens to player 2 has the impact of increasing the value of the tokens. In typical trust
experiments, the number of tokens sent by player 1 is doubled or tripled (see point 5 for an
example). After the experimenter has increased the amount that player 1 has sent, this amount
is sent on to player 2. Once player 2 has received this amount, player 2 makes a decision.
Player 2 has to decide on the number of tokens to return to player 1.
8.2 The trust game 83

Figure 8.3: Trust game with the payoffs used by Berg et al.

4. Outcomes. The outcomes are the size of the funds allocated to the two persons in light of the
decisions they have made.
5. Action-outcome linkages. The amount invested by player 1 in player 2 earns a rate of return
(supplied by the experimenter) of 1 + r. If r = 2, then if player 1 gives 10 tokens to player 2,
player 2 receives 30 tokens. If player 1 gives 2 tokens, player 2 receives 6 (= 2 * 3) tokens.
6. Information. Both players are informed of the complete range of possibilities and that their
own identity will remain anonymous to the other player (and to the experimenter if a double
blind experiment is performed).
7. Potential payoffs. The payoffs are affected by the rate of return (1 + r). In most trust
experiments, r is assumed to be 2. This means that the amount that player 1 sends to player 2
is tripled by the experimenter. The payoff to player 1 is (10 - X) + Y where Y is the number of
tokens that player 2 returns to player 1. The payoff for player 2 is (1 + r)*X - Y (note if X is
zero, player two’s payoff is zero).

If this game seems a bit strange to you, consider for a moment your bank account. Your
relationship with your bank represents an action situation very similar to this game. Namely, you,
player 1, invest your money in the bank (put it in a savings account or “send” it to the bank). The
bank, player 2, may or may not give you your money plus interest, i.e., 1 + r times what you gave the
bank back.
A self-interested trustee (player 2) wanting to maximize returns (behaving as homo economicus)
would return zero to the investor (player 1). If the investor expects this, no funds will be invested
(transferred to the trustee) in the first place. Let’s reconsider the bank example again. If your bank
wants to maximize its returns, and only plays with you once, it will just keep your money. If you
know that this is a possibility, you won’t give them your money. Establishing trust between you
and your bank is thus essential for potentially valuable transactions to take place. Loss of this trust
causes runs on banks and precipitated such catastrophic events as the great depression after the crash
of the New York Stock Exchange in 1929.
The baseline trust game is barren of many of the social factors that are thought to affect trust.
The players are strangers and do not even see one another. There is no way that they can establish a
link to one another through making promises or the like. The trust game is similar in structure to a
prisoner’s dilemma if players make decisions sequentially. The baseline game represents a situation
requiring trust in about as pristine a form as one can imagine.
Figure 8.3 shows the trust game once more, using r equal to 2. The investor gives X tokens to
the trustee who then earns 3X - Y where Y is the amount returned by the trustee to the investor. The
investor thus earns 10 - X + Y. If the investor does not trust the trustee she may choose to invest 0,
i.e., set X = 0. As a result the trustee will not receive anything.
When the first experiments were performed by Berg and her colleagues, they used a double-blind
experimental protocol. In this protocol the experimenter will not know the identity of the participants.
Each token in the game was worth one dollar. From the 32 persons who were in the position of
investor (player 1), 30 sent money ($5.16 of the $10 on average) to the trustee (player 2). Of the 30
84 Social Dilemmas in the Laboratory

subjects in the position of trustee, 18 returned more than $1.00 ($4.66 on average) and one-third
of them sent more funds to the investor than the investor sent them (before being multiplied by
three). On average those investors who sent $5.00 or more received an average return in excess of the
amount they sent. It was those investors sending less than $5.00 who received a negative net-average
return. In other words, those investors who trusted their counterparts the most were the ones who
left the game with more wealth than those who were less trusting.
Although this experimental design of the trust game is simple, it captures the essence of trusting
and reciprocal behavior so effectively that it has been replicated and extended in many different
settings and countries.
Some of you may be surprised at the high levels of trust exhibited in this experiment. The
common theory used to make predictions in this action situation assumes selfish rational behavior
and predicts no investment. The investor is not expected to trust an unknown stranger enough to
send any funds. The empirical data challenges the conventional theory.

8.3 How people make decisions
We have mentioned that theory predicts that participants will act as selfish rational human beings
when faced with a decision that involves costs and benefits to themselves. Yet we have just discussed
a set of experiments that show that people do not make decisions in that way. So, why do we make
these assumptions? These assumptions relate to rational choice theory, an important theory in
the social sciences. Rational choice theory is a framework for understanding economic and social
behavior (we define it more precisely below). It has proven to be valuable in predicting human
behavior in stable, competitive market settings. In stable and repetitive settings, individuals are
able to learn about the full, relevant structure of the situation and attach preferences to actions and
outcomes.
Predictions from these models are empirically supported at an aggregate level in open competitive
market settings and at an individual level in carefully designed experimental settings of competitive
market situations (see the work of Nobel Laureate Vernon Smith). What are the assumptions of
rational choice theory? They are:

1. Individuals possess as much information about the structure of a situation as is contained in
the situation itself. That is, they have perfect information about the world around them and the
situation they are in.
2. Individuals assign a complete and consistent, internal valuation to outcomes that is a monotonic
function of an individual’s own net external payoff. Put simply, this means that individuals
will always prefer more units of a product to less units. Furthermore, if an individual likes
product A more than product B and product B more than product C, then this individual will
also like product A more than product C.
3. After making a complete analysis of the situation, individuals choose an action in light of their
resources to maximize expected material net benefits to themselves given what they expect
others to do.

We will call participants whose behavior can be predicted by these three assumptions rational
egoists. When we study institutions, individuals in many situations do not meet the specific as-
sumptions of rational choice theory. The rational egoist might be a good starting point to predict
human social and economic behavior, but we know it cannot explain all observed behavior. Unfortu-
8.3 How people make decisions 85

nately, there is not a widely accepted alternative theory. Such a theory should include assumptions
concerning

1. the way that participants acquire, process, represent, retain, and use information;
2. the valuation that participants assign to actions and outcomes, especially the outcomes that
others experience; and
3. the processes (heuristics) that participants use for selecting particular actions or strategic
claims of actions in light of their resources.

Given the many different assumptions that are involved in a more mature theory of human
behavior, a more realistic theory would not be as simple, elegant and decisive as rational choice
theory. We cannot discuss all the possible assumptions of this enormous body of literature that could
be involved. We do, however, discuss some of the most important ones.

1. There is increasing evidence that a crucial element in decision making is the evaluation of
the outcomes for the decision-maker and others whom their decisions might affect. Many
people value fairness and equity and prefer a more equal distribution of payoffs rather than
maximizing their own earnings. Why people have these values is not clear. An important
factor is the social norms people acquire during their lifetime regarding what they are supposed
to do. Not surprisingly we find norms like “Do unto others as you would have them do unto
you” that serve as guiding principles in educating young people in how to behave in their
interactions with others.

“Love your neighbor as yourself”
“Do unto others as you would have them do unto you”

2. One of the main strategies that we find people use is conditional cooperation. People cooperate
if others do so too. This means that in repeated situations involving social dilemmas, people
may adjust their decisions if they observe others are more or less cooperative than they had
expected. “One bad apple can spoil the whole bunch” is a common saying that illustrates the
principle that one persistent defector can trigger others to change to defection too.
3. People have incomplete information and an incomplete understanding of the world around
them. Humans develop mental models as a consequence of experiences and information they
have received. As such, mental models are a product of experiences during childhood, the
culture people come from, and the ability to process the information received through life
experiences. In an action situation participants have incomplete information but use their
mental models to infer what others will do, what others will think about them, whether others
can be trusted, and what the payoff matrix means for them. Communication may play an
important role because communication enables people to share information to update the
expectations they have about each other.
4. What is clear from the empirical analysis of many social dilemma experiments is that there
is not one method by which people make decisions. We can detect various strategies, and
allowing for heterogeneity of strategies across individuals is crucial in explaining observations
from experiments. If you are in a social dilemma with people you know very well, it is easier
to judge what they will do and vice versa. Under these circumstances, it is more common that
groups will cooperate.
86 Social Dilemmas in the Laboratory

8.4 Ultimatum and dictator games
We will now consider two other games that provide additional insights into how people make
decisions in social dilemmas. Both games are called one-shot games, meaning that they are played
for only one round. The first game is the ultimatum game, and the second game is called the dictator
game.
In the ultimatum game there are two players who interact anonymously, without communication
and only once. The first player receives a sum of money (called the “endowment”) and then has the
following decision to make: “decide how to share the endowment you received between you and the
other person.” It is important to know that there is no communication and the participants do not
know each other’s identity and will never know each other’s identity. The offer the first player makes
to the second player can be accepted or declined by the second player. If the second player declines
the offer, both players receive nothing. If the second player accepts the offer, the endowment is
divided the way the first player proposed.
Suppose that player 1 receives 10 one-dollar bills. A selfish rational player would accept any
offer. Player 1, thinking that player 2 would behave selfishly and rationally would therefore offer a
minimum amount—say one dollar to player 2. Player 2, being rational would accept 1 dollar, because
1 is better than 0, which is the only other available option (i.e., decline the offer). The ultimatum
dilemma was proposed in 1982 by Werner Güth and his colleagues and today is a frequently used
experimental design. Experiments on ultimatum games have been performed all around the world.
In many cultures, people offer a fair split (e.g., 50:50), and offers of less than 20% are often rejected.
The persistence of this result indicates that, in general, people do not make decisions as the rational
egoist model would predict.
A common explanation of these empirical findings is that people evaluate aspects of the potential
outcomes of their decisions other than monetary benefits, such as the relative distribution of benefits
between the player herself and others. Let’s now consider the ultimatum and dictator game using the
terminology of the action situation:

1. Participants. Two persons play the game.
2. Positions. The two positions are participant 1 and participant 2.
3. Actions. Participant 1 is given X at the beginning of the game. Participant 1 then decides how
much to keep and how much to offer participant 2. In the ultimatum game participant 2 can
choose to decline the offer, which results in zero earnings for both participants. In the dictator
game participant 2 does not make a decision.
4. Outcomes. The potential outcomes are the different allocations of the endowment to the two
persons in light of the decisions they have made.
5. Action-outcome linkages. The amount offered by participant 1 affects whether participant
2 will accept the offer in the ultimatum game. In the dictator game, the amount offered by
participant 1 affects only the final allocation.
6. Information. Both players are told the full tree of possibilities and that their own identity will
remain anonymous to the other player.
7. Potential payoffs. The payoffs are affected by the decision of participant 1, and in the ultimatum
game whether participant 2 will accept the offer or not.

The dictator game is described schematically in Figure 8.4. Player 1 makes a decision about how
much (denoted by “X”) to give to player 2. Once player 1 makes the decision that marks the end of
the game. Player 2 does not make a decision. Figure 8.5 shows the ultimatum game. In this game,
8.5 Bringing the lab to the field 87

Figure 8.4: Dictator game.

Figure 8.5: Ultimatum game.

player 2 can decide to accept the offer or not. If accepted, the distribution of the endowment is the
same as proposed by player 1, otherwise both players have zero dollars.

8.5 Bringing the lab to the field
How general are the findings of the experiments performed
with undergraduate students at American and European
universities? Do these decisions represent the entire pop-
ulation or are they just artifacts of a western cultural her-
itage? Anthropologist Joseph Henrich started to perform
ultimatum experiments in the Amazon with communities
who have not had many interactions with western societies
(Figure 8.6). Doing such experiments is very challeng-
ing. One has to work with a population who might not
read or write, who may not use money, or possibly speak
a rare language. Furthermore, field experiments are not
conducive to computer-based experiments like those per- Figure 8.6: Joseph Henrich in the field.
formed in laboratories as discussed previously.
In the field, experiments are translated into the local
language. If money is not relevant to the community (there is nothing to spend it on), rewards will
be given in physical objects such as food items. Experiments are performed with pencil and paper,
or physical objects. People make decisions one at the time in private, and only later the decisions of
player 1 are matched with a randomly drawn player 2.
Henrich and colleagues used the same experimental design in 15 places around the world and
found a much wider diversity of decisions than earlier experiments that were held with people in
modern market economies. Some communities had a much lower level of offers in the ultimatum
game, and accepted a much lower level of offers (Figure 8.7). For example, the Machiguanga people
from Peru made decisions in the ultimatum game most in line with the prediction of selfish rational
persons. Others, such as the Lamelara in Indonesia offer more than 50% and do not reject any offer.
In Papua New Guinea, experiments in two communities led to a very high rejection rate. Often
participants rejected high offers. A possible explanation for this is that their custom regarding gift
giving is that if one receives a gift one must give a gift in return. Although they did not know the
88 Social Dilemmas in the Laboratory

Figure 8.7: Mean offer versus rejection rate in ultimatum game (adapted from Henrich et al. 2001).

identity of the other player, participants in these experiments still did not want to be obligated to
give a gift back.
The large variation in the data is, to some extent, related to the type of activities the com-
munities performed in their daily lives, ranging from hunting and gathering, to fishing, farming,
and pastoralism. Having said that, in general the populations’ familiarity with a market exchange
affected the level of the offerings more strongly that did other factors. In 2010, Henrich et al. (2010)
published a follow-up study where market integration was measured as the percentage of calories
purchased on the market (Figure 8.8). In the U.S. such a percentage is about 100 percent, while
some hunter-gatherer groups where completely self-sufficient and purchased 0% of their calories
from the market. Henrich and colleagues found that the more communities were integrated with the
market, the more they offered in dictator games.
Do markets lead to cooperation? The publication of the initial results led to a huge controversy.
The results seem to suggest that capitalism can lead to cooperative behavior. This is not the right
conclusion to make. What the results indicate is that in societies where it is common to exchange
goods with strangers, something we do in a market-based society, people become more comfortable
cooperating with strangers. For those who gather and hunt their own food and never have any
interaction with people beyond their extended family, participating in an experiment in which they
make a decision with an unknown other person in the broader community is a strange experience.
Since this is an uncommon experience, participants without market experience may not be generous
to the other participant.
In market-based societies there is a strong norm of fairness. But what about recent news reports
of bankers receiving exorbitant bonuses during a time of economic crisis? The protests in response
to this greedy behavior suggests that there is a strong sense of fairness. The fact that some do not
follow the social norms has also to do with power inequality. In the experiments all participants
are randomly allocated to player 1 and player 2. Since players do not know who their partner is,
reputation and power differences do not have a major effect on the experimental results. If players
8.6 Critical reflections 89

Figure 8.8: Mean dictator game offers for each population plotted against mean value of market integration
(Henrich et al. 2010).

did meet face to face, the outcomes of the experiment would no doubt be considerably different.

8.6 Critical reflections
Three social-dilemma experiments were introduced in this chapter: the trust game, the dictator
game, and the ultimatum game. These games can be analyzed using the action situation frame-
work. Experiments with these games in the lab and the field show that theoretical predictions that
people will act as rational egoists are falsified. There is not yet a general alternative theory of
decision making, but it is clear that most humans take into consideration the well-being of others
(including strangers) when making decisions. We also discussed how experiments can be used to
explore how the level of cooperative behavior is influenced by social context such as the level of
market integration.

8.7 Make yourself think
1. What do you expect will happen if two bankers play the ultimatum game?
2. And what if an ultimatum game is played where members of the 1% are player 1 and members
of the 99% are player 2?
3. Have you ever experienced a situation similar to the trust game?
90 Social Dilemmas in the Laboratory

8.8 References
Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity and social history. Games and
Economic Behavior, 10, 122–145.
Güth, W., Schmittberger, R., & Schwarze, B. (1982). An experimental analysis of ultimatum bar-
gaining. Journal of Economic Behavior and Organization, 3, 367–388.
Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., & McElreath, R. (2001). In
search of Homo Economicus: behavioral experiments in 15 small-scale societies. American
Economic Review, 91(2), 73–78.
Henrich, J., Ensminger, J., McElreath, R., Barr, A., Barett, C., Bolyanatz, A., . . . Ziker, J. (2010).
Markets, religion, community size, and the evolution of fairness and punishment. Science, 327,
1480–1484.
Ostrom, E. (2005). Understanding institutional diversity. Princeton, NJ: Princeton University Press.
Sally, D. (1995). Conversation and cooperation in social dilemmas: a meta-analysis of experiments
from 1958 to 1992. Rationality and Society, 7, 58–92.
Key Concepts
In this chapter we will:
– Be introduced to public good and common pool resource experiments
– Learn about the important roles of costly sanctioning and communication
– Observe that the findings in the lab are confirmed in the field
– Become aware of more dynamic and interactive experiments as the
frontier of experimental methods

9 — Self-governance in the laboratory

9.1 Introduction
In the last chapter we explored social dilemmas
involving only two players, playing one time only, Learning from peasants
and making very simple decisions. In this chapter
we will look at more complex experiments involving During field work in Nepal in the 1980s Elinor
public goods and common-pool resources. These Ostrom learned from interviews that farmers mon-
dilemmas are the main focus of this course. In Chap- itored each other’s water withdrawals and contri-
ter 4 we saw that these types of goods are difficult butions to maintenance. She also observed a dra-
to exclude others from using. Public goods, like matic event in which a group of farmers dropped
their work to repair a mud-based canal from which
Wikipedia, national defense, public radio and clean
a farmer was withdrawing water illegally.
air can be used by everybody, yet there are costs to Back in the U.S. Ostrom discussed these events
individuals to invest in the provision of public goods. with colleagues in economics. The initial response
Hence we can expect an under-provision of public from the economists was that it was illogical to
goods. With experiments we can test the assump- invest so much time in monitoring and enforce-
tion that public goods cannot be provided voluntarily. ment. Maybe peasants would do this, but students
The same is true for common pool resources, which never would do this self-policing. New exper-
iments were designed where participants could
are vulnerable to overharvesting.
give up earnings to reduce the earnings of others.
Although it is irrational for rational egoists to do
9.2 Common-pool resource action situa- so, actual participants did use this costly sanction-
ing. The important role of costly sanctioning is
tions
now well established.
In Chapter 4 we introduced the idea that with
common-pool resources, it is difficult to exclude or
limit other users’ consumption of the resource and
each person’s consumption of the resource reduces the availability for others. We will now discuss
how the basic dilemma in a common-pool resource setting can be translated into an action situation
that we will subsequently use in a laboratory experiment.
For ethical reasons, we cannot do experiments with real resources. Therefore we create an
artificial resource. Participants in the experiments earn a certain amount of money based on their
decisions, just as in the simple ultimatum and dictator games discussed in the last chapter. Below we
introduce a model that demonstrates that when more people harvest from the common-pool resource,
92 Self-governance in the laboratory

less money is earned per unit of effort by each individual. In the experiments, participants receive
tables in which they see the consequences of their decisions, given the decisions others make.
We need to start with a static, baseline situation that is as simple as possible without losing
crucial aspects of the problems that real appropriators face in the field. This static, baseline situation
is composed of:

1. Participants. A set of n individuals.
2. Positions. No differentiation exists in the positions these participants hold relevant to the
common-pool resource. In other words, there is only one position of resource appropriator.
3. Actions. Appropriators must decide how to allocate tokens assigned to them in each time
period. Basically, one can think of these appropriators as being “endowed” with a number of
tokens, e, which they are free to allocate in any proportion during each time period to each
of the two activities. To simplify the problem, we posit that all appropriators have the same
endowment (just as we all have only twenty-four hours per day), and face the same outside
opportunity. Thus, each round they have to decide how much of their endowment to devote to
appropriation from the common-pool resource or in gaining returns from an outside activity.
For example, if each appropriator is given 10 tokens (e.g., 10 hours of labor), they can use 5
hours to go fishing (use the common-pool resource) and 5 hours working for a fixed wage.
4. Outcomes. The action they take affects the amount of resource units that can be appropriated
from the common-pool resource or returns earned from the outside option.
5. Action-outcome linkages. Here we use a mathematical function to translate the actions of
all the appropriators into outcomes. While these functions are in reality frequently affected
by many variables in addition to the actions of individuals, we will consider only simple,
deterministic relationships between appropriation actions and their effects on the resource in
the baseline setting. For wage labor, the relationship (function) is simple: the amount of time
allocated to wage labor is simply multiplied by the standard wage. The resource function is a
concave function, F, which depends on the number of tokens, x, invested in harvesting the
common-pool resources.
6. Initially, the sum of all of the individuals’ harvesting actions, Σx, generates better outcomes
than the safe investment in wage labor. If the appropriators decide to allocate a sufficiently
large amount of their available assets (e.g., time) to harvesting the common-pool resource, the
outcome they receive is less than the alternative. In other words, allocating too many assets to
harvesting the common-pool resource is counter-productive.
7. Information. As an initial information condition (because of the instructions that are carefully
given to all participants), we assume that appropriators know the shape of the function
linking actions to outcomes and know that they are symmetric in their access to assets and
opportunities. Information about outcomes is generated after each decision round is completed.
Appropriators may not communicate with one another. It is assumed that each appropriator
will assume that all other appropriators are rational actors and will adopt the “best response” as
their own actions. If all players do, in fact, act as rational egoists, the best response functions
should lead all appropriators to overharvest from the resource.
8. Potential payoffs. Payoff functions specify the value of the wage rate and the value of
the resource units obtained from the common-pool resource. Specifically, the payoff to an
appropriator is the payoff from wage labor plus the payoff from using the common-pool
resource. If we call the total endowment of tokens (i.e., units of labor) e, and the wage ω, then
9.2 Common-pool resource action situations 93

the payoff function is given by

e·ω if xi = 0
(9.1)
ω · (e − xi ) + (xi / ∑ xi ) · F(∑ xi ) if xi > 0

Basically, if appropriators put all of the assets into the fixed wage option, they receive a certain
monetary return equal to the amount of their endowment times, an unchanging rate of return (ω). If
appropriators put some of their endowed assets into wage labor and the rest in harvesting from the
common-pool resource, they get part of their return from the wage labor (ω · (e − xi ) ) and the rest
from their proportional harvest of the common-pool resource as determined by function F.
Suppose participants receive 10 tokens (e.g., 10 hours of labor) each round (a round could
represent a day, so setting e equal to 10 is equivalent to allowing the players to work 10 hours per
day.) What share should we expect will be invested into the common-pool resource that maximizes
income of each player in the group? The function F() is depicted in Figure 9.1 for an experiment
with 8 participants. You can see that as more people invest in harvesting the common-pool resource
F will start declining at a certain point (blue line). Since F is the earnings people get from harvesting
from the common-pool resource, it follows that at a certain point participants will get less returns if
they keep on investing their effort in harvesting the common resource. We use a value of 0.05 for ω,
which means that participants earned 5 real cents for each token used for wage labor.

Figure 9.1: The value of the function F () for different levels of Σx is depicted as the blue line. The red line is
the value of F plus ω · (10 · e − Σx).

If the group could agree to work together, what would be the best thing to do? From Figure 9.1
it is clear that the best earning from the groups perspective would be to invest a total of around 40
tokens. At that point the red line, the net earnings, reaches the maximum level. To be precise, we
can calculate that the group earnings are maximal when the total investment is 36 tokens. This is an
average of 4.5 tokens per person. In that case the total earnings would be 7.24 units, which is 0.91
94 Self-governance in the laboratory

units per person. At the end of the experiment, the participants are paid in cash based on their payoff
“units.” Often, the units are cents. So after playing 20 rounds and earning 0.91 tokens per round as in
the example above, the participant would earn 20 x $0.91 = $18.20.
However, from an individual perspective, the problem looks different. If others do not appropriate
from the common-pool resource, a maximum harvesting investment of 10 will be most productive.
So if you invest labor in harvesting and no one else does, you will do well. But if everybody thinks in
this way, the total level of harvesting goes up. When each participant invests 8 tokens in harvesting
there is no opportunity for individuals to improve their earnings. Note that the individual earnings
in this case is a little above 0.5 which is substantially lower than the earnings of 0.91 realized if
everybody had invested an average of 4.5 units of labor in the common-pool resource and 5.5 units
in wage labor. However, if people are rational egoists, we expect they invest 8 tokens per person.
Like the tragedy of the commons, people keep adding sheep as long as the individual benefits are
larger than the costs that are shared by everyone else.
We will now discuss some of the results from the common-pool resource (CPR) experiments
that have been conducted using this action situation. Participants know that they are participating
in an experiment that will not take more than 2 hours. The number of rounds in each experiment
varies between 20 and 30. In addition to being told the payoff function specifically, participants were
provided with look-up tables that eased their task of determining outcomes depending on their own
and others’ decisions.
In the baseline experiments, the actual average investment level in the CPR is around 8 tokens,
as predicted if we assume all players act as rational egoist. Figure 9.2 shows that the investment level
stays the same over 30 rounds. However, although the average investment is 8 tokens, most people
don’t invest 8 tokens. Figure 9.3 shows that there is a large spread of investments. Most frequently
individuals invest 10 tokens. It never occurred, with any group or in any round, that all 8 participants
invested 8 tokens, as the theoretical equilibrium calculations for rational egoists predicted. Thus
although individuals experience the same action situations, the results differ among the action arenas
because people make different decisions.

9.3 Changing the CPR action situation
Now that we have the baseline results of the experiments in a CPR context we can explore how the
results will change if we make some small modifications to the action situation. These changes are
operationalized in the set of instructions given to participants and in the procedures adopted within
the experiment.

9.3.1 Communication
The first change that we will look at relates to the information component of the action situation.
Instead of forbidding communication among participants as in the baseline experiments, participants
are now allowed to communicate with one another on a face-to-face basis in a group setting before
returning to their own enclosed terminals to make their private decisions. The participants can talk
about whatever they want as long as they do not promise side payments or threaten each other. This
kind of communication is known as “cheap talk.” “Cheap talk” describes agreements and promises
that can not be enforced. If participants behave according to the traditional model of rational egoists,
the option of cheap talk should have no effect. Promises cannot be enforced and the decisions are
private. If participants’ decisions were known to each other, there might be reputation effects, like
people getting bad feedback on eBay.
9.3 Changing the CPR action situation 95

Figure 9.2: Average investment level per person in the common-pool resource. The dotted line represents the
average +/- the standard deviation.

Figure 9.3: Distribution of the investment levels into the CPR.
96 Self-governance in the laboratory

In the basic communication experiments held at Indiana University in the late 1980s, participants
first made ten rounds of decisions in the context of the baseline appropriation situation with no
communication. After the tenth round, participants listened to an announcement that told them
they would have an open group discussion before each of the next rounds of the experiment. The
participants left their terminals and sat in a circle facing one another. After each discussion period,
they returned to their terminals to enter their anonymous decisions. Participants used face-to-face
communication to discuss which strategy would gain them the best outcomes and to agree on what
everyone should invest in the subsequent rounds. After each decision round, they learned what
their aggregate investments had been, but not the decisions of individual players. Thus they learned
whether total investments were greater than the total investments they had earlier agreed upon.
While in many rounds participants did exactly as they had promised one another they would do,
some defections did occur. If promises were not kept, participants used this information about the
aggregate investment levels to castigate the unknown participant who had not kept to the agreement.
This opportunity for repeated face-to-face communication was extremely successful in increasing
joint returns. In the 10-token experiments, participants obtained close to 100 percent of the maximum
available returns. Only in 5% of the cases did a participant invest more in the common-pool resource
than was agreed upon. The results show that a random group of undergraduate students quickly
reached the group maximum, while the prediction using the rational egoist model was that they
would still overharvest the resource. Although there was no formal punishment possible, participants
used strong language if they detected somebody had invested more than what was agreed upon.
The participants had probably internalized norms regarding the importance of keeping promises.
They were also not shy to express anger if participants broke their promises and those who did
promised to do better next round.

9.3.2 Costly sanctioning
Participants in field settings are frequently able to communicate with one another on a face-to-face
basis, at least from time to time, either in formally constituted meetings or at social gatherings. In
many field settings, where the resource has been sustained over a long time, participants have also
devised a variety of formal or informal ways of monitoring and sanctioning one another if rules
are broken. Engaging in costly monitoring and sanctioning behavior is, however, not consistent
with the theory of a norm-free, perfect rationality. Costly sanctioning means that individuals have
the opportunity to pay to reduce earnings of another participant in the experiment. Thus, new
experiments were held to test whether participants would actually pay from their own earnings in
order to sanction the less cooperative behavior of other participants.
As in the communication experiment, participants first participated in ten rounds of the baseline
game. Participants were then told that in the subsequent rounds they would have an opportunity to
pay a fee in order to impose a fine on the payoffs received by another player. The fees ranged across
many different experiments from $0.05 to $0.20 and the fines from $0.10 to $0.80. In brief, the
finding from this series of experiments was that much more voluntary costly sanctioning occurred
than the zero level predicted by the rational egoist model.
Participants react both to the initial cost of sanctioning and to the fee-to-fine relationships. They
sanction more when the cost of sanctioning is less and when the ratio of the fine to the fee is higher.
Sanctioning is primarily directed at those who invested more in the common-pool resource. A few
sanctions, however, appear to be a form of “blind revenge.” These were fines made by participants
who had themselves been fined by unknown others for their high levels of investment. In these few
9.4 CPR experiments in the field 97

cases, the sanctioners picked on those whose investments were lower than others, and thus were
suspected of being the ones who had previously sanctioned them.
In this set of experiments, participants were able to increase their returns since participants
reduced their investments, although they did not reach the maximum earnings. However, if we look
at the net earnings, after subtracting the costs of the sanctioning system, we don’t see any benefit
from sanctioning. On the other hand, the performance was greatly increased if the participants had
the opportunity to communicate for just one round.
Outside of the lab, participants rarely impose sanctions on one another that have been devised
by an external entity, as in the experiments above. In the field, sanctions are much more likely
to emerge from an endogenous process of communities crafting their own rules along with the
punishments that should be imposed if these rules are broken. Spending time and effort in a linked
collective-choice situation designing rules creates a public good for all of those involved. Crafting
rules for an operational situation is thus a second-level dilemma.
The traditional theory of the rational egoist predicts that people will not invest time and effort to
create rules. That is the reason why many people argue that rules must be designed and imposed by
external authorities who then assume official responsibility for monitoring and enforcing these rules
and are paid a salary for their work. Since self-organized rules are found in many local common-pool
resource situations, it appears that participants frequently do design their own rules contrary to the
theoretical prediction. But these processes are difficult to witness in the field.

9.4 CPR experiments in the field
One of the great advantages of laboratory experiments is that they can be replicated very accurately,
and modified for new experiments so that one can gain even greater confidence in the findings. One
of the critiques of laboratory experiments is the use of undergraduate students, instead of “real”
people (i.e., will we get the same findings if we do experiments with people who work on natural
resource management every day?).
Juan-Camilo Cardenas has performed a series of field experiments by setting up CPR action
situations in school buildings in rural Colombia where participants made decisions in paper and
pencil experiments (Figure 9.4). Cardenas invited local villagers who were actual users of local
forests for the extraction of firewood, natural fibers, and log timber as well as local water resources.
In order to do these experiments, the instructions were
written in Spanish and in a manner that would be easily
understood by villagers. Instead of tokens—which are an
easy medium for undergraduates to understand—he asked
villagers to decide on how many months a year they would
spend in the forest gathering wood products as contrasted
to using their time otherwise. Each villager had a copy
of a payoff table, which was the same as that of the other
seven participants, showing that as the number of months
that each individual would spend in the forest increased,
she would gain more returns, but that the return to all of
them depended on keeping their harvesting time to a very Figure 9.4: Juan Camilo Cardenas explain-
low level. ing an experiment in the field in Colombia.
Would the findings from experiments with villagers
in rural Colombia be similar to those conducted with undergraduates at universities in the U.S.?
98 Self-governance in the laboratory

In the baseline, no-communication experiments, Cardenas found a similar pattern as in laboratory
experiments. Villagers substantially overinvested in the CPR. The average earning for a participant
was about one day of paid labor for an exercise of two or three hours. However, they could double
their earnings if the participants could agree to cooperate. Face-to-face communication significantly
increased the performance of the groups, although not quite to the maximum level.

9.5 More dynamic CPR experiments
So far, the experiments all have the following sequence: (1) Participants each make private decisions
on how much to appropriate from the CPR by investing tokens (time) in appropriation activites. (2)
When everybody has made their decision, the aggregated information about the consequences of
the decisions of the group are shared with the participants. (3) Then a new round starts where the
participants have the same payoff information.
In reality people make decisions all the time. Some make more decisions than others. And some
decisions cause immediate consequences (if the irrigator takes too much water upstream, downstream
irrigators immediately experience a shortage of water). To better approximate real-world decision
making, new experiments were developed using computer simulations where participants make
decisions in real-time in a spatially explicit dynamic environment (Janssen, Holahan, Lee, & Ostrom,
2010). This experiment includes more realistic ecological dynamics. Will the results of earlier
experimental designs still hold?
Let’s consider this more complex environment in a bit
more detail. The participants collect tokens from a shared
renewable resource environment. Groups consist of five
participants who share a 29 x 29 grid of cells (borders not
wrapped around) (Figure 9.5). At the start of the experi-
ment, 25% of the grid space is filled with tokens, thus 210
tokens are randomly allocated on the board. Each partic-
ipant is assigned an avatar, which they are able to control
using the arrow keys on the computer keyboard to move
up, left, right and down. The avatars are initially placed
in the middle row of the screen with equal distances be-
tween the avatars. When the participant wishes to harvest
a token, they must position their avatar on top of a token
and press the spacebar. Each token harvested is worth
a certain amount of money defined by the experimenter.
Figure 9.5: Spatially explicit resource.
Participants have complete information on the spatial po-
sition of tokens and can watch the harvesting actions of
other group members in real time (the entire screen is visible to the participants). Furthermore, they
can see the total number of harvested tokens of all the participants at the top of the screen.
Every second, empty cells have a possibility of generating a new token (in the same way that
land can generate new resources over time). The probability, pt , that a given empty cell will generate
a new token is dependent on the number of adjacent cells containing tokens. The probability pt is
linearly related to the number of neighbors: pt = P ∗ nt /n where nt is the number of neighboring
cells containing a token, n is the total number of neighboring cells (n = 8), and P = 0.01). If an
empty cell is completely surrounded by eight tokens, it will generate a token at a higher rate than an
empty cell that abuts only three tokens (can you think of the ecological equivalent of this process?).
9.6 Public good action situations 99

The model space is not toroidal, so cells located at the edges of the screen have fewer neighbors than
interior cells. For example, a corner cell only has 3 neighbors (N = 3). At least one adjacent cell
must contain a token for a new token generation to occur. Therefore, if participants appropriate all
of the tokens on the screen, they have exhausted the resource and no new tokens will be generated
(i.e., they have caught every last fish, so there are no more fish to reproduce, and the fish population
becomes extinct). By designing the environment in this manner, the experiment captures a key
characteristic of many spatially-dependent renewable resources.
The results of the experiments show that we see a tragedy of the commons in experiments
without communication or costly sanctioning. Within two minutes the individuals have collected the
last available token on the screen and spend the remaining 2 minutes of time in the round staring
at a black screen (Figure 9.6). If we repeat this situation, we get the same results. This means that
there is no learning that might avoid the tragedy of the commons. If we allow participants to chat
for a few minutes, the performance is significantly increased. We can see this in Figure 9.6 since
the resource stays at a higher level, leading to more regrowth of the resource. The earnings increase
from 53 to 94 tokens per round per person. Note that here the participants do not communicate
face-to-face but only exchange messages via a text screen. Nevertheless the group can achieve a
result that is substantially better than without communication. Interestingly however, if we allow
costly punishment, we do not see an improvement if it is not combined with communication. These
experiments show that communication is much more important than the ability to sanction each
other.

Figure 9.6: The resource size (number of tokens) over time for rounds with and without communication.

9.6 Public good action situations
Public goods are goods that people can consume without reducing the available amount left over for
others. Examples are national defense, broadcast television, information on the Internet, levies, etc.
However, the provision of the public good requires investments by individuals. The dilemma is to
get individuals to invest effort in creating a public good that everybody, even those who may not
100 Self-governance in the laboratory

have contributed, will enjoy.
Public goods are widely studied in experimental settings. We can describe the action situation in
the following way:

1. Participants. A set of n persons.
2. Positions. No differentiation exists in the positions these participants hold relevant to the
public good. In other words, there is only one position of contributor.
3. Actions. Contributors must decide how to allocate tokens assigned to them in each time period.
Basically, one can think of these contributors as being “endowed” with a number of tokens, e,
which they are free to allocate in any proportion during each time period to two activities. In a
field setting one can think that every day, each appropriator must decide between spending
time trying to write articles for Wikipedia or using their time to earn money in an office job.
To simplify the problem, we posit that all contributors have the same endowment (just as we
all have only twenty-four hours per day), and face the same outside opportunity. Thus, they
have to decide how much of their endowment to devote each round to investing into the public
good or in gaining returns from an outside option.
4. Outcomes. The actions participants take affect the size of the public good generated and shared
among all the participants.
5. Action-outcome linkages. The function maps the actions of all the contributors into outcomes.
Initially, the sum of all of the individuals’ contributions, Σx, generates a public good. The
more the contributors decide to allocate to the public good, the larger the public good will be.
6. Information. As an initial information condition (because of the instructions carefully given to
all participants), we assume that contributors know payoff function. Information about out-
comes is generated after each decision round is completed. Contributors may not communicate
with one another. It is assumed that each contributor will assume that all other contributors are
rational actors and will adopt the “best response” as their own actions. These best response
functions should lead all contributors not to contribute.
7. Potential payoffs. Payoff functions specify the value of the wage rate (ω ) and the marginal
per capita return (r). An experimental linear public-good game involves a free-rider problem if
r < 1 and n · r > 1. Suppose, in a given round, individual i contributes xi of ω for the provision
of the public good. The participant’s payoff is:

n
ω − xi + r ∑ x j (9.2)
j=1

You can see that the structure of this public good action situation is similar to the CPR action
situation. The key difference is the potential payoff from different allocations. The equilibrium
prediction, assuming individuals maximize their own monetary payoffs, is that the public good will
not be provided at all. The persistent finding from experiments is that initially people invest roughly
half their endowment in the public good. In further rounds, the level decreases over time to a level
that depends on the marginal per capita return.
Figure 9.7 shows the basic results with a group size n = 4, r = 0.4 and ω = 20. Thus when all
participants invested their complete endowment to the public good, each participant earned 32 units,
which is 60% higher compared to the situation in which none of them invest anything at all.
These results are published in Fehr and Gächter (2002). In this series of experiments, they tested
9.6 Public good action situations 101

the effect of costly sanctioning (called punishment in the original article). With costly sanctioning
the following decision was added after the investment decisions were made. Each participant got the
information about the investment decisions of every other group member, and the participants could
then decide how much to invest in sanctioning. Up to 10 units could be invested in costly sanctioning.
For each token invested in sanctioning, the other participant’s earning was reduced by 3 units.
A clever aspect of this particular study was that many groups played during the same time.
Each round, the computer rearranged the groups during a series of 6 rounds. As a consequence, a
participant was never in a group with the same person for more than one round during a treatment.
This was done to avoid reputation effects. Variations in the level of cooperation could have been
caused by the reputations of participants in the group.
In rounds without costly sanctioning, we see a rapid reduction of the investment level, while
inclusion of costly sanctioning leads to an increase of cooperation. The effect is the same independent
of the order of when costly sanctioning is included. This suggests that inclusion of costly sanctioning
increases cooperation, but that this effect is removed the moment costly sanctioning is not possible
anymore.

Figure 9.7: a) During the first six periods, subjects have the opportunity to punish the other group members.
Afterwards, the punishment opportunity is removed. b) During the first six periods, punishment of other group
members is ruled out. Afterwards, punishment is possible. (Source: Fehr and Gächter, 2002).
102 Self-governance in the laboratory

9.7 Communication and costly sanctioning
Why does communication and costly sanctioning increase cooperation? Why does costly sanctioning
increase cooperation but not the net earnings? These questions are part of several current research
efforts. You may find it surprising to learn that there are various possible explanations and, therefore,
there is no common agreement on the best possible explanation.
Costly sanctioning happens spontaneously in many of these experiments. People invest in
sanctioning others and this leads to a higher level of cooperation. But there is, well, a cost to
costly sanctioning! The net earnings are not higher than those compared to treatments without any
possibility to sanction. So why do people sanction in the first place? Interestingly, sanctioning
experiments in which people only get warnings with no monetary penalty also show an increase in
cooperation. Furthermore, if experiments last many rounds, we see net gains when using sanctions is
allowed. There is also an ongoing debate on the effectiveness of using penalties for not cooperating
compared to rewards to increase cooperation (the old adage of the carrot versus the stick). Including
the option to give other participants a reward for cooperative behavior, for example, also increases
cooperation.
What makes communication so effective? One explanation is that people start feeling like they
belong to a group. Instead of competing against each other as is seen in the no communication
rounds, communication seems to lead to a group effort. Others argue that communication leads to
further explanation of the experiment through group discussion (i.e., learning) so that everybody
understands what is best to do. Another explanation is that communication enables people to guess
how much they can trust others and confirm social norms. Experiments have been performed where
participants watch a video recording of a discussion on the best strategy, or a recording of a professor
explaining the possible strategies. In the end, mutual communication within the group was found to
be most effective.
Given the experimental work on governance in laboratory settings, at a minimum we can say
that the specific ways costly sanctioning and communication were implemented in the experiments
will have a significant impact on the results. Nevertheless, the basic notion that communication and
costly sanctioning have positive effects on governance is an important topic in subsequent chapters
in which we will analyze practical case studies on public goods and common-pool resources.

9.8 Critical reflections
When experiments are performed with common-pool resources or public goods in the lab or in the
field, the results with actual humans typically contrast significantly with the theoretical predictions
of what rational egoists would do. There is some cooperation, although we also see under-provision
and overharvesting in the experiments. Inclusion of communication and costly sanctioning increases
the level of cooperation.

9.9 Make yourself think
1. Come up with an example in which you used sanctioning?
2. Provide examples when communication alone might be sufficient to achieve cooperation (in
contrast with writing a contract).
9.10 References 103

9.10 References
Brosig, J., Ockenfels, A., & Weimann, J. (2003). The effect of communication media on cooperation.
German Economic Review, 4, 217–241.
Fehr, E., & Gächter, S. (2002). Altruistic punishment in humans. Nature, 415, 137–140.
Janssen, M., Holahan, R., Lee, A., & Ostrom, E. (2010). Lab experiments for the study of social-
ecological systems. Science, 328, 613–617.
Ostrom, E., Gardner, R., & Walker, J. (1994). Rules, games and common-pool resources. Ann Arbor:
University of Michigan Press.
Ostrom, E. (2005). Understanding institutional diversity. Princeton, NJ: Princeton University Press.
Part IV

Rules of the Games
Key Concepts
In this chapter we will:
– Learn which types of rules can be used to address different elements
that structure action situations.

10 — Classifying Rules

Organizing a surprise party

A surprise party can be considered an action situation where different participants have different
positions, information and choices. When you organize a surprise party there are different rules and
norms to take into account. These rules and norms are informal; they are not written down on paper.
Nonetheless, it is advisable to follow these rules for a successful event. Once a date has been set for
which the guest of honor is available, an assistant is appointed to create a fake activity. Guests are invited
and are explicitly told that it is a surprise party. All guests need to arrive about half an hour before the
arrival of the guest of honor and they need to park their cars out of sight.
The informal rules define the positions of the participants (guest of honor, organizer, assistant, and
guests), the boundary rules when people can participate in the event (needs to be invited), define the
information one has (do not share information with guest of honor!), define the choices (guest of honor
has limited choices). If these rules are followed the payoff is a successful surprise party.
This example shows that there are different types of rules that define specific aspects of the action
situation. In this chapter we will discuss what kinds of rules relate to the different variables of the action
situation.

10.1 Introduction
In this chapter we will discuss consistent ways to group rules. In empirical research many different
types of rules are found. Elinor Ostrom developed a way to classify the rules in order to study and
understand institutions.
The motivation for classifying rules relates to addressing questions such as “which rules need to
be changed to solve a particular problem?” or “which are the most cost effective rules to change
to achieve a certain outcome?” Such questions may arise in response to a vast array of problems
from global challenges such as how to change the incentives facing users of fossil fuel to reduce
CO2 emissions so as to reduce the impact of climate change, to local issues such as how to change
the parking regulations on a university campus to increase accessibility.
Policy analysts may want to fix a dysfunctional action situation as a doctor wants to heal a sick
patient, a mechanic wants to repair a car, or a computer technician wants to solve a telecommu-
nications problem. The policy maker needs to understand the workings of the action situation if
she wants to attempt to solve problems of poor performance. What is the actual problem leading
participants in an action situation not to cooperate to sustain the commons? Because action situations
can be extremely complex, policy makers (an to an even greater extent, politicians) can fall back on
ideology, which may lead to policy panaceas such as “cut taxes and everything else will take care of
108 Classifying Rules

itself.” Ideology might not be the best guidance to dissect the problems that societies face. Based on
a physician’s sophisticated understanding of the workings of the human body, she will perform a
number of diagnostic checks to identify the fundamental causes of the illness.
The policy analyst needs to understand the action situation and the consequences of different
interventions on the outcomes. This understanding is ideally based on experience and a good
theoretical framework. Instead of assuming all situations are the same and we just have to apply a
blue print solution, the policy maker needs to understand the local context to identify the appropriate
response. The classifications of rules as discussed in this chapter may help the policy analyst to
perform a proper diagnosis.

10.2 How to classify rules
In Chapter 4 we identified the components of action situations that are used to construct a wide variety
of analytical models of markets, hierarchies, firms, neighborhood associations, common-property
governance regimes, etc. The elements are participants, positions, actions, outcomes, information,
control, and costs/benefits. They are related in the following manner:

• Participants and actions are assigned to positions
• Outcomes are linked to actions
• Information is available about action-outcome linkages
• Control is exercised over action-outcome linkages
• Costs and benefits are assigned to action-outcome linkages

Participants, who can either be individuals or any example of a wide variety of organized entities,
are assigned to positions. In these positions, they choose among actions allowed for these positions
in light of the information available to them, the control they have over action-outcome linkages, and
the benefits and costs assigned to actions and outcomes.
For example, within a sport league, participants include players, coaches, teams, etc. Each of
those participants is permitted to undertake a limited number of actions. A coach decides which
players will play in the game, while the player is making operational decisions allowed by the rules
in split second intervals during the game. Information available in an action situation is not available
for everyone. Tactics of teams are not publicly shared, and a player may not reveal some shin splint
problems to his coach because she wants to play that night. Winning and losing games has major
consequences for teams. Will they win a key tournament and get promoted to a higher-level league?
Although teams and team members will benefit from winning the game, individual players also want
to shine to get more lucrative contracts.
The relationships among the various parts of the action situation are represented within the square
in Figure 10.1. Potential outcomes from a particular pattern of actions can be evaluated using criteria
like efficiency and equity. If an actor wants to influence the outcomes of this action situation, rules
may be crafted in order to do so. For example, if the action situation is a one-shot prisoners’ dilemma
without communication, an increase in cooperation could be achieved by allowing communication
between the players. If there is a desire to increase the efficiency of a university, more strict admission
standards can be crafted.
We are now in a position to start classifying rules. The classification of a rule is based on the
element in the action situation that is most directly impacted by that rule. Many rules also indirectly
affect other components of the action situation via the interactions among the components involved
10.2 How to classify rules 109

Figure 10.1: Rules as exogenous variables directly affecting the elements of an action situation (Adapted from
Ostrom 2005).

in the action situation.
We use the action situation to classify seven broad types of rules: position, boundary, choice,
aggregation, information, payoff and scope. Position rules create positions (e.g., member of a
legislature or a committee, voter, etc.). Boundary rules affect how individuals are assigned to or
leave positions and how one situation is linked to other situations. Choice rules affect the assignment
of particular action sets to positions. Aggregation rules affect the level of control that individual
participants exercise at a linkage within or across situations. Information rules affect the level of
information available in a situation about possible actions and the link between actions and outcomes.
Payoff rules affect the benefits and costs assigned to outcomes given the actions chosen. Scope rules
affect which outcomes must, must not, or may be affected within a given domain.
In Figure 10.1, arrows illustrate how different types of rules relate to specific parts of the action
situation. We can also look at the different types of rules from a language perspective. Which verbs
are most likely used to express the rule? In Table 10.1 we provide verbs that closely resemble the
intent of specific rules. For example, most boundary rules will specify the conditions for when
somebody can enter or leave to a position.

Type of Rule Basic AIM Verb Regulated component of
the action situation
Position Be Positions
Boundary Enter of leave Participants
Choice Do Actions
Aggregation Jointly affect Control
Information Send or receive Information
Payoff Pay or receive Costs/Benefits
Scope Occur Outcomes

Table 10.1: The AIM component of each type of rule
110 Classifying Rules

Specific rules can have an impact on the whole action situation. Take for example the local
medallion ordinance for taxis in urban areas. A boundary rule might be that only taxis who display
a purchased medallion from the city can legally use the city streets to attract customers. This has
been the case in New York City for decades. When a city limits the number of medallions it will
authorize, the entry costs for putting a taxi on the streets rise substantially if demand is rising. But if
the medallions are strictly limited, the potential returns per unit of time also increase substantially,
and the time the traveler has to wait for a taxi may also increase.
Around 2000, the Dutch government changed the boundary rules for taxis in the Netherlands.
Until that time, there were strict boundary rules for cities. The minister of Economic Affairs argued
that it would be more efficient to open the market and let taxis freely attract passengers within the
whole country. It was expected that competition between taxis would increase performance and
reduces prices. However, taxis flocked to Amsterdam, the city with the most potential clients, and the
original taxi company acted to defend their territory. The so-called taxi war led to taxi drivers fighting
physically for passengers and some taxis were burned in the process. Further, many taxi drivers
without much knowledge of Amsterdam city streets ended up driving tourists, who also did not know
their way around. After ten years, the government concluded that the policy change (boundary rule
change) led to higher prices and reduced the quality of the taxi service. Could the government have
used a policy analysis framework like the IAD to foresee this unintended consequence?

10.3 The seven types of rules in detail
10.3.1 Position rules
As discussed in Chapter 3, an important component of an action situation is the set of positions that
participants can occupy. Position rules create these positions. Since various rules relate to actions
that participants in certain positions may, must or must not make, these position rules have important
consequences. The position rule creates the scaffolding for a wide diversity of situations that can
vary significantly in terms of the number of positions and the relative authority assigned to each
position in a given action situation.
A position rule could be defined such that all people in the action situation occupy the same
position. Other position rules define more specific positions, such as the chair person or referee.
A position rule may also state whether there is a limit to the number of participants that can
occupy a certain position. Most sports have rules regarding how many players of a team are allowed
to be on the field. For example, soccer teams have a maximum of eleven field players and a minimum
of eight field players. For a senate to make decisions, a certain quorum needs to be met, meaning that
a minimum number of senators needs to be present in order to make binding decisions. Sometimes
the maximum number of participants in a certain position is restricted by the biophysical conditions.
For example, the maximum number of students who can take a course is often defined by the official
number of seats available in a classroom.

10.3.2 Boundary rules
Boundary rules are often called entry and exit rules. They define (1) who is eligible to enter a position,
(2) the process that determines which eligible participant may enter (or must enter) positions, and
(3) how an individual may leave (or must leave) a position. For example, during a soccer game a
coach can bring in a new player if an existing player is taken off the field. Depending on the sport,
there are specific rules regarding how many switches a team is allowed to make during a match.
10.3 The seven types of rules in detail 111

Thus, some entry rules specify the criteria to be used to determine whether an actor is eligible to
fill a particular position. Ascribed and acquired attributes are frequently used in this type of entry
rule. Individuals may have to meet certain physical standards, such as height and weight. Think
about boxing (weight) or the military (weight and height). They may have to meet a certain wealth
standard or pay an entry fee (a golf country club). Individuals may be required to possess a certain
range of experiences, to be above a minimum age, to have graduated from certain schools, to be
the descendants of a particular group, to possess certain abilities, or to live in certain geographical
areas. Public employment under civil service systems and patronage systems differs substantially
with respect to the entry rules that are applicable.
When those crafting a rule hope to increase the skills and knowledge held by those in a position,
they may list one or more acquired characteristics, such as holding a college degree or passing a
test. For example, to become a tenure-track professor at most public universities within the U.S. one
is required to have acquired a relevant PhD degree. Contentious debates have and will continue to
erupt over whether one or another attribute should be included in a rule. For example, in the past,
race was used to deny access to many public places for black citizens within the U.S.
Sometimes boundary rules are not specific about the attributes of individuals who may occupy
positions. For example, boundary rule for a party might simply be “those invited may attend.”
Boundary rules do not always include a choice. A suspect that has been arrested by the police has no
choice as to whether to participate in court proceedings concerning charges made by the state (i.e.,
be in the position of defendant). The suspect also has no choice in exiting his/her position. Other
examples of compulsory positions taken by participants are those who are drafted into the army,
serve on a jury or pay taxes.

10.3.3 Choice rules
Choice rules specify what a participant occupying a position must, must not, or may do at a particular
point in a decision process in light of conditions that have, or have not been met at that point in the
process. The actions that participants must, must not, or may do are dependent both on the positions
they hold, prior actions taken by others and/or themselves and attributes of relevant state variables.
A suspect in a criminal case has limited choices: what to plea, whether to hire a lawyer, whether
or not to testify on his or her own behalf. A parent has many choices to make and often needs to
make choices for their children, especially if they are under 18 years of age.
The choices available also depend on the conditions of particular actions. If a carpenter injures
his hand while hammering a nail on the job, the carpenter may be elegible for compensation because
of the arrangements related to activities of the workplace. If the carpenter gets injured at home,
outside of the work situation, no such compensation may be received. The carpenter may be a
member of a club that provides members with disability insurance, but that choice is related not to
the position of “employee” but, rather, to the position of “member of group insurance plan.”
In many bureaucratic action situations, no one participant is authorized to take particular positive
actions unless specific conditions are met. A power-plant employee, for example, may not be
authorized to open a turbine unless water levels are above a minimum. A social worker cannot
authorize food stamps or welfare payments unless an applicant’s income is below some defined level
given the size of the family and other conditions.
Choice rules affect the basic rights, duties, and liberties of individuals involved in an action
situation. Choice rules may be allocated to positions with high levels of control, such as the President
of the United States.
112 Classifying Rules

10.3.4 Aggregation rules
Aggregation rules determine whether a decision, made by a single participant or multiple participants,
is needed prior to an action at a node in a decision process. In chess each participant is authorized to
make a move when it is his or her turn. The player’s action set at that juncture includes the specific
physical moves to be made. While no single player fully controls the final outcome, individual
players do control the decisions to be made at individual decision moments.
However, in many legislative action situations, multiple participants jointly control which actions
will be taken. Members of the legislature may amend or not amend a bill, and can cast votes that
are then aggregated by an aggregation rule. No single participant has full control over the move to
amend or not to amend the bill.
Aggregation rules are necessary whenever choice rules assign multiple positions partial control.
Aggregation rules need to clarify for a group “who is to decide.” “A new president is elected if one
candidate gains 50% of the votes plus one vote,” is an example of an aggregation rule. This is the
majority rule. Another example is the unanimity rule, where all participants in an action situation
need to agree with a proposed action in order to continue with that action (e.g., in a trial by jury, the
jurors must agree unanimously on the verdict). Another example is that one actor can block actions
using a veto. For example, some members of the United Nations Security Council can veto proposed
resolutions.
Sometimes different participants have complementary attributes. A bank clerk, for example, is
not authorized to open safety deposit boxes unless the owner of the box or an authorized agent signs
a registration form and produces a second key to fit the box.

10.3.5 Information rules
An important part of any action situation is the informa-
tion available to participants about the overall structure
of that situation, the current state of the attributes of the
participants, the previous and current decisions of other
participants in related positions, and their own past deci-
sions. Information rules affect the level of information
available to participants. Information rules authorize chan- Figure 10.2: Sellers must earn the designa-
nels of information flow among participants, assign the tion of top-rated seller based on customer
obligation, permission, or prohibition to communicate to feedback.
participants in positions in particular situations, and the
language and form in which communication must take place. Information rules are particularly
important in generating information about past actions of participants so that other participants can
know who is, or is not, trustworthy.
On eBay, for example, you can see feedback about past transactions of the buyers and sellers that
enable you to make decisions whether to pursue the transaction or not (Figure 10.2). Each state and
country has rules on which languages can be used to write official documents. To board a plane, your
boarding pass must have your legal name exactly as it appears in your passport or drivers license.
How open is the U.S. federal government about how decisions are made? When are minutes
of meetings made available? Can the general public observe a court case or meeting? What sort of
actions must someone who is on probation report to a probation officer on a regular basis?
10.3 The seven types of rules in detail 113

Figure 10.3: The US government service (GS) pay-scale. There are very specific rules to enter and leave the
different positions (GS-1, GS-2), etc. Within each GS position there are 10 different salaries (e.g, a GS-8, step
5 salary is $42,647).

10.3.6 Payoff rules
Payoff rules assign external rewards or sanctions (fines) to particular actions that have been taken
by participants or to the assessment of particular outcomes. An example of a set of payoff rules
is the pay schedule that is used by a government agency or by a private firm to assign salaries
to participants in particular positions (Figure 10.3). Payoff schedules may vary in terms of the
variables taken into account and the complexity of the schedule. Hourly wage payoff rules are very
simple. Someone being paid according to piecework will, on the other hand, be paid by formula.
Performance contracts for corporate exectutives are frequently much more complex and state very
explicity when executives qualify for certain kind of bonuses and how they will be paid (in terms of
stocks, options, cash, etc.).

10.3.7 Scope rules
Scope rules affect a known outcome variable that must, must not, or may be affected as a result of
actions taken within the situation. For example, for a professor to get tenure, they are evaluated
on different outcome measures (quantity and quality of publications and quality of teaching). The
performance of schools can be evaluated based on many different outcomes as witnessed by the
various criteria used for college rankings. Most sports are straight forward in the outcome evaluation,
one wins a competition or not. Some sports are based on evaluations by a jury, such as gymnastics,
who follow a detailed set of rules to define the score (Figure 10.4). An example of the application
of such a scope rule would be as follows: in order for a female competitor to win a sanctioned
gymnastics competition (a particular outcome of an action situation), the competitor must achieve
the highest combined score on all events (the floor exercise, vault, uneven bars, and balance beam).
114 Classifying Rules

Equating achieving the highest combined score on all events with “win” is a scope rule that defines
how actions lead to outcomes. Another set of detailed scope rules defines how judges assign points to
the actions of participants. Note that both gymnists and judges hold positions in the action situation.
What are the boundary rules for them to occupy those positions?
Another example of a scope rule used to manage fishing by Native Americans on the Northwest
Pacific coast is that tribe members cannot begin fishing for salmon until the chief has caught and
dried the first salmon caught that season.

10.4 Default conditions
Not all possible actions and outcomes of each action situation have
rules associated with them. If there were, following the rules would
become such a complex task that we would waste all of our time on
monitoring and sanctioning the rules and would get nothing done.
This raises the quesiton of what are the default situations when there
are no explicit rules and what is allowed in those circumstances?
To answer this question, we can start defining the default condi-
tions from what is physically possible. Table 10.2 below defines
the default conditions for each of the rule types if there are no ex-
plicit rules. If the participants adopt internal norms that promote
cooperative behavior, such a default position may lead to a positive
outcome. However, if some participants cheat, manipulate, or do
not cooperate, it becomes clear that explicit rules would be needed
to reach a positive outcome in the action situation. Figure 10.4: Is this a “good”
How do people create rules when they experience a default sit- stag leap? Scope rules deter-
uation that produces undesirable outcomes? A possible action is mine criteria so judges can an-
to create the position of a leader or a group of leaders who coordi- swer this question consistently.
nate communication and mediate differences in interests among the
participants.

Default positions
Default Position Conditions One position exists.
Default Boundary Condition Anyone can hold this position.
Default Choice Condition Each player can take any physically possible action.
Default Aggregation Condition Players act independently. Physical relationships
present in the situation determine the aggregation
of individual moves into outcomes.
Default Information Condition Each player can communicate any information via
any channel available to the player.
Default Payoff Condition Any player can retain payoffs from any outcome that
the player can physically obtain and defend.
Default Scope Condition Each player can affect any state of the world that is
physically possible.

Table 10.2: Default positions of each type of rule.
10.5 Critical reflections 115

10.5 Critical reflections
Within action situations we can distinguish different types of rules that define who is part of the
action situation, the position she can take, the information that is available to her, and what the
payoffs are for different decisions. Defining the different types of rules that provide the “software”
for action situations is important if we wish to analyze them in more detail.

10.6 Make yourself think
1. What kind of rules do you have within your household or student dorm? Are there position
rules? Choice rules? Boundary rules? Scope rules?
2. What is an example of information rule that holds for students at ASU?

10.7 References
Ostrom, E. (2005). Understanding institutional diversity. Princeton, NJ: Princeton University Press.
Key Concepts
In this chapter we will:
– Learn to analyze the text of an institutional arrangement
– Be introduced to ADICO
– Recognize the differences between strategies, norms, and rules

11 — Rules, Norms and Shared Strategies

Consequences

What are the consequences if you don’t follow the rules? For example, driving far above the speed
limit might have consequences if you are caught. What about your civic duty to vote in elections? What
will happen if you don’t vote? In the U.S. you don’t have to vote, and there are not formal consequences
if you don’t vote. In a country like Belgium you MUST vote and there are penalties for not voting
without a valid reason.
Not all statements that say what you MUST do and MUST NOT do are rules. In the statement in
Figure 11.1, there are no consequences specified if you do take pictures. Of course it would be very
impolite not to follow the request, but is the exhibit organizer allowed to throw you out of the exhibit and
confiscate your camera?
In the immigration warning in Figure 11.2 Malaysia states clearly that people who are dressed in
shabby clothes will be ordered to leave the country. Hence you MUST dress properly OR ELSE you will
be ordered to leave the country.

Figure 11.1: Statement when entering exhibition.
118 Rules, Norms and Shared Strategies

Figure 11.2: Warning on entering Malaysia.

11.1 Introduction
Throughout this book we have used the term “rules,” but we have not been very precise in what we
mean by it. In fact, scholars have different interpretations of what makes something a rule rather
than a shared strategy or a norm. In this chapter we will discuss in more detail how we define shared
strategies, norms and rules. Moreover, we will learn some tools for looking at shared strategies,
norms and rules in a more systematic way. This will help you to analyze action situations and unravel
what the rules and norms are. In fact, a better understanding of the structure of rules will help you
with decisions you make in your life.
We will look at written statements about what is allowed or not. If you look carefully around
you, you will find many signs and statements that define what is or is not allowed: “only people with
a valid permit can park here,” “you must show your ID to buy alcohol in this store,” “please recycle,”
etc. In this chapter, we will learn how to analyze these statements in more detail to define what they
mean in the context of institutional statements.
An institutional statement describes opportunities and constraints that create expectations about
other actors’ behavior: what may, must and must not be done. In other words, an “institutional
statement” encompasses a broad set of shared linguistic constraints and opportunities that prescribe,
permit, or advise actions or outcomes for participants in an action situation. The importance of these
statements cannot be underestimated for the smooth functioning of human societies. Just imagine if
you could not predict at all what others are likely to do in the action arenas in which you participate
each day. Not only would you not get much done, you might end up seriously injured or even dead.
The institutional grammar tool, developed by Sue Crawford and Elinor Ostrom, allows analysts
to distinguish more systematically between institutional statements that are best understood as shared
strategies, norms or rules. Rules can be written down on paper, such as formal laws, but might not
11.1 Introduction 119

be common knowledge to everyone. Do you know all the laws that apply in the jurisdiction that
you now inhabit? Rules do not have to be written down on paper. In fact, many rules are actually
unwritten, so-called rules-in-use. These are the rules that people actually use in an action situation.
Let’s provide some examples of institutional statements:

1. In order for a couple to marry in the state of Arizona they must be of the opposite sex or else
no marriage license will be issued. Is this the same in every state?
2. A customer must produce a valid I.D. if asked by an establishment that sells alcoholic beverages
in order to be served alcohol. An underage person who tries to buy liquor with a fake I.D. may
be charged with a Class 3 misdemeanor.
3. All lobster fishers in Maine must not put their traps in the fishing territory of neighboring
harbor, or else their traps may be cut loose.
4. If you use the microwave, you must clean up your own mess!
5. The person who places a phone call is the one who calls back when the call gets disconnected.

If we look carefully at these institutional statements we can detect regularities. The statements
are about certain types of persons who are or are not allowed to do certain activities. And if they do
something that is not allowed there might be consequences.
Let us now turn to the syntax that we will use to analyze these and other examples. With a syntax
we mean that we can detect common components or building blocks in these statements, and these
components have a particular function.
The general syntax of this grammar includes five components and is known as ADICO: AT-
TRIBUTE, DEONTIC, AIM, CONDITIONS, and OR ELSE. You might compare this to the SUB-
JECT, VERB, OBJECT, or “SVO” structure of English.
ATTRIBUTES (A) is a place holder for any value of a participant-level variable that distinguishes
to whom the institutional statement applies. Examples include eighteen-year-old, female, college
educated persons, or persons with one year of experience; or a specific position, such as employee or
chairperson.
DEONTIC (D) is a place holder for “may” (permitted), “must” (obliged), and “must not” (forbidden).
AIM (I) is a place holder that describes particular actions or outcomes in the action situation to
which the deontic is assigned. An AIM may include a formula specifying an amount or intensity
of actions or ranges of outcomes or a description of a process for an action. An example would be
harvesting 10 kilograms per day.
CONDITIONS (C) is a place holder for those variables that define when and where an action or
outcome is permitted, obligatory, or forbidden. Examples include between 6 am and 9 am, in area B,
after the first fish is dried, etc.
OR ELSE (O) is a place holder for the assigned consequence for not following a rule.
We will now define the three types of institutional statements using the syntax above. In the same
way that a sentence does not need to contain all the possible parts of speech, institutional statements
need not contain all the building blocks we have described. The possible institutional statements are:

• Shared strategies contain components AIC; the attributes, aim and conditions. Strategies can
thus be written as [ATTRIBUTES] [AIM] [CONDITIONS].
• Norms contain the elements ADIC. Compared to shared strategies, norms include whether
120 Rules, Norms and Shared Strategies

an activity is permitted, obliged or forbidden. But norms do not explicitly include the
consequences for doing something that is forbidden or not doing something that is required.
Norms can thus be written as [ATTRIBUTES] [DEONTIC] [AIM] [CONDITIONS];
• Rules do contain all five elements (ADICO). Regardless of how institutional statements are
expressed in natural language, they can be rewritten or summarized in the ADICO format:
[ATTRIBUTES] [DEONTIC] [AIM] [CONDITIONS] [OR ELSE].

By writing the statements in a consistent manner, we can then better compare the institutional
statements in use in a variety of settings.
Let’s now apply the ADICO syntax to the five examples above.

In order for a couple to marry in the state of Arizona they must be of the opposite sex or else no
marriage license will be issued.

ATTRIBUTES: [a couple who wants to marry and have their marriage be legally
recognized by the state of Arizona]
DEONTIC: [must]
AIM: [be of opposite sex]
CONDITIONS: [in the state of Arizona]
OR ELSE: [no marriage license will be issued]

A customer must produce a valid I.D. if asked by an establishment that sells alcoholic beverages
in order to be served alcohol. An underage person who tries to buy liquor with a fake I.D. may be
charged with a Class 3 misdemeanor.

ATTRIBUTES: [a customer]
DEONTIC: [must]
AIM: [produce a valid I.D.]
CONDITIONS: [if asked by the establishment in order to be served alcohol]
OR ELSE: [no liquor will be served][may be charged with a Class 3 misdemeanor]

All lobster fishers in Maine must not put their traps in the fishing territory of neighboring harbor,
or else their traps may be cut loose.

ATTRIBUTES: [All lobster fishers in Maine]
DEONTIC: [must not]
AIM: [put their traps in the fishing territory of neighboring harbor]
CONDITIONS: [-]
OR ELSE: [traps may be cut loose]
11.2 The syntax components 121

If you use the microwave, you must clean up your own mess!

ATTRIBUTES: [If you use the microwave]
DEONTIC: [must]
AIM: [clean up your own mess]
CONDITIONS: [-]
OR ELSE: [-]

The person who places a phone call is the one who calls back when the call gets disconnected.

ATTRIBUTES: [The person who places a phone call]
DEONTIC: [-]
AIM: [calls back]
CONDITIONS: [when the call gets disconnected.]
OR ELSE: [-]

11.2 The syntax components
We now turn to a discussion of the components of the ADICO syntax. Understanding the components
helps analysts to develop common methods of distinguishing between rules, norms, and shared
strategies to cumulate knowledge for key questions such as: “what is the difference between a rule
and a norm?” and “what is the difference between a shared strategy and a norm?” We will see that
there will still remain some ambiguity in the definitions, but it will be more precise as the terms are
commonly used. Especially the distinction between a norm and a rule has some gray areas, which
we will touch upon when we discuss the OR ELSE component.

11.2.1 ATTRIBUTES
All institutional statements apply to a subset of participants in an action situation. The subset can
range from one participant to all participants. A set of ATTRIBUTES define the participants to
whom the institutional statement apply. If individuals make up the participants in an action situation,
the ATTRIBUTES will be individual-level values. Individual level ATTRIBUTES include values
assigned to variables such as age, residence, sex, citizenship, and position. When the participants
governed by a set of institutions are corporate actors, rather than individuals, the ATTRIBUTES
refer to organizational variables such as size of membership or geographic location.
In the first example, no specific ATTRIBUTES are specified, but in the long list of marriage
regulations there are additional attributes related to age, citizenship and blood relationships. For
example, in the state of Arizona people below the age of 16 are allowed to marry if they have
the notarized consent of their parents or legal guardian as well as a court order. Some version of
the fourth example, the microwave cleanup statement, can often be found taped to the door of a
community microwave. The beginning of that statement, “If you use the microwave,” could be
interpreted as the ATTRIBUTE of “microwave users.” In the last example, the ATTRIBUTE is the
caller who placed the call. The other examples list no specific attribute. When no specific attribute is
listed, the default value for the ATTRIBUTE component is all members of the group. This means
that the ATTRIBUTE component always has something in it, even when a specific attribute is not
122 Rules, Norms and Shared Strategies

contained in the statement. Thus, the second example applies to all people who want to buy alcohol,
and the third example apples to all fishers in Maine.
The ATTRIBUTE component can specify which positions in an action situation are referred to
in an institutional statement, such as “physicians must maintain confidentiality with regard to patient
information.”

11.2.2 DEONTIC
The DEONTIC component draws on deontic logic to distinguish prescriptive from non-prescriptive
statements. The complete set of DEONTIC operators consists of “permitted,” “obliged” and “forbid-
den.”
Institutional statements use the operative phrases may, must/should, and must not/ should not to
assign these operators to actions and outcomes. “Should” and “must” are both commonly used to
oblige a person to act. Similarly, “must not” and “should not” both forbid. For the sake of simplicity,
we use “must” and “must not” throughout this chapter in nearly all examples. However, the deontic
terms can be used equally well for “should” statements. Generally, in everyday language, “must”
obliges someone more strongly than “should,” and “must not” forbids someone more strongly than
“should not.”
One may wonder whether the element “may” is that important. What does it mean that certain
actions are allowed? In the United States, consumers above the age of twenty-one may purchase
alcohol in most states. In some states, Sunday liquor laws constrain this permission further and
allow consumers above the age of twenty-one to purchase alcohol only Monday through Saturday.
In effect, these rules establish the settings in which permission exists and thus forbid the action in
circumstances that do not meet the stated CONDITIONS. A rule that grants permission to cut trees
with a permit from a forestry agency implies that in the absence of a permit tree cutting is forbidden.
If we look at the examples, we see that in examples 1, 2, 3, and 4 participants must or must not
do certain activities. The fifth example has no deontic element, which means that example 5 is a
strategy. Both norms and rules have deonitic elements.

11.2.3 AIM
The AIM is the specific description of a working part in an action situation to which an institutional
statement refers. The description can include information about a process (filling out a form at the
U.S. Post Office) or a formula (pay $10 per hour worked). In order for an institutional statement
to influence behavior, the AIM must be physically possible and the contradiction need also to be
physically possible. An individual cannot logically be required to undertake a physically impossible
action and prescribing an action can only influence behavior if it is physically possible to not do that
action. The capability of voting implies the capability of not voting. Voting for candidate A implies
the option of not voting for candidate A. The AIM sometimes specifies states of affairs in the world
or an outcome instead of an action. Outcomes, like actions, must also be possible and avoidable
to be parts of a well-formed institutional statement. Moreover, any particular outcome implies the
existence of the contradicting outcome.
In example (1), the AIM is the sex of two people who want to get married. Same-sex marriage
is not allowed in the state of Arizona, and couples of the same sex therefore do not have the right
to be married. In the second example, the AIM is the action of producing a valid I.D. The AIM in
the third example is to put traps at the right place. The AIM in the fourth example is cleaning the
microwave. Finally, the AIM in the fifth example, the strategy, is the action of calling back.
11.2 The syntax components 123

The AIM often supplies the focus for studies and debates such as the same sex marriage debate.

11.2.4 CONDITIONS
CONDITIONS indicate the set of variables that define when and where an institutional statement
applies. For example, the CONDITIONS for a statement might indicate when a statement applies,
such as during certain weather conditions, at a set time, or at particular step in some process. Likewise,
the CONDITIONS might indicate where a statement applies, such as a particular jurisdictional
area. If an institutional statement does not specify particular variables, the default value for the
CONDITION is in all situations at all times and in all places covered by that rule, norm, or strategy.
Thus like the ATTRIBUTE, the CONDITION component always has some value in it even when the
institutional statement fails to overtly specify it.
The CONDITIONS component in the first example refers to the state of Arizona, since the
institutional statement refers to a state specific regulation. Marriage among same-sex couples is
allowed in various other states in the U.S. The CONDITIONS component in the second example
indicates that a valid I.D. needs to be provided when the establishment where liquor is bought
requests this. The strategy in the fifth example applies when a telephone call is disconnected. The
third and fourth examples do not specify specific CONDITIONS; therefore, we assume that the rules
apply for Maine fishers and microwave users under all circumstances.

11.2.5 OR ELSE
The final component of the institutional syntax is the consequence that an institutional statement
assigns to detected noncompliance with the other components of the statement. In some cases,
the OR ELSE specifies a range of possible sanctions if a rule is not followed. Individuals in the
community know that if they violate a rule, they face the probability that a sanction in a specified
range will be applied and that others in a similar situation face the same range. Only rules include an
OR ELSE. This component, consequently, plays a crucial role in discerning what a rule is and how
rules differ from other institutional statements.
Description of sanctions for breaking a rule are a common type of OR ELSE statement, but the
OR ELSE may take other forms. The OR ELSE might also shift the DEONTIC assigned to some
other action. For example, a violator might be forbidden to vote or engage in some other action
that would otherwise be permitted. The OR ELSE might also shift the DEONTIC assigned to some
activity for an individual from permitted to obligatory. The violator might be required to allocate
resources to a public jurisdiction (i.e., pay a fine), or another actor might be required to check on the
violator. Those actions might be permitted under all other CONDITIONS, but obligatory when the
CONDITION of a violation of the rule is met.
Although the OR ELSE often refers to physical punishments, the OR ELSE may also in-
volve institutional actions, such as taking away a position or refusing to accept an amendment.
For example, one of the rules governing the amendment process may state that legislators with
[ATTRIBUTES][must][take a particular action][when voting for an amendment][OR ELSE—the
amendment fails].
Three qualifications must be met for an OR ELSE to exist. First, the consequence stated in
the OR ELSE must be the result of collective action. A collective decision must have been made
in a relevant collective-choice arena to determine the consequence. Hence individual revenge is
not a valid component of an OR ELSE statement. Second, the threat in the OR ELSE component
of a rule must be backed by another rule or norm that changes the DEONTIC assigned to some
124 Rules, Norms and Shared Strategies

AIM, for at least one actor, under the CONDITION that individuals fail to follow the rule. Often
the actions threatened in the OR ELSE are forbidden under most CONDITIONS (e.g., imposing
a fine, incarcerating a citizen, or taking someone’s livestock and putting them in a village pen).
The prescription backing the OR ELSE makes these actions permitted or required under in the
CONDITION that someone breaks a rule.
Third, in order for an OR ELSE to exist, a prescription must affect the constraints and opportuni-
ties facing an actor or actors with the responsibility of monitoring the conformance of others. The
actors who monitor frequently sanctioned nonconforming actors may only report nonconformance
to someone else responsible for sanctioning. We do not consider government sponsorship to be a
necessary condition for a statement to include an OR ELSE. Many self-organized, communal, or
private organizations develop rules that include (1) sanctions, (2) another rule or norm that changes
the DEONTIC assigned to some AIM for at least one actor if that individual fails to follow the rule,
and (3) a norm or rule (a sanctioning prescription) that affects the constraints and opportunities
facing an actor or actors who take the responsibility to monitor the conformance of others to the
prescription (a monitoring prescription).
Turning again to the example of institutional statements listed above, the first three examples
appear to contain an OR ELSE. Of course, we would want to check to be sure that there are rules
or norms regarding monitoring and sanctioning that back the stated OR ELSE. For example, the
potential punishment for fishers who put traps in a different territory is an OR ELSE only when rules
or norms accepted in that harbor prescribe others to monitor and to employ the sanctions defined
in the OR ELSE. Without the establishment of positions with the authority for monitoring and
sanctioning, phrases that contain the words “or else” fail to constitute an OR ELSE that distinguishes
an institutional statement as a rule as defined herein.
Institutional statements with content in the OR ELSE slot, then, are institutional statements
that add information to the action situation about what will happen if a participant violates the
prescription. The rule on producing a valid I.D. indicates that participants who violate this know
that if he or she violates the rule they will face the consequence and cannot buy alcohol. In the
first institutional statement, same-sex couples do not get a marriage license as a consequence of not
meeting the CONDITIONS.
Compare the information about the consequences of the marriage rule to that in the microwave
example. The sign on the microwave does not provide any specific information about what will
happen to a participant who chooses to ignore the norm and leave a puddle of soup at the bottom of
the microwave. Office workers seeing the sign on the microwave have only information about their
own internal costs and benefits of following the norm and their beliefs about how others in the office
are likely to respond when they open the door and see the crusty soup spill there.
The OR ELSE component of a rule is frequently linked to a CONDITIONS component that
specifies the number of times that a rule has been violated. The range of sanctions is likely to
be lowest when someone has committed their first offense against a specific rule. A different
rule, specifying a more stringent range of sanctions, will then be applicable if that individual has
committed a second, or a third, or more offenses. Many rules against drunk driving use this form
of graduated sanctions and increase the sanction for this offense substantially for second and third
offenders.
11.3 How to use the grammar of institutions? 125

11.3 How to use the grammar of institutions?
In the coming chapters we will discuss various examples of how to use the grammar of institutions
to analyze concrete cases. You, like Kyle in South Park (http://www.southparkstudios.com/full-
episodes/s15e01-humancentipad), may not be used to reading the terms and conditions of new
purchases but it is probably a good idea to do so when the costs of not reading them may be high.
At a minimum reading terms and conditions is a good
exercise from the perspective of learning to use the gram-
mar of institutions. What attributes of the consumer are
specified, what are the conditions and the aims. What are
the consequences, if any, if the conditions are not met?
When you look at concrete examples of norms and
rules you experience every day, you probably will find
that there are no explicit consequences. Does that mean
that these statements are always norms? Absolutely not.
Many statements are imbedded in regulations at higher or-
ganizational levels. So, many rules add special conditions
to already existing collections of rules. Not all rules that Figure 11.3: South Park
are involved in taking a course are explicitly stated in the
syllabus. There are many rules at the college or university
level that hold for many courses and which are not explicitly stated in the syllabus that are implied by
the fact that students are taking a class. Furthermore, universities need to meet regulations imposed
by the state and the federal government. These obligations impact the syllabus, but are not stated. If
all relevant rules were stated for a class, the syllabus would be 50 pages long. In fact, the syllabus
mainly focuses on the specific rules related to the specific course. Thus, to be fully aware of the
structure of an action situation, participants may need to be aware of several different collections of
rules and how they are nested.

11.4 Critical reflections
In this chapter you learned about a systematic way to analyze statements that can be classified as
strategies, norms and rules. Strategies do not include a specification that some behavior may, must, or
must not occur. The difference between rules and norms is the explicit statement of the consequences
if a rule is not met so that a third party can enforce the rules.

11.5 Make yourself think
1. Look at the terms and conditions of a recent purchase. What are the rules and norms? What
are the consequences if a rule is not followed?
2. What are rules and norms in the syllabus of the course you are following?
3. Are you aware of all the rules that might apply in a given action arena you are in frequently?
Under what circumstances would it be reasonable to not know the “letter of the law” governing
the situations you may be in often?
126 Rules, Norms and Shared Strategies

11.6 References
Crawford, S., & Ostrom, E. (1995). A grammar of institutions. American Political Science Review,
89(3), 582–600.
Ostrom, E. (2005). Understanding institutional diversity. Princeton, NJ: Princeton University Press.
Part V

Applying the Framework
Key Concepts
In this chapter we will:
– Be introduced to public health as a collective action problem
– Learn that vaccination is only effective if enough people
get vaccinated and is thus a public good provision problem
– Become aware that millions of people die each year due to lack of sanitation
– Notice the difference between opt-in and opt-out policies for organ donation

12 — Public Health and Collective Action

Bareheaded motorcyclist dies in helmet protest

On July 2, 2011, 55-year-old Philip Contos died after a motorcycle crash near McClary Road and
Route 11 in LaFayette, New York. Contos was part of a protest against motorcycle helmets. Police say
Contos suddenly hit the brakes and lost control of his motorcycle. Contos was not wearing his helmet,
as required by law, and doctors say he would have survived if he had been wearing a Department of
Transportation-approved helmet.
Most states in the U.S. require motorcyclists to wear a helmet, and about half the states require this
of bicyclists as well. Not wearing a helmet not only puts the cyclist in danger, it can also affect other
people involved in the accident. Think about a car driver who is involved in a deadly accident with a
motorcyclist.
Some motorcyclists, like Philip Contos, argue that it is a restriction of the freedom of the individual
to require helmets. In a federal court case in 1972, the State of Massachusetts told a motorcyclist who
objected to the law:
Provision of this section requiring motorcyclists to wear protective headgear is not violative
of due process, notwithstanding claim that police power does not extend to overcoming
right of an individual to incur risks that involve only himself, since public has an interest
in minimizing resources directly involved, in that from moment of injury, society picks
the person up off the highway, delivers him to a municipal hospital and municipal doctors,
provides him with unemployment compensation if, after recovery, he cannot replace his
lost job, and, if injury causes permanent disability, assumes responsibility for his and
his family’s continued subsistence. Simon v. Sargent, D.C.Mass.1972, 346 F.Supp. 277,
affirmed 93 S.Ct. 463, 409 U.S. 1020, 34 L.Ed.2d 312.
The decision was approved by the U.S. Supreme Court. This is a quintessential public goods problem,
and the motorcyclist who does not wear a helmet can be seen as a free-rider (no pun intended).

12.1 Introduction
Both the ways in which we live and the places in which we live affect our health. Diseases spread
from person to person. Unsafe situations created by someone may harm others. Just being sick and
going into public places increases the risk that others will also become sick. Likewise, when one
person takes risks while driving—for example speeding because they are late—they increase the risk
others will get injured in an accident. Both of these are examples of negative externalities related to
health. Thus, we all have the potential to generate health problems for others. On the other side of
the equation are those people that take care of our health problems. After a moment’s reflection, a
130 Public Health and Collective Action

reader familiar with the material presented in earlier chapters should see that public health involves
numerous social dilemmas. In fact, we live in a society with a very complex health system, which
has largely evolved to cope with these social dilemmas.
First, and perhaps most obvious, the public health care system can be viewed from a public good
perspective. If you get in an accident you will be taken to an emergency room that is financed by
both public and private means. The care you receive will be paid for either by an insurance company,
the government, you, or a combination of the three. Few individuals could afford to pay their health
care costs by themselves in exactly the same way that no single individual can build an irrigation
system by themselves. Thus, in the same way that people work together in groups to build water
infrastructure, typically we rely on group insurance plans provided by our employers to provide
health infrastructure. Such plans are club goods (which functions the same as public goods but
control for free-riding). Those who are retired rely often on publicly provided health care such as
Medicare in the U.S. But is Medicare a pure public good (remember we have discussed the rarity of
“pure” goods of any type)? No—Medicare is also a club good where club membership is defined by
being a U.S. citizen and of a certain age. Thus, given these different modes of health care provision,
not all health care is available for everybody. In fact, as with Medicare, there are clearly defined
boundary rules as to who receives which type of health care.
How health care is organized differs between countries (Table 12.1). The expenditure on health
as a percentage of Gross Domestic Product is more than 15% in the United States, which is high
compared to other countries. As of 2012, one out of every six people in the U.S. is uninsured,
meaning that this person will not share the cost of health care with others. Most other developed
countries have so-called socialized health care systems in which each citizen receives health care
as a basic right. The health-related expenditures in these countries with universal health care (i.e.,
Canada, Germany, and France) are lower than the U.S. In developing countries, the expenditure
levels are much lower than in the U.S., but the consequences are also measurable in terms of lower
life expectancies at birth.
Note that some of the basic public infrastructure investments that increase health outcomes many
not be directly related to health care. Examples include the provision of safe drinking water and
improved sanitation to reduce the transmission of diseases. Provision of such infrastructure is a
classic public good problem.
Table 12.1 indicates that the U.S. health care system is very expensive. In fact health-related
financial stress is one of the main sources of personal bankruptcy. Although we do not present a
detailed study of the U.S. health care system in this chapter, the numbers suggest that the way health
care is organized in the U.S. may not be efficient, at least at the population level. Some possible
reasons for this inefficiency include an increasingly large aging population, expensive advanced
health care options, expensive treatments that are the result of uninsured people delaying treatment
due to lack of resources, and expenses due to litigation over medical errors. It is important to note
that not every actor in the health care system will consider the U.S. health care system as problematic.
The insurance industry, lawyers, and producers of medicine and medical equipment all benefit from
the current system and will protect the current configuration. This explains the difficulties various
U.S. presidents have had in attempting to change the health care system.
Health care is an extraordinarily complex issue well beyond the scope of this chapter. To illustrate
the nature of the collective action challenges in health care, we will now focus on three examples of
social dilemmas in public health: vaccination, organ donations and sanitation.
12.2 Vaccination 131

Country Expenditure Government Life Improved Improved
on health as vs private expectancy drinking sanitation (%
% of GDP spending at birth water (% population)
population)
USA 15.2 48 / 52 79 99 100
Canada 9.8 70 / 30 81 100 100
Mexico 5.9 47 / 53 76 94 85
Germany 10.5 75 / 25 80 100 100
France 11.2 76 / 24 81 100 100
Sweden 9.4 78 / 22 81 100 100
India 4.2 32 / 68 65 88 31
China 4.3 47 / 53 74 89 55
Ethiopia 4.3 52 / 48 54 38 12

Table 12.1: Source: World Health Organization, 2011 Health Statistics

12.2 Vaccination
Vaccinations protect us from exposure to many diseases, like the annual flu. Once vaccinated, the
likelihood of getting sick and then spreading the disease is reduced. If enough people in a population
are vaccinated, there develops a so-called herd immunity to contagious diseases (Figure 12.1). Once
herd immunity is established, even people who have not been vaccinated are protected because the
disease cannot spread through the population.
For example, in order to establish herd immunity for measles, about 90% of the population needs
to be vaccinated. For polio, this number is about 85%. Establishing herd immunity is important since
not every individual gets vaccinated. Some people refuse vaccination for religious reasons. Others
may feel that the risks posed by side effects from vaccination such as allergic reactions outweigh the
benefit of the reduced risk of infection and choose not to get vaccinated. This cost-benefit calculation,
however, rests on the assumptions that most everyone else has been vaccinated. If everyone made the
choice to not vaccinate based on concerns of side effects, the risk of infection would increase rapidly,
reversing the cost-benefit calculation. This is a social dilemma with very serious consequences.
Again, providing herd immunity boils down to a public goods problem, and people who don’t get
vaccinated are free-riding on the public good contributions of others.
For example, in 1998 a paper was published in the top journal Lancet that showed the results of
a study of side effects to the MMR vaccine (measles, mumps and rubella). The study argued that the
vaccination could lead to autism. A few years later the publication was retracted because of scientific
misconduct, but the conclusions created a scare for parents considering MMR vaccinations for their
children. The vaccination rate declined below the threshold for herd immunity within the U.K. and
parts of the U.S. This led to the return of measles. After initial outbreaks, measles was declared
endemic again in 2008 after being eradicated for more than 14 years (See Wikipedia aticle on the
MMR vaccine controversy).
132 Public Health and Collective Action

Figure 12.1: Diagram illustrating how herd immunity works. When more people are vaccinated, the spread of
the illness can be contained.

12.3 Organ donation
Increasingly, the health of individuals depends on organ transplants from fellow humans. According
to the statistics of the World Health Organization, around 66,000 kidney transplants, 21,000 liver
transplants and 6,000 heart transplants were performed globally in 2005. Still there is a shortage of
organs and people die waiting for an available organ.
There are different ways to facilitate the provision of organs. Most organs can be harvested only
when the host dies. As a consequence, people need to indicate in advance whether they are willing to
have their organs available for organ donations. Although many people approve of organ donations,
there is still a shortage. This is mainly because people do not indicate they are willing to donate their
organs.
Johnson and Goldstein (2003) showed that the institutional arrangements for organ donations
strongly affect the availability of donated organs. They distinguish between opt-in and opt-out
conditions. In an opt-in situation you are a potential organ donor only if you have indicated this
explicitly through various methods such as filling out a form that you want to be an organ donator. In
countries where the opt-in mechanism is available, the percentage of the population who are potential
donors is between 4 and 8 percent. The other option is opt-out. In that case everybody is a donor by
default except if you have explicitly indicated that you are not willing to be a donor. The difference
in consent percentages is significant (Figure 12.2). If people are donors by default, and have to take
action not to be a donor, the consent percentage is close to 100 percent in most countries.
Note that in both cases being a donor is a voluntary choice. In terms of institutional statements,
a person MAY be a donor. However, the initial conditions differ in the action situation of potential
organ donation in the opt-out and opt-in cases.
How is organ donation organized in Arizona? Arizona has an opt-in policy. Since 2008 you
can sign up for organ donation when you get your driver’s license or an identity card. You can also
register at the following website: https://www.azdonorregistry.org/. When you register as an organ
donor you give consent for the following actions:
• A physician needs to provide a legal declaration of death.
• There is no cost to the family related to the donation.
• The identity of the donor will only be shared with organizations and individuals related to the
donation process, such as hospital staff, but not to the receivers of the donated organs.
• Donors will be tested for transmissible diseases to ensure safety of transplant recipients.
12.3 Organ donation 133

Figure 12.2: Effective consent rates, by country. Explicit consent (opt-in, gold) and presumed consent (opt-out,
blue). (Source: Johnson and Goldstein, 2003).

• Arizona laws require that positive confirmed tests will be reported to the Arizona Department
of Health Services.
• The placement of organs and tissues is directed by local, state, and federal guidelines and
guidelines in transplantation countries.
• Tissues may be sent to help people in other countries.
• Recovered tissues may be reshaped and reformed to accommodate transplant needs.
• Both nonprofit and for-profit organizations may be involved in the process to increase the
benefits of organ donation programs for organ recipients.
• Donation includes the careful restoration of the donor, but bruising and swelling can occur.
• The donation process does not normally interfere with funeral arrangements.
Note that these statements are mainly information of what may happen. The donor gives consent
with potential consequences and use of the organs and tissues.
You can specify which organs you want to donate. Menu items exist for heart, heart valves,
eyes, small intestine, kidneys, bone and associated tissue, liver, veins and arteries, lungs, skin tissue,
pancreas and pericardium. Furthermore, one can specify whether the donation is for transplantation
and research or transplantation only. Hence, a potential donor may have to specify from among
many options to complete an opt-in donation form. This large number of decisions requires some
work and may be enough of a barrier to prevent people from opting-in.
Due to the shortage of donors and the poverty of many potential organ donors around the
world, some organs like livers and kidneys are harvested from live donors who receive financial
compensation. In countries like India and the Philippines, trade of organs was legal until 1994. In
fact, India became one of the world’s leading kidney transplant centers in the world. Due to problems
in this sector such as organ harvesting without consent or with lower compensation than agreed
upon, organ trade was banned in 1994. This is a good example of a case in which the attributes of
the community play a major role in the outcomes.
In China organs are mainly harvested from prisoners who sign an agreement before execution. In
Iran selling one’s kidney for profit is still legal. There is no waiting list for kidneys and the average
134 Public Health and Collective Action

reward received by the donor is 1200 dollars.

12.4 Sanitation
According to the World Health Organization there are about 2.6 billion people without access to
basic sanitation. This means that the most basic human necessities are carried out in open fields
without proper disposal. This contributes to the spread of pathogens and deteriorates the health of
billions of people. Every day 5000 children die from diarrheal diseases. The reason for this impact
can be explained using the F-diagram (Figure 12.3).

Figure 12.3: The F-diagram (Wagner and Lanoix, 1958).

In the F-diagram, we see the spread of pathogens coming from feces. If there is no proper
sanitation pathogens spread via four routes: (1) fluids from the human waste will be spread in the
environment and this can lead to contamination of food; (2) human waste in an open field can be
used as manure or an energy source, but also can come into contact with food; (3) flies can bring
pathogens from feces to food and other persons; and (4) “fingers” refer to the lack of proper hygiene
after disposing the human waste. In areas without proper sanitation people consume about 10 grams
of human waste per day, which contains millions of pathogens.
How can the spread of pathogens be reduced? One can invest in the use of clean water for
preparing and washing food (and reduce the pathogen input through fluids and fields). Another
option is to increase hygiene, such as hand washing, which reduces the spread of pathogens via
fingers. Proper sanitation cuts the pathogens spread via fields, fluids and flies.
Why doesn’t everybody have proper sanitation? It is likely that most of the readers of this chapter
make every day use of clean and modern toilets and flush the toilet with water that is kept separate
from drinking water. But proper sanitation facilities in the western world have only become broadly
available since the mid 1800s, starting in England. In 1854, John Snow showed the relationship
between contaminated water and the spread of diseases such as cholera and triggered an awareness
of the importance of proper infrastructure for both sanitation and drinking water. It was only after
World War II that, with the support of the federal government, proper sanitation infrastructure was
12.5 Critical reflections 135

installed and wastewater began to be treated instead of dumped into rivers from which drinking water
was being tapped.
These large investments in infrastructure have not yet been carried out in many developing
countries. It is expensive, there are cultural barriers, and not all places in the world have sufficient
access to water to flush human waste. A typical error that is made in development projects is to focus
on the technology. Providing latrines to poor people does not solve the sanitation problem. Often,
these latrines end up as storage facilities.
Sanitation is not only a technical problem; it is also
a problem of provision of public goods. Sanitation sys-
tems need to be built, used in an orderly fashion, cleaned,
and maintained. Cleaning toilets is typically not seen as
the most honorable activity. Once communal latrines are
provided, the problem of proper maintenance needs to be
addressed. If latrines are not properly maintained, people
will go back to using open fields (Figure 12.4).
There are also cultural barriers. Disposing your waste
in the same location as others is not considered a proper
activity in various cultures. In some cultures, such as
some tribes in Madagascar, human waste is not supposed
to enter the soil (which happens in latrines) since the soil Figure 12.4: Contaminated open fields
is the place where their ancestors live. spread pathogens throughout the environ-
In order to increase the level of sanitation infrastruc- ment.
ture performance, new approaches have been developed
that shift the focus away from technology and toward cultural practices and collective action. Efforts
to promote awareness in the community of the benefits of sanitation and to develop practices to
maintain the latrines have produced successes in various places in India and Bangladesh.

12.5 Critical reflections
Individual health care and public health are very complex issues that involve social dilemmas on
many different levels. Caring for those who are ill, preventive measures to reduce the spread of
diseases, and the provision of organs for transplantations are all examples of public good provision
problems that affect public health.

12.6 Make yourself think
1. Do you know people without health insurance?
2. Did you get a flu shot during the last flu season? If not, why not?
3. Why can’t our sanitation system necessarily be applied in countries with low levels of sanita-
tion?
4. How is organ donation organized in your state? Is it an opt-in system or opt-out system?

12.7 References
Johnson, E., & Goldstein, D. (2003). Do defaults save lives. Science, 302, 1338–1339.
136 Public Health and Collective Action

Rosen, G. (1993). A history of public health, expanded edition. Baltimore and London: John Hopkins
University Press.
Wagner, E., & Lanoix, J. (1958). Excreta disposal for rural areas and small communities. WHO
Monograph series No. 39. Geneva: World Health Organization.
Key Concepts
In this chapter we will:
– See the consequences of new technology for institutional arrangements
– Learn how the concepts we have developed in this book can be
applied to the new digital commons

13 — The Digital Commons

Death in the digital world

Increasingly we share a great deal of personal infor-
mation online. We talk with friends and colleagues on our
Facebook pages and twitter accounts, share photos through
Flickr, make payments with PayPal and store our music col-
lection in iTunes. What happens with this information when
we die? Who has access to our accounts and who owns our
music collection? New legislation, i.e., a new set of rules, is
needed to deal with the digital belongings of the deceased.
Individual social network sites may have their own policies.
Twitter, for example, will delete a user’s account once it has
received a death certificate. Since legislation is lacking some
people are being advised by their legal counsel to appoint Figure 13.1: Digital death.
someone in their will to manage their digital legacy includ-
ing account details and passwords. There are also firms, such as Entrustnet and Deathswitch, which act
as an intermediary and will send details of your accounts to the people you have specified in case of your
death.

13.1 Introduction
Since the mid-1990s there has been an increasingly rapid development of the digital commons. By
digital commons we mean software and hardware that is connected via the Internet. The Internet
is a piece of infrastructure that makes information easily available. Thus, as with many examples
in this book, infrastructure makes a resource available. The resource in this case is information.
Every file you download is, in the end, a string of 0’s and 1’s. Once you have the file, your use
of it does not impact others use of the same file. What resource does the Internet infrastructure
actually provide? Does it provide the information, i.e., the music or video content that is available
for download? No, it does not. A person or group of people somewhere provides the information
encoded in the string of 0’s and 1’s. What the Internet provides is information transport capacity,
and this is why the Internet generates a “digital commons.” It provides a vehicle that people can use
to make information available. And much of the information that is made available has no formal
private property rights associated with it, and thus qualifies as a “commons.” This information varies
from free encyclopedias, open source software, cloud computing, and buying and trading via the
web to sharing photos, videos and ideas on social networks, blogs and twitter. Conflict over the use
138 The Digital Commons

of the digital commons and those whose intellectual property is being made available has emerged
around downloading music, books and movies, and many more activities.
The increasing number of opportunities to share information has led to many governance
challenges. Existing regulations are not appropriate for many of the current uses observed. Which,
if any, sales tax should apply to a product purchased on the Internet? If a buyer in Australia is a
victim of a suspicious transaction in the U.K. and loses 5,000 dollars, which jurisdiction, if any, can
act and make a case against the person in the U.K.? Consider if someone stole your car in Phoenix.
Which police department would you call? This is obvious, right? There is a very clear physical area
associated with car ownership and the boundaries are clear (remember Design Principle 1 in Chapter
7?). With digital information, the boundaries are not so clear.
The governance of the digital commons is in constant flux due to the many new problems people
experience, from virtual rape to stolen credit card information, from cyber bullying to cyber attacks
on government servers. In this chapter we will discuss three examples of digital commons most
of you will be familiar with and will go into a bit more depth on what you have actually agreed to,
what the rules are, and how these digital commons are governed. We will look at the public good
provision of Wikipedia, the privacy aspects of Facebook and other social networks, and intellectual
property of music files.

13.2 Wikipedia1 : a global public good.
Wikipedia is an encyclopedia freely available at
www.wikipedia.org (Figure 13.2). Wikipedia started on
January 15, 2001 and was created by Larry Sanger and
Jimmy Wales. Using wiki, websites that can be edited and
created via a simple web browser, 20,000 articles were cre-
ated by the end of 2001 in 18 different languages. At the
moment (April 2013) there are articles in 286 languages,
100,000 regular active contributors, and 26 million articles,
all written by volunteers from all over the world.
Figures 13.3 and 13.4 show the growth and rate of
growth of the number of articles in the English language.
The number of new articles peaked in 2006 and has been
rapidly declining since. Today, most topics are established,
and most new articles relate to current events. Still a lot of
contributions are needed in existing articles to keep them
up to date and at a high quality level. Just like irrigation
canals need to be cleaned and repaired each year, so does
information in the digital commons.
Figure 13.2: Wikipedia logo.
Wikipedia is a global public good. The information it
makes available is truly a pure public good. More people
consume (i.e., read) articles from Wikipedia than contribute. In fact, most people free ride on
the contributions of others. Still Wikipedia has enough contributions to be up to date and of high
quality. When a random set of articles were compared with articles from Encyclopedia Britannica
the accuracy of the information was the same, though the style of writing was better in Encyclopedia
1 Much of the information on the governance of Wikipedia is derived from the Wikipedia site
13.2 Wikipedia: a global public good 139

Britannica (Giles, 2005, pp. 900-901).
What explains the success of Wikipedia? There are different motivations for why people might
choose to contribute. Some people make small edits of articles on topics they know a lot about,
others spend many hours a week on editing and writing. Those who make large contributions develop
a certain reputation within the Wikipedia community. Although articles have no authors, one can see
the user names of Wikipedians who make many contributions. Those who develop a good reputation
may get invited to take on more responsibilities. Those with more responsibilities can mediate
disputes on articles. Although Wikipedia looks like an encyclopedia constructed by contributions of
many individuals, there is a global community with hierarchical layers.

Figure 13.3: Number of articles on Wikipedia since 2001.

Since the open wiki editing model is used, everyone can edit an article anonymously or through
an account. There are exceptions for some pages that are prone to vandalism, such as pages related
to celebrities or politically sensitive topics. Wikipedia pages in different languages have adopted
some changes to this open access situation. For the English language version of Wikipedia only
registered users can make edits.
Articles are not owned by anybody. Once an article is posted to Wikipedia, anybody can edit
the text. Each edit page clearly states “If you do not want your writing to be edited, used, and
redistributed at will, then do not submit it here.” Editing wars can erupt where participants remove
and adjust edits of others in an iterative way. If this is noticed, editors can receive help in dispute
resolutions. There are various levels of dispute resolution, depending on the severity and kind of
dispute (http://en.wikipedia.org/wiki/Wikipedia:Resolving_disputes).
The default situation of any edit to an article is that it becomes immediately available as edited.
Therefore articles can contain errors, advocacy or nonsense until another editor makes edits. For
each article, the history of edits can be viewed. You can see which changes are made when and by
whom (Figure 13.5).
Each language version of Wikipedia can make some adjustments to this default setting. The
140 The Digital Commons

Figure 13.4: Wikipedia’s monthly growth rate.

Figure 13.5: Example of Wikipedia article history comparison.
13.3 Social networking 141

German version, for example, contains stable versions of many articles that have passed a certain
level of reviews.
Although Wikipedia is a global website with contributions from all over the world, it is under
the jurisdiction of the laws of the United States and the U.S. state of Florida where the servers of
Wikipedia are hosted.
Wikipedia distinguishes different positions of users who have different rights of access. The
first distinction is whether a user is registered or not (an example of a boundary rule). Every-
body can edit pages (the default choice rule), but registered users can start new pages (a choice
rule restricted to registered users) and gain a reputation within the community of Wikipedians.
There are about 19 million editors active in the English version of Wikipedia as of April 2013
(http://en.wikipedia.org/wiki/English_Wikipedia)
Those editors who are very active and have gained a good reputation can apply to become an
administrator with more rights to change content and resolve disputes. An application is discussed
in an open forum for seven days after which a bureaucrat makes a final decision on the application.
There are 1445 administrators in the English language version of Wikipedia. There are 13 arbitrators
who have even more rights to change content on the website and resolve disputes. Finally there is
a group of 43 stewards from the various Wikipedia language groups. The stewards have complete
access to the Wikipedia website.
This brief look into Wikipedia shows that it is not a spontaneous collaboration among millions of
people. There are clearly defined rules at different levels of organization. Let’s use the rule typology
of Chapter 10 and see whether we can find examples from Wikipedia for all those rule types:
• Position: Unregistered and registered users. Among the registered users there are four levels:
editor, administrator, arbitrator, and steward.
• Boundary: To become an administrator, one must be a registered user (editor).
• Choice: Administrators are Wikipedia editors with access to restricted technical features such
deleting and restoring pages. That is, an administrator may delete a page.
• Aggregation: Stewards are elected at least once a year by the global Wikimedia community.
Candidates must have a support/oppose ratio of at least 80% with at least 30 supporting users.
• Information: The history of edits is stored for each article. You can see which changes are
made when and by whom.
• Payoff : An Arbitrage committee may impose sanctions on editors. Such sanctions include the
restriction of access to Wikipedia.
• Scope: The outcome of an application for becoming an administrator is either that one is
approved to be an administrator or one remains an editor.

13.3 Social networking
People are increasingly sharing information on social networking sites such as MySpace, Facebook,
Google+, etc. This information can be personal or professional. The pages of a particular individual
are linked with others, i.e., the friends of the individual.
There are various challenges with social networking websites. Who owns the information on
the site, who has access to participants’ information, and what kind of material is allowed to be
put on such sites? Recall one of the rule classes was information rules. Information rules are of
critical importance in the digital commons. We will look at these issues by analyzing the terms and
conditions of Facebook, which can be found at http://www.facebook.com/terms.php.
142 The Digital Commons

Facebook was launched in February 2004, and was developed by Mark Zuckerberg and his
collaborators (Figure 13.6). By the time of writing this chapter, one billion people have a Facebook
account. People with an account, which is provided at no cost, can post personal information and
share it with friends or the rest of the world, depending on the privacy settings the user has adopted.
The rules of Facebook deal mainly with how to use the Facebook
site:
• Position: Facebook, Facebook users, friends, developers of
Facebook applications, and advertisers.
• Boundary: A person can become a Facebook user if (s)he is
13 years or older and is not a convicted sex offender.
• Choice: A developer of an application has restricted access
to user data and this data cannot be sold to a third party.
• Aggregation: By agreeing to the terms of Facebook, you
allow Facebook to make decisions regarding your account
(aggregation rules determine who is to decide when multiple
agents interact). In the case of a dispute, the state or federal
court located in Santa Clara County, California, will preside.
• Information: A user can define the group of other users with
whom he or she will share what kind of data via the privacy
settings.
Figure 13.6: Mark Zuckerberg.
• Payoff : Access to the website can be withheld if a user breaks
the rules.
• Scope: The users are responsible for the information they
share and with whom they are sharing it.
Although there are clear rules on paper, are you confident that you actually have control over it?
Do you have control over your television or your car?

13.4 Intellectual property
If you create something whose value is based solely on its information content, (e.g., a song, a
video, software, etc.) who owns it, who can access and use it and receive revenues? This happens
to be a very complicated topic considering that now it is easy to copy a creation (since they are in
digital formats) and it is not always clear what or who the source of the creation is. The information
transport infrastructure we call the Internet allows people to easily share digital files at extremely
low cost and has led to various new developments regarding how to define intellectual property and
how to protect it.
In June 1999, 18-year-old college student Shawn Fanning launched Napster (Figure 13.7).
Napster was an independent peer-to-peer file sharing service. Peer-to-peer file sharing simply means
that people can download files from the computers of other peers. Napster made it easy for people to
share their mp3 files. Due to the Napster service, millions of people started to share their music files,
which led to copyright violations on a massive scale. Once they had created an account on Napster,
people could download music for free, creating their own CDs without paying royalties to the artists
and composers. Furthermore, computer networks became overloaded due to the high volume of file
transfers.
13.5 Critical reflections 143

The music industry became upset about the activities
of Napster. Recording companies and music icons like
Metallica and Madonna started to sue Napster. In July
2001 the service was closed by court order.
The music that talented musicians make could be con- Figure 13.7: Napster logo.
sidered to be a contribution to the public good (because
music exhibits the standard attributes of a public good), but they can only make music if their
creations and performances are rewarded. The free downloading of files is an example of free riding
on a massive scale. The music industry would not survive if they did not get rewards for their
products. An interesting question is how much reward is enough.
Napster triggered innovation in the music industry. In January 2001, Apple launched iTunes,
where you can buy mp3 files that you then own legally. The mp3 file becomes a private good. Still
it is relatively easy to copy mp3 files. The music business has therefore changed in recent times.
Nowadays music groups do not depend for their income solely from selling recorded music but,
rather, largely from live performances. In fact, distributing music freely on the Internet has become a
method of promoting an artist. This is in stark contrast to concert tours being used largely to promote
album (cd) sales in the last three decades of the 20th century.
The problem faced by the music industry is also visible in other industries. Companies that rent
DVDs and stream video recordings, like Netflix, pose a challenge to the movie industry. In the fall
of 2011 Netflix experienced a large decline of their stock value due to customer revolt in response to
the price hike Netflix implemented to cover the increase in copyright fees it now had to pay to movie
companies.
The book industry is also experiencing many challenges. Besides the fact that many people now
buy their books online (through vendors such as Amazon.com), the increasing popularity of e-books
and audio books makes it seem as if the traditional book store may disappear. The Google project to
scan all published books and make them available online adds another factor that needs to be sorted
out in terms of intellectual property.
The examples above describe situations that work in favor of small groups who have made a
finished product that they want to distribute. There are other types of activities for which it can
be beneficial not to have limited intellectual property rights. In software development there is a
movement called “open source software.” In contrast to software that is protected by intellectual
property rights, open source software can be used for free. If people make adjustments to the software
they are required to make it available again as open source software.

13.5 Critical reflections
With the rapid digitalization of many parts of our lives, new challenges emerge for which traditional
institutional arrangements do not hold. Basic concepts like ownership face new challenges in the
digital world. Nevertheless the tools of institutional analysis, including the careful definition of
action situations and the rules that apply as discussed in this book, can be applied to effectively
manage the new digital commons.
144 The Digital Commons

13.6 Make yourself think
1. Read the terms and conditions of one of the social network services you are involved in.
2. How do creators of music and books benefit from their creations if they are freely distributed?

13.7 References
Giles, J. (2005). Internet encyclopedias go head to head. Nature, 438(7070), 900–901.
Key Concepts
In this chapter we will:
– Evaluate action situations in sports from an institutional perspective
– Learn how to apply the syntax of institutional statements to sports
– See that sports are a fruitful topic to which to apply lessons from this book

14 — Rules of the Game: Applications to Sport

Corruption in sumo wrestling

Sumo wrestling is a traditional sport that goes back
more than 1000 years (Figure 14.1). For years speculations
over widespread match fixing circulated. The Japan Sumo
Association always denied that there were any problems.
However, in 2011 an investigation uncovered cell phone
text messages that proved that matches were being fixed.
Various wrestlers admitted wrongdoing and the March 2011
tournament in Osaka was cancelled.
In 2002 economists Duggan and Levitt published an arti-
cle that showed that the incentive structure of the tournament
stimulates corruption. In a tournament, each competitor will
participate in 15 fights in 15 days. If the wrestler has a record
of 8 wins and 7 losses, he will increase in rank, while a 7-8
score leads to a decrease in rank. Hence for a player who has Figure 14.1: Sumo wrestler Asashōryū
7 wins and 7 losses, a win in the last game is very valuable, fighting against Kotoshogiku at the Jan-
especially compared to an opponent who already lost or won uary Tournament 2008.
8 fights. Duggan and Levitt did a statistical analysis over a
12-year period of match data and found that 77% of the 7-7
wrestlers win this last match, which is unusually high. In fact, if the winners face the same opponent in a
next tournament they lose 74% of the time. This example shows that the structure of the tournament
affects how the athletes approach the matches.

14.1 Introduction
We can apply the lessons from this course to study sports. More importantly, sports can serve as
simple test cases for the theories that help us understand more complex situations. Sports are to the
study of institutions as fruit flies are to the study of evolutionary biology.
Sports as entertainment is a good. When we pay to watch professional sports, the good is a
club good. When we watch the Olympics when covered by a standard network TV broadcaster,
sports become a public good because everyone can enjoy it for free (except for time you might spend
watching advertisements). Sports leagues can also be commons in the sense that the reputation of a
sport or league (and thus its value) is affected by the contribution of the participants. Doping scandals,
for example, have reduced the popularity of cycling, baseball and some other sports. A sport league
is a productive commons when the participants play fair, and the differences in performance among
146 Rules of the Game: Applications to Sport

the participants (individuals or teams) are not too big.
A benefit of looking at sports is that the action situation is self-contained and well documented.
Further, sports are a ubiquitous cultural currency; most of the readers of this book are familiar with
sports as a spectator or practitioner. Thus, many readers will be able to identify with the ideas we are
developing in this book as they apply to sports more readily than in more specialized situations such
as fisheries, forests, or groundwater systems with which they are less familiar. Thus, in this chapter,
we will discuss some examples from sports from an institutional analysis perspective. The main goal
is to provide you with a new lens to look at your favorite sports and provide opportunities to work
with institutional analysis tools.

14.2 Individual sports
In individual sports, like sprinting 100 meters, tennis, or boxing, individuals are competing against
each other. The action situation is a defined time and space where two or more individuals compete
against each other. The positions include the players, but can also include referees who are monitoring
the action situation.
How decisions are made in sports depends on the type
of sport being played. For example, in a 100-meter sprint,
the decisions sprinters make might not be conscious. After
many years of training, responding to the starting signal
might be an automatic reaction, and the sprinter will only
realize what happens after the finish. In a boxing match,
the boxer makes many decisions: where and when to hit
the opponent, to initiate an attack, retreat, or to give up if
injuries become severe.
Who decides who has won? In a 100-meter sprint, the
moment a runner’s chest crosses the finish line defines the
race time of the individual (a scope rule) (Figure 14.2).
Sometimes a jury needs to study a photo of the finish to
determine the winner. In figure skating, a jury will award
points according to a table of points for certain types of ac-
tions. Jury members can be subjective in interpreting how
many points to assign to an action during a performance.
Figure 14.2: A close finish at the 2007 world
To reduce the level of subjectivity, there are a number of championship race of the 100 meters.
jury members who make independent decisions which are
aggregated according to the following aggregation rule:
the highest and lowest scores are taken out of consideration, and the remaining scores are summed
up to compare with other skaters. The use of a jury can lead to strategic manipulation.
In various sports, individuals use equipment and the rules of the game specify what equipment
is allowed (scope rules). For example, a javelin for adult men in international track and field
competition consists of a metal head, a solid or hollow shaft and a cord grip. The javelin must weigh
at least 800 grams (why is there no maximum weight specified?) and must be between 2.6 and 2.7
meters long. The shaft can be made of either wood or metal. As we will see below, there are a very
large number of scope rules governing equipment, and the number of scope rules increases with the
level of professionalism.
As an example of rules in individual sports we will look at amateur boxing in more detail
14.2 Individual sports 147

(Figure 14.3). These rules are also used during Olympic boxing events and can be found at
http://www.aiba.org. The official rules of the International Boxing Association are documented in
a 71-page booklet, which is divided into three categories: rules governing the competition, rules
governing the referees, and rules governing the equipment. Let’s look at just one of the rules from
the first category, medical examination. For boxers to compete, one has to determine their weight,
verify the weight class the boxer is in, and determine if the boxer is fit to compete (a type of boundary
rule applying to the position of being a “boxer fit to compete”).

4.1. At the time fixed for the Weigh-In, the Boxer must be considered as fit to compete
and examined by a Physician appointed by a member of the Medical Jury before being
weighed-in. With a view to ensure a smooth running of the Weigh-In, the Technical Del-
egate may decide to begin the Medical Examination at an earlier time. The Chairperson
of the Medical Jury can select local or Team Doctors to assist with the duties of the
Medical Jury only for the Medical Examination.

This boundary rule defines the specifications concerning when
a boxer can enter the position of “boxer eligible to weigh in.” There
is a scope rule for the time frame for examination, and who may
perform the examination. Note that the rule specifies options by
using the term MAY in various places.

4.2. At the Medical Examination and Weigh-In, an up-
to-date Competition Record Book with correct informa-
tion must be presented by the Boxer. This Competition
Record Book must be signed and/or stamped by the
Secretary General or Executive Director of the Boxer’s
National Federation, where the requisite entries shall
be made by the Officials in charge. If a Competition
Record Book is not presented by the Boxer at the Med-
ical Examination and Weigh-In, this Boxer shall not be
allowed to compete.
Figure 14.3: Headgear is re-
Rule 4.2 focuses on information and defines that the boxer must quired in amateur boxing.
provide the right information in order to compete. Rule 4.3 (below)
also provides an option for the referee to consider.

4.3. Gender Test—For AIBA Approved Events, a gender
test can be done.

Rule 4.3 is a choice rule associated with the position of approved medical examiner. That is, the
medical examiner may conduct a gender test.
148 Rules of the Game: Applications to Sport

14.3 Team sports
Compared to individual sports, team sports are
much more complicated. One simple reason is
that there are more possible positions held by
the players. Each position can have different
choice rules. Fore example, the goal keeper in
soccer may touch the ball with her hands under
particular circumstances while the field players
must not or else they may be penalized by, for
example, conceding a free kick to the other team.
Furthermore, there are rules concerning when
players in positions can be replaced (a boundary
rule), how many players in different positions
are allowed (a scope rule—e.g., there may be no
more than nine field players on the field at any Figure 14.4: The official measurements of a soccer
field.
one time during a match), etc.
Let’s look at a particular sport and discuss
the rules in more detail. The world’s most popular sport is soccer. The official rules of soccer
can be found on the website of the Fédération Internationale de Football Association (FIFA) (http:
//www.fifa.com/aboutfifa/officialdocuments/doclists/laws.html). The document, titled “Laws of the
Game” has 138 pages!! The main text of the official rules consists of the following sections:
1. The Field of Play (scope rules) (Figure 14.4)
2. The Ball (scope rules)
3. The Number of Players (position rules)
4. The Players’ Equipment (scope rules)
5. The Referee (position and choice rules)
6. The Assistant Referees (position and choice rules)
7. The Duration of the Match (scope rules)
8. The Start and Restart of Play (scope rules)
9. The Ball in and out of Play (scope rules)
10. The Method of Scoring (scope and payoff rules)
11. Offside (scope rules)
12. Fouls and Misconduct (choice and payment rules)
13. Free Kicks (scope and payoff rules)
14. The Penalty Kick (scope and payoff rules)
15. The Throw-in (scope and choice rules)
16. The Goal Kick (scope and choice rules)
17. The Corner Kick (scope and choice rules)
Looking at this list from an institutional analysis perspective we see that a lot of emphasis is
given on defining the positions (players, referee), the pay-off rules (method of scoring, fouls and
misconduct), and the physical boundaries (scope rules) of the game such as the duration of the match,
start and restart of play, ball in and out of play, etc. We can also see that the regulations of the game
are actually rules since there are clearly defined consequences assigned to the act of breaking the
rules (free kicks, penalty kick). However, in practice it is up to the referee to interpret the rules.
14.3 Team sports 149

Let’s look at some of these rules in more detail such those governing the number of players. The
first rule is

“A match is played by two teams, each consisting of not more than eleven players, one
of whom is the goalkeeper. A match may not start if either team consists of fewer than
seven players.”

We can rewrite this using the grammar of institutions:
1. A match is played by two teams. This is a scope rule.
2. Each team MUST have a goal keeper. This is a position rule that could be written as “on each
team there will be the position of ‘goal keeper’ OR ELSE a match cannot start.”
3. Each team MUST have field players. This is a position rule that could be written as “on each
team there will be the position of ‘field player’ OR ELSE a match cannot start.”
4. Each team must have at least six field players and MUST NOT have more than ten field
players, OR ELSE a match cannot start. This is a scope rule.
We see here that each team has two positions: field players and a goalkeeper, and that there is
a maximum and minimum number of players defined by scope rules. There are additional rules
affecting the number of players:

“Up to a maximum of three substitutes may be used in any match played in an official
competition organized under the auspices of FIFA.”

This is a scope rule. This rule also implies that there are three positions on a team: field player,
goal keeper and a substitute. There is an additional rule related to the substitutes.

“In all matches, the names of the substitutes must be given to the referee prior to the
start of the match. Any substitute whose name is not given to the referee at this time
may not take part in the match.”

This is an information rule that states that all information regarding names of players who may
occupy positions need to be defined at the start of the game and this information needs to be provided
to the referee prior to the start of the match. Players can change their position during the match.
Substitutes can enter the match, but this transition has to follow a set of specific (boundary) rules

To replace a player with a substitute, the following conditions must be observed:
• the referee must be informed before any proposed substitution is made,
• the substitute only enters the field of play after the player being replaced has left
and after receiving a signal from the referee,
• the substitute only enters the field of play at the halfway line and during a stoppage
in the match,
• the substitution is completed when a substitute enters the field of play, from that
moment, the substitute becomes a player and the player he has replaced becomes
a substituted player,
• the substituted player takes no further part in the match, and
• all substitutes are subject to the authority and jurisdiction of the referee, whether
called upon to play or not.
150 Rules of the Game: Applications to Sport

This set of rules includes information rules (referee must be informed), boundary rules (when
substitute is replaced).
The goal keeper is a specific position, and there is a specific boundary rule on how to change the
position of the goal keeper.

Any of the other players may change places with the goalkeeper, provided that:
• the referee is informed before the change is made, and
• the change is made during a stoppage in the match.

Also here we see a combination of information and boundary rules.
Note that while we are referring to rules, most of the rules above did not include an explicit or
else. The OR ELSE condition is necessary if we want to call them rules instead of norms. The OR
ELSE condition is defined separately in the case that any of the above situations is not met.

If a substitute or substituted player enters the field of play without the referee’s permis-
sion:
• the referee stops play (although not immediately if the substitute or substituted
player does not interfere with play);
• the referee cautions him for unsporting behaviour and orders him to leave the field
of play; and
• if the referee has stopped play, it is restarted with an indirect free kick for the
opposing team from the position of the ball at the time of the stoppage (see Law
13—Position of Free Kick).
If a player changes places with the goalkeeper without the referee’s permission before
the change is made:
• the referee allows play to continue and
• the referee cautions the players concerned when the ball is next out of play.
In the event of any other infringements of this Law:
• the players concerned are cautioned and
• the match is restarted with an indirect free kick, to be taken by a player of the
opposing team from the position of the ball at the time of the stoppage (see Law
13—Position of Free Kick).

One of the current debates in soccer
is whether referees should use information
from video cameras in making decisions (Fig-
ure 14.5). If it happens that from the audience’s
perspective it looks as if a goal has been made,
but it was not clear from the referee’s perspec-
tive, the referee can ask the assistant referees
for information in making decisions, but cannot
use other information. (It also happens that the
audience may see foul play that is not noticed by
the referees.) This particular information rule
can affect which country wins the World Cup.
Figure 14.5: England’s shot went behind the line but
goal was not approved by referee.
14.4 Sport leagues 151

14.4 Sport leagues
A sport league is a group of individuals or teams
that are performing in an organized competition.
By having a league there will be additional rules that determine how teams qualify for the league
(scope rules) and the organization of the matches that define the winner of the league (scope and
pay-off rules).
Besides these extra rules, we now get additional po-
sitions in the action arena that play an important role,
namely the owners of the league, the owners of the teams,
and the fans of the teams. Since huge resources are at
stake for teams and the league to be successful, we see
specific rules on transfers of players, drafting new players,
and rules related to advertisement.

14.5 Performance-enhancing drugs
One of the main controversies in sports is performance-
enhancing drugs. Although the main objective of ath-
letes in competitive sports is to elevate their abilities
above everybody else, it is prohibited to take performance-
enhancing drugs in this pursuit. Not all performance-
Figure 14.6: Due to disagreements between
enhancing drugs are illegal (e.g., coffee), so what makes players and club owners there was a 2011
a substance illegal? How is this measured? At what level NBA lockout. A shorter version of the regu-
is somebody considered to have an illegal dose? lar season was implemented when finally an
Many of the performance enhancing substances are agreement was made.
found in regular products such as coffee and meat. That’s
why a certain level needs to be measured in the body to
demonstrate that a person purposely took the substance for
performance enhancing reasons (a scope rule). However, concentrations in humans vary considerably,
and it is possible to unjustly accuse a person of illegal activities. The challenge with governing the
use of performance enhancing drugs is one of monitoring.
Many of the performance enhancing substances are artificial products that are difficult to trace
since users try to mask their use. For example, EPO (Erythropoietin) is naturally produced in the liver.
It is also ingested since it enhances performance in endurance sports. In a sport like cycling, EPO
has been used by most competitive athletes over the last 20 years, although most were not detected
(a problem of monitoring). Reasons for this include the initial difficulty of detecting ingested EPO
as well as the use of EPO during the training season. Retired cyclists nowadays admit to the use of
EPO.
Since doping is difficult to detect during competitions, there is an increasing use of out-of-
competition controls. Professional athletes need to notify governing boards where they are each day
(a combination of a choice and an information rule) so that they can be tested randomly. Depending
on the sport, there are sanctions when somebody is caught with certain substances in their body,
including a lifelong ban from competitive sports. More information on doping and anti-doping can
be found at http://www.wada-ama.org/.
152 Rules of the Game: Applications to Sport

14.6 Boundary rules
In organized sport competitions there are clear rules regarding who is allowed to participate. Even so,
there are some interesting examples where boundary rules are challenging to apply. One of the most
prominent is the issue of gender. Due to physical differences, only females are allowed to participate
in female competitions. This may sound simple, but it so happens that not everyone can be clearly
defined as male or female.
During the 2009 world championship in track and field, Caster
Semenya, an unknown 18-year-old, won the women’s 800-meter
competition by a large margin (Figure 14.7). Her masculinity imme-
diately caused rumors regarding her gender. Gender tests have been
performed but the results have not been revealed. She has continued
to participate in women’s 800 meter competitions ever since.
Why is a gender test not simple? One would think it is simply
a case of checking for XX vs. XY chromosomes to determine
whether an athlete is a woman or a man. However, fetuses start out
as undifferentiated, and the Y chromosome turns on a variety of
hormones that differentiate the baby as a male. Sometimes this does
not occur, and XX people with two X chromosomes can develop
hormonally as a male, and XY people with an X and a Y can develop
hormonally as a female.
Moreover, people can change gender. Transsexuals are allowed
Figure 14.7: Caster Semenya.
to compete in the Olympics after having completed sex reassignment
surgery and having undergone two years of hormonal therapy.
Oscar Pistorius, a young man from South Africa, is the
fastest man without legs (Figure 14.8). He is also known
as the “blade runner” because of his artificial limbs made
from carbon fiber. In fact he is fast enough to compete
in regular events, not just the competitions for amputees.
But the able-bodied runners argued that Pistorius has an
unfair advantage due to the carbon fiber limbs and initially
prohibited his participation in regular competitions. This
decision was reversed when it became clear that his arti-
ficial limbs provided no net advantage over able-bodied
athletes. Pistorius participated in the World Championship Figure 14.8: Pistorius during the semi final
of the World Championship of 2011.
of 2011 in the 400-meter relay.

14.7 Rule changes
Official rules are not fixed. Rules change over time. It is especially common in sports to see rules
changes due to the emergence of new technologies and as a result of innovations intended to increase
the value and competitiveness of the sport. For example, the field of swimwear went through a period
of very rapid technological development during the first decade of the 2000s. World records were
broken everywhere. Since 2010, non-textile (non-woven) swimsuits have been banned in competitive
events, although the records of previous years with the performance enhancing swimwear remain
standing (Figure 14.9). Men are only allowed to wear suits from the waist to above the knees and
women are only allowed to wear suits that do not go past the knees or shoulders.
14.7 Rule changes 153

Another example comes from the U.S. National Hockey League. During the 1998-99 and
1999-2000 NHL seasons, the league conducted an experiment in which some games had two referees
instead of one. Only a subset of all the NHL games played in those seasons were played with two
referees. This makes it possible to measure the net effect of monitoring. If there is more monitoring,
we would expect there would be a decrease of rule breaking.
Several people have studied the effect of an extra referee. It
was found to lead to better monitoring and therefore more penalty
minutes for players who broke the rules (Allen, 2002; Heckelman
& Yates, 2003; Levitt, 2002). These findings are consistent with
the hypothesis that increased monitoring leads to increased detec-
tion of illegal behavior. But these studies did not find changes in
behavior at the individual level. In other words, individual players
did not change their behavior because of increased monitoring and
sanctioning (serving a penalty time that disadvantages their team).
Ahn et al. (2005) looked at the NHL experiment on a longer time
scale and at both the individual player level and the team level. They
observed a significant drop in penalty minutes during the seasons
having two referees compared to the seasons having one referee
(Figure 14.10). They also argued that the extra referee did change Figure 14.9: Forbidden
swimwear on the left.
the pattern of behavior in hockey games, but not because the players
adapted to the new rules. In fact, individual players’ behavior did
not change and, therefore, the dominant short-term effect of the added referee was on the side of
detection, not deterrence. They found, rather, that teams adjusted the composition of player types on
their teams.

Figure 14.10: Average penalties in minutes (PIM) per team per game for 9 seasons, where the average is
depicted as solid dots, and the dotted lines show one standard deviation around the average PIM per team
(Ahn et al. 2005).
154 Rules of the Game: Applications to Sport

The players of particular interest are the so-called “goons.” Goons are private enforcers or
protectors who punish the opposing team either for breaking the rules when such transgressions
are not detected and punished by the referees or for attacking a protected player. This private
enforcement by goons frequently takes the form of rough behavior that is itself a violation of the
rules. Thus, if the referees catch the goon’s retaliation, it is costly to the team because the team
suffers a penalty. Goons are also costly inasmuch as their scoring and defensive skills are not as well
developed as the other players.
With more referees on the ice, there is more public enforcement of the rules and there is less
need for goons. At the same time, goons become more costly since their illegal actions will be
caught more frequently. We would therefore expect the role of goons to diminish when there are
more referees. At the aggregate level, the data show an overall decline of penalties in minutes per
team per game since the introduction of the additional referee. The data also indicate that a large part
of the decline is due to the selection of players by teams, which is shown by the differential patterns
of salary changes between goons and non-goons before and after the introduction of the additional
referee. The salaries of the goons increased less than the salaries of players who are highly skilled in
skating and scoring.
This rule change is an example that illustrates how a change in the regulations can lead to
adjustments in behavior and strategies of individuals and teams. We now see fewer large goons on
the ice who are less skilled skaters but who are able to punish players of the opposing team.

14.8 Critical reflections
The rules of the game are part of the sport. Although the goal of sports is ostensibly about staying fit,
having fun, learning sportsmanship, and learning to work together in teams, when sports become
professional, it is about winning. Given strong profit incentives, rules are very important in maintain-
ing the fairness of sports competitions. Sports also illustrate very clearly that the crafting of rules
is never finished. New technologies, such as performance-enhancing drugs, gear and monitoring
technology, affect the game. This, in turn, leads to adjustments to existing rules and the introduction
of new rules.

14.9 Make yourself think
1. What rule changes have you observed in your favorite sport?
2. What rule changes would you suggest for your favorite sport?
14.10 References 155

14.10 References
Ahn, T., Janssen, M., Reinders, D., & Stake, J. (2005). Learning to play nice: strategy evolu-
tion in the National Hockey League. unpublished mimeo, (http://www.public.asu.edu/ ma-
jansse/pubs/learningtoplaynice1.pdf).
Allen, W. (2002). Crime, punishment, and recidivism: lessons from the national hockey league.
Journal of Sport Economics, 3(1), 39–60.
Duggan, M., & Levitt, S. (2002). Winning isn’t everything: corruption in sumo wrestling. American
Economic Review, 92(5), 1594–1605.
Heckelman, J., & Yates, A. (2003). And a hockey game broke out: crime and punishment in the nhl.
Economic Inquiry, 41(4), 705–712.
Levitt, S. (2002). Testing the economic model of crime: the national hockey league’s two-referee
experiment. Contributions to Economic Policy and Analysis, 1/1(article 2).
Key Concepts
In this chapter we will:
– See a summary of the lessons learned from this book
– Learn why we still cannot solve all collective action problems
– Get exposed to the big challenges we are still facing in governing the commons

15 — Challenges ahead

15.1 What have we learned?
This book has provided an introduction to the study of institutions and governance in general and
of governance of the commons in particular. Common-pool resources often face the challenge
of overharvesting, while public goods face the problem of under-provision. Despite the difficult
challenges associated with governing these types of goods, we see successful performance of many
common-pool resource systems and successful provision of public goods. We need to extend
the lessons learned from these successes to better understand the general properties of different
approaches to successfully governing the commons.
Elinor Ostrom developed a coherent theoretical framework that enables scholars to clearly
articulate how institutional arrangements can facilitate successful governance of the commons. By
institutions we refer to the prescriptions that humans use to organize all forms of repetitive and
structured interactions. The prescriptions are rules and norms. They apply not only to common-pool
resources such as groundwater, but also to other types of social dilemma situations like traffic,
Wikipedia, sports, and health care.
Rules can be written laws, or agreed upon and commonly understood verbal rules in a community.
Norms do not include explicit consequences if forbidden activities are performed or requirements
are not met. Even though not explicit, not following social norms may have negative consequences
since people may decide to avoid interacting with people who have bad reputations.
A key concept in studying institutions is the action arena. An action arena consists of people
as participants and an action situation in which they participate. When people interact in an action
situation, they make decisions based on the choice rules associated with the position they occupy in
that action situation. In a given action situation, people may hold different positions and therefore
may not be able to make the same decisions, or have the same information. The interactions of the
participants lead to outcomes that can be evaluated.
Figure 15.1 shows the schematic representation of the Institutional Analysis and Development
(IAD) framework and highlights the key components necessary for studying how institutions structure
action situations. The IAD framework emphasizes the fact that action situations are influenced
by broader contextual variables. The biophysical conditions—whether you live in a desert or a
rainforest—affect rules and norms concerning how to build houses and how to organize health care
(e.g., due to different diseases that are prevalent in a given area). The attributes of a community such
as the age and income distribution, education, and kin-relationships, affect which kind of interactions
one can expect in action situations.
158 Challenges ahead

The rules-in-use are one of the key foci of the IAD framework. Rules on paper are important, but
if those rules are not known, understood, and accepted by participants in the action situation, they
will not effectively guide behavior. In studying the governance of the commons, we are interested in
which rules people actually use, how they monitor rule compliance, sanction rule infractions, and
how contextual variables impact how the rules function.

Figure 15.1: A framework for institutional analysis (adapted from Ostrom, 2005).

We have illustrated the application of the framework through several examples. The framework
is just that—a framework. Frameworks are an articulation of key elements that should be considered
when trying to understand the impact of institutional arrangements on human behavior and social
interactions. The framework provides a set of concepts and language that enables scholars to
communicate effectively about the key working parts of an action situation. Thus, if a student has
developed a working knowledge of the IAD framework, they should be able to translate observations
of social phenomena into the language of the IAD Framework and action situations. This process of
translating phenomena into a formal language enables us to compare different cases and uncover
regularities.
It is important to understand that the IAD framework is not a theory. It does not suggest a
hypothesis about how different parts of action situations relate to outcomes. Theory relevant to
understanding social phenomena is an additional layer related to how people make decisions in
different action situations. We discussed one theory of human decision making called rational choice
theory. We then discussed experimental work that shows that this theory of behavior holds only
in certain circumstances. Human decisions take more into account than narrow self interest. We
illustrated the importance of this fact through experiments of social dilemmas.
Much of the discussion and the majority of examples focused on a particular context that we
call the commons. We study the commons because many examples of the most difficult problems
we face today are commons that involve social dilemmas. In those situations there are incentives
for individuals to free ride on the cooperative actions of others. We experience social dilemmas in
many action situations in our daily lives. For example, who is doing all the work in a group project,
how do we pay for the highways we use, how do we make sure there is health care available when
we need it, who writes the articles on Wikipedia, are our bridges being inspected for safety, who
reduces their energy use to help reduce pollution?
How do we organize incentives such that we reduce free riding in problems associated with
issues we care about? One option is to use coercion. If people have a tendency to free ride on the
15.1 What have we learned? 159

cooperative behavior of others, then privatization of common-pool resources and public goods is
an option. The reasoning is that individuals will make better decisions regarding the use of private
goods. We illustrated the problems associated with common-pool resources with an example in
which multiple people share a meadow. Each individual has an incentive to add animals to the
meadow and when everyone in the group does so, this will lead to overgrazing. If, on the other hand,
everybody owns their own a part of the meadow, everybody will take care of his own property and
won’t damage others’ property. Another policy might be to tax the use of resources, so that people
will not overuse common-pool resources.
Both of these economic instruments (privatization and taxation) are used in managing the
commons in practice. However, these instruments face several practical limitations and are not the
only options available. There are many examples of self-governance, meaning that the users and
producers of the commons are crafting, implementing and maintaining the institutional arrangements
themselves. Based on these institutional arrangements, communities can successfully govern the
commons without privatization or taxation from an outside governmental body.
The challenge that the tools developed in this book are meant to address is to understand what
kind of institutional arrangements are successful in which circumstances. A coercive approach is
not necessarily a productive approach. Coercion may demotivate participants. Providing monetary
incentives may also not always be beneficial. An illustrative example is a study by Gneezy and
Rustichini (2000) on daycare centers. Parents often come late to collect their kids from daycare. To
reduce the number of people who are late, an experiment was performed that imposed a monetary
penalty when parents were late. Surprisingly, parents came late more often. Why should anyone
complain when they have paid for it? Parents who were willing to pay the price could come late
without feeling guilty. When the daycare centers wanted to revert back to the original situation and
remove the penalty, the number of parents who came late remained high. A behavior that is a moral
obligation (coming on time to collect your child) became an economic transaction (paying a fee).
This is a risk of using economic incentives to stimulate behavioral change; it may have unintended,
difficult-to-reverse consequences.
The study of successful institutional arrangements shows that it is important that participants
in action situations are involved in the creation of rules, that there are low-cost conflict-resolution
mechanisms, and that there are clear rules about who and when people can use the commons.
Effective institutional mechanisms stimulate personal interactions that facilitate trust relationships
and allow participants to build reputations. When people on e-Bay provide feedback based on their
experience in the transaction they just finished, it contributes to information about the reputation
of their economic partners. When people edit the English text of Wikipedia articles, they gain
respect and a good reputation in the community that may enable them to occupy a special role in
the community. When a tennis player, who just lost a match, shakes the hand of the opponent, it
reinforces the respectful relationship they have with each other.
The emerging picture of effective institutional arrangements is that in order to have success,
it is important that people can develop trust relationships, gain a reputation, experiment with new
arrangements, tolerate mistakes people make, and have commonly understood rules-in-use. Most of
these insights have been derived from studies of communities who share common-pool resources. If
we know so much about successes, why are there still so many problems?
160 Challenges ahead

15.2 Why are there still so many problems in governance?
In this book we have discussed insights relating to the ability of communities to solve collective
action problems. If we know so much about what leads to effective institutional arrangements, why
are there still so many problems? What prevents us from sustaining the commons?
More than a billion people around the world do not have sanitation or access to clean water. Many
species go extinct each year and human activities cause long-term disruption to biogeochemical
cycles in nature. Many of us waste hours each week in traffic jams, download illegal music files, and
complain about the performance of elected officials.
Knowing what leads to better institutional arrangements will not solve all these problems.
What are the main challenges? What are the open questions in our understanding of institutional
arrangements that require further research? In the following paragraphs, we attempt to list some of
the most important challenges. This list is not exhaustive but, rather, represents only a starting point.
One of the big challenges in our modern society is the scale of the problems we face. We are no
longer living in small communities where we know exactly what everybody is doing. We may not
even know who our neighbors are. In an increasingly urbanized world, we interact with many people
who are strangers to us. Even so, there is still an incredible level of cooperation in most modern
economies. A moment’s reflection should give the reader a sense of astonishment at the fact that
hundreds of millions of people can effectively coordinate their behavior every day. How do we do
this? Institutions are a big part of the story. We are able to signal to each other our reputation and
trustworthiness because of the uniforms we wear (in the position of police officer, you must wear an
official uniform), the tattoos we have, the certificates we have earned (positions defined by boundary
rules) and the gossip that is spreading about us. It is not uncommon for us to give a stranger our
credit card (backed by an enormous stock of institutions and organizations) to make a payment. We
are accustomed to conditionally trusting strangers.
Nevertheless, the larger scale of our interaction spheres increases the possibility that we may
lack the appropriate information to make good decisions. Think about people accepting the terms of
home loans that they could not understand. Think about large institutional investors who purchase
investments for which they cannot assess the risks. The financial crisis of 2008 demonstrated just
how calamitous and how much suffering such information failures may generate. There is also
the possibility of misunderstandings about our intensions, motivations and the meaning of rules.
An important condition of well-functioning institutional arrangements is that rules are commonly
understood.
Being in larger groups makes it more difficult for individuals to be involved in rule crafting. In
the position of a U.S. citizen of eighteen years or older, you may vote but you may also feel that
your vote is insignificant. You may not be able to have an impact on the outcomes at the national
level, but you still can participate in local governance issues whether this is through an elected office,
community service project, or a volunteer activity for your children’s school. Individual actions in
the community add up. Because the impact of such activities is difficult to measure, the incentives to
take them are weak. This is one of the fundamental problems of society—the under-provision of
public goods.
Further, larger groups will make it easier to be invisible as a free-rider. You can be one of the
many who do not volunteer. Larger groups make it more likely that there are different opinions and
more disagreement among the participants. Disagreement makes it easier not to act, even though we
know we should.
How can we stimulate cooperation in large populations? Can we apply the insights from this
15.2 Why are there still so many problems in governance? 161

book to an urbanized and globalized world? New technologies may provide solutions. Many of us
have a mobile phone with us; a small computer that can register where we are and can be used to
make photos and exchange information with friends in social networks. Can we use these devices
to improve the information we have about each other in order to improve trust in relationships and
monitor the actions of each other? How we may be able use the crowd to govern the crowd is an
important, open question.
Another big challenge is that new problems always emerge. With every new technology there
are benefits but there also come new problems. There was no cyber bullying before the Internet.
It is more difficult to bully someone in person than virtually. There was no illegal downloading
before digital recording. To illegally obtain a music recording 40 years ago, it was necessary
to walk into a record store and walk out with a vinyl disc! Again, before the Internet, stealing
was a more personal affair—you had to actually see the victim. Now it has become impersonal.
New problems also emerge due to new insights from science. Improved technology allows better
measurements and enables new discoveries, such as the emergence of the hole in the ozone layer. Our
understanding of chlorofluorocarbons enabled us to determine that they were responsible. Reducing
chlorofluorocarbons was fairly easy—the problem was clear, measurable, and well understood.
The solution was technologically feasible and economical. This is in stark contrast to climate
change, which poses a much more difficult collective action problem. If and when we develop
global governance arrangements to deal with climate change, what will be the next problem to
emerge? Will human society ever have enough time to solve its existing set of social dilemmas
before being presented with another new problem? Or, put in another way, will humans ever learn to
craft institutions and governance structures fast enough to address new challenges?
History suggests there are some reasons to be hopeful—e.g., the Montreal Protocol, which deals
with chlorofluorocarbons—but the challenges are many. Globalization will bring with it global scale
problems. These will require global-level solutions. This will require cooperation between people
from many different cultures. New mixtures of populations may require generations to develop
commonly understood well-functioning regulations, slowing our capacity to respond. Further,
because solving social problems is difficult and complex, people tend to stick with institutions that
have worked in the past. The No Child Left Behind program is a good example. The tried and true
solutions (based on the Protestant work ethic) of trying to create incentives for more discipline and
harder work through higher standards and more measurement simply does not work for a public
good like education today. Why? The social context is completely different and “education” is
complex. In order to learn more material more quickly as No Child Left Behind demands, children
need mentoring. In the past when parents had the time to mentor, No Child Left Behind may have
been a great success (at least by its own measure of improved standardized test scores). At present,
when in many households both parents work and have little time or energy to mentor their children,
higher standards and more testing will have little effect. Old solutions do not translate well to new
situations and simple panaceas will fail. Rather, we must perform small scale experiments to get
experience with new institutional arrangements in new contexts. Because such experiments are
costly and require patience, developing effective institutions will require considerable collective will
on the part of society.
The third challenge we face is that it is often not in everybody’s interest to solve a problem.
Different people have different positions and interests. A problem for one participant can be an
opportunity for another. Hence not everybody has an incentive to solve a problem. Problems don’t
exist in a vacuum, there is already a social and ecological context for every problem we face. If the
poor and unemployed don’t receive health care, it is not a direct benefit for those who have health
162 Challenges ahead

care to pay for and share their health care benefits. The status quo, although not perfect, might be
beneficial to many participants as compared to an alternative.
Finally, sometimes constitutional choice rules make it difficult to change a regulation. The
European Union now consists of more than 25 nations. The EU employs an aggregation rule by
which decisions are made by a unanimity vote. In a unanimity vote everybody needs to vote in favor
in order for a proposal to be accepted. If the group is relatively small and people are sufficiently
aligned in terms of their understanding and preferences, this will work. But in large groups, one
individual country can take negotiations hostage to receive benefits for voting in favor of a motion.

15.3 Closing
In closing we can say that there has been significant developments over the past 50 years in our
understanding of institutional arrangements and the way they structure social interactions. This book
provides ways to study and analyze institutions. After reading this book, we hope you will view the
problems we face everyday and the very diverse ways we are solving collective action problems
through a new lens and in a different light.
Different disciplines contribute to our understanding of human behavior in the context of complex
social and ecological systems. Unfortunately we cannot provide a blueprint for how to solve all the
problems we experience. Experimentation at the small scale and finding mechanisms to connect
successful solutions to larger scales are key. Although we cannot provide simple solutions to
complex problems, we have provided you a powerful set of tools to make more informed decisions
and recognize the importance of your own role in society.

15.4 Critical reflections
The rules and norms that govern human interactions can be studied with the framework that is
presented in this book. The framework can be applied to many different topics including sustainability,
health care, sports, education and the digital commons. Despite our increasing understanding of
institutions and lessons regarding the conditions of successful collective action, there are still many
failures.
Major challenges exist in governance in modern society since the scale at which we interact with
others is much larger than it has ever been in human history. This makes lessons about success from
studies of small groups difficult to apply. Furthermore, we experience misunderstandings if we don’t
speak the same language or live in different social and ecological contexts. Finally, new problems
are constantly emerging due to rapid environmental and technological change.

15.5 Make yourself think
1. How can you make a difference in addressing major problems in society?
2. Ask older family members how they made arrangements for going out (in a time before mobile
phones and texting). Do you see changes in rules and norms?
3. What do you see as the most challenging topic of governance in the future?

15.6 References
Gneezy, U., & Rustichini, A. (2000). A fine is a price. Journal of Legal Studies, 29(1), 1–17.
Credits

Cover Photo: Licensed from Shutterstock.
Chapter Header Images: Licensed from Shutterstock.
Chapters 10 & 11: Copyright: Iurii Osadchi

Images appear courtesy of the owner, except in cases where author is unknown (after extensive
research) or permission was repeatedly requested without response. If you believe a copyright has
been violated, please contact CSID at csid@asu.edu.

Chapter 1
Figure 1.1: Photo credit Mike Willis. Licensed under Creative Commons.
Figure 1.2: Accessed on May 6, 2013 from the Garret Hardin Society Website at http://www.
garretthardinsociety.org/gh/gh_pictures.html. Photographer unknown.
Figure 1.3. Relative biomass estimates from the beginning of industrialized fishing (Myers, R.A. and
B. Worm (2003) Rapid worldwide depletion of predatory fish communities, Nature 423: 280-283.)
Figure 1.4: First published in William R. Shepherd, Historical Atlas, New York, Henry Holt and
Company, 1923. Public domain.
Figure 1.5: Reprinted courtesy of Indiana University.

Chapter 2
Figure 2.1: Image licensed from Shutterstock.
Figure 2.2: Image licensed from Shutterstock.
Figure 2.3: Photo credit the Vincent and Elinor Ostrom Workshop in Political Theory and Policy
Analysis.
Figure 2.4: Adapted from Ostrom, E. (2005) Understanding Institutional Diversity, Princeton Uni-
versity Press.
Figure 2.5: Adapted from Ostrom, E. (2005) Understanding Institutional Diversity, Princeton Uni-
versity Press.
Figure 2.6: Original to this manuscript.
164 Challenges ahead

Chapter 3
Figure 3.1: Adapted from Ostrom, E. (2005) Understanding Institutional Diversity, Princeton Uni-
versity Press.
Figure 3.2: Photographer unknown.
Figure 3.3: Image licensed from Shutterstock.
Figure 3.4a: Photo credit Eckhard Pecher. Licensed under Creative Commons.
Figure 3.4b: Photo credit Steiner. Licensed under Creative Commons.
Figure 3.4c: Image licensed from Shutterstock.
Figure 3.4d: Image licensed from Shutterstock.
Figure 3.5a: Photo credit Pete Souza. Licensed under Creative Commons.
Figure 3.5b: Photo credit Steve Petteway, Collection of the Supreme Court of the United States. This
work is in the public domain.
Figure 3.5c: Photo owned by American Idol, Inc. Permission requested.
Figure 3.5d: Photo credit United States Department of Interior (photographer unknown). This work
is in the public domain.
Figure 3.6: Photo credit United States Coast Guard (photographer unknown). This work is in the
public domain.
Figure 3.7: Image licensed from Shutterstock.
Figure 3.8: Image licensed from Shutterstock.
Figure 3.9a: Photo used courtesy of TrueStory Films.
Figure 3.9b: Photo used courtesy of Ampersand Advisory Group.
Figure 3.9c: Photo used courtesy BWOG at Columbia University.
Figure 3.9d: Photo used courtesy of Arizona State Board of Regents.

Chapter 4
Figure 4.1: Cartoon used with permission from Cartoon Bank.
Table 4.1: Original to this manuscript.
Table 4.2: Original to this manuscript.
Figure 4.2: Wikipedia Logo owned by the Wikimedia Foundation. Permission requested.
Figure 4.3: Image credit Benjamin D. Esham. Licensed under Creative Commons.
Table 4.3: Original to this manuscript.
Figure 4.4: Photo credit Andrew*. Licensed under Creative Commons.
Table 4.4: Original to this manuscript.
Figure 4.5: Photo credit Midloa. Licensed under Creative Commons.
Table 4.5: Original to this manuscript.
Figure 4.6: Cartoon owned by Kirk Anderson. Permission requested.

Chapter 5
Figure 5.1: Data first appeared in United Nations World Water Development Report 3 available
at http://www.unesco.org/new/en/natural-sciences/environment/water/wwap/wwdr/wwdr3-2009/
downloads-wwdr3/.
Figure 5.2: Data first appeared in Natural Disaster Hotspots: A Global Risk Analysis available at
15.6 References 165

http://go.worldbank.org/G1IE72BKU0.
Figure 5.3: Global Map of Irrigation Areas version 4.0.1. Johann Wolfgang Goethe University,
Frankfurt am Main, Germany / Food and Agriculture Organization of the United Nations, Rome,
Italy.
Figure 5.4: Photo used courtesy of About.com.
Figure 5.5: Photo credit U.S. Department of Interior Reclamation. This work is in the public domain.
Figure 5.6: Map owned by NedWater. Permission requested.
Figure 5.7: Photo credit Lucas Hirschegger. Licensed under Creative Commons.
Figure 5.8: Photo credit Sono Adriano. Licensed under Creative Commons.
Figure 5.9: Image created by National Oceanic and Atmospheric Administration. This work is in the
public domain.
Figure 5.10: Photo used courtesy of N. Rabalais, LUMCON.

Chapter 6
Figure 6.1: Photo used courtesy of Torbel Moosalp Tourimus.
Figure 6.2: Photo used courtesy of Gita Mehra.
Figure 6.3: Patten White, a fisherman out of York, Maine. Photo used courtesy of Gulf of Maine
Lobster Foundation.
Figure 6.4: Photo used courtesy of the AVC Lobster Science Centre.
Figure 6.5: Graphs created based on data from Maine.gov “Historical Summary of Maine Lobster
Fisheries.” http://www.maine.gov/dmr/rm/lobster/lobdata.htm.
Figure 6.6: Photo used courtesy Sajid Pareeth.

Chapter 7
Figure 7.1: Photo used courtesy of Rhett Butler.
Figure 7.2: Image used courtesy of Hawk Eye Innovations.

Chapter 8
Figure 8.1: Photo owned by LMU Experimental Economics Laboratory (LEEL), Department of Eco-
nomics, Bellarmine College of Liberal Arts, Loyola Marymount University. Permission requested.
Figure 8.2: Photo owned by Team Building and Events. Permission requested.
Figure 8.3: Original to this manuscript.
Figure 8.4: Original to this manuscript.
Figure 8.5: Original to this manuscript.
Figure 8.6: Photo used courtesy Dr. Joseph Henrich.
Figure 8.7: Adapted from Henrich, J., R. Boyd, S. Bowles, C. Camerer, E. Fehr, H. Gintis, and
R. McElreath (2001) In Search of Homo Economicus: Behavioral Experiments in 15 Small-Scale
Societies, American Economic Review. 91(2) 73-78.
Figure 8.8: Henrich, J., J. Ensminger, R. McElreath, A. Barr, C. Barrett, A. Bolyanatz, JC Cardenas,
M. Gurven, E. Gwako, N. Henrich, C. Lesorogol, F. Marlowe, D. Tracer, and J. Ziker (2010) Markets,
166 Challenges ahead

Religion, Community Size, and the Evolution of Fairness and Punishment , Science 327: 1480-1484

Chapter 9
Figure 9.1: Original to this manuscript.
Figure 9.2: Original to this manuscript.
Figure 9.3: Original to this manuscript.
Figure 9.4: Photo used courtesy of Daniel Castillo.
Figure 9.5: Original to this manuscript.
Figure 9.6: Original to this manuscript.
Figure 9.7: “Time trend of mean cooperation.” Fehr, E. and S. Gächter (2002) Altruistic punishment
in humans, Nature 415: 137-140.

Chapter 10
Figure 10.1: Adapted from Ostrom, E. (2005) Understanding Institutional Diversity, Princeton
University Press.
Figure 10.2: Image owned by eBay Corporation. Permission requested.
Figure 10.3: Data collected by the U.S. Office of Personnel Management.
Figure 10.4. Photo credit Skubik. Licensed under Creative Commons.

Chapter 11
Figure 11.1: Photo used courtesy of Mark Wilson.
Figure 11.2: Photographer unknown.
Figure 11.3: South Park episode still used courtesy of Viacom, Inc.

Chapter 12
Figure 12.1: Image owned by Emergency Management and Safety Solutions. Permission requested.
Table 12.1: 2011 Health statistics of World Health Organization.
Figure 12.2: Johnson, E.J., D. Goldstein (2003) Do defaults save lives, Science 302: 1338-1339.
Figure 12.3: Adapted from Wagner EG & Lanoix JN (1958). Excreta Disposal for Rural Areas and
Small Communities, WHO Monograph series No 39, WHO, Geneva. WHO.
Figure 12.4: Photo owned by Water Journalists Africa. Permission requested.

Chapter 13
Figure 13.1: Photo appears on Served Fresh Media website (http://servedfreshmedia.net). Permission
requested.
Figure 13.2: Wikipedia Logo owned by the Wikimedia Foundation. Permission requested.
Figure 13.3: Image credit HenkvD. Licensed under Creative Commons.
Figure 13.4: Image credit HenkvD. Licensed under Creative Commons.
Figure 13.5: Image of screen shot of Wikipedia page is licensed under Creative Commons.
Figure 13.6: Photo credit Guillaume Paumier, CC-BY. Licensed under Creative Commons.
Figure 13.7: Napster Logo owned by Napster, Inc. Permission requested.

Chapter 14
Figure 14.1: Photo credit Eckhard Pecher. Licensed under Creative Commons.
Figure 14.2: Image used courtesy Seiko Holdings Corporation.
Figure 14.3: Photo credit Cpl. Megan L. Stiner, United States Marine Corp. This work is in the
public domain.
Figure 14.4: Created by Chandler. Released to public domain by author.
Figure 14.5: Photo owned by Associated Press. Permission pending.
Figure 14.6: Photo credit Keith Allison. Licensed under Creative Commons.
Figure 14.7: Photo credit Chell Hill. Licensed under Creative Commons.
Figure 14.8: Photo credit Jim Thurston. Licensed under Creative Commons.
Figure 14.9: Photo credit Cameron Richardson, The Daily Telegraph. Permission requested.

Chapter 15
Figure 15.1: Adapted from Ostrom, E. (2005) Understanding Institutional Diversity, Princeton
University Press.