Ecological Economics 131 (2017) 312–321

Contents lists available at ScienceDirect

Ecological Economics

journal homepage: www.elsevier.com/locate/ecolecon

Local rulemaking, enforcement and compliance in state-owned

forest commons
Graham Epstein
The Vincent and Elinor Ostrom Workshop in Political Theory and Policy Analysis, 513 North Park Ave., Bloomington, IN 47408, USA
Environmental Change and Governance Group, School of Environment, Resources and Sustainability, University of Waterloo, 200 University Avenue West, Waterloo, ON N2L 3G1, Canada

a r t i c l e i n f o a b s t r a c t

Article history: The literature on rule compliance is divided between those urging greater autonomy for stakeholders in
Received 6 January 2016 rulemaking processes; and those arguing for increased enforcement. However recent experimental evidence
Received in revised form 9 September 2016 highlights the potential for synergies between participatory rulemaking and enforcement. This paper therefore
Accepted 12 September 2016
seeks to build upon these findings to explore the relationship between local rulemaking, local monitoring and
Available online 28 September 2016
compliance in field settings. The results which draw upon data about the behavior of 93 fuelwood user groups
Keywords:
in state-owned forest commons in Asia, Africa and Latin America suggest that the average group is more likely
Rule compliance to comply with rules when local rulemaking is combined with local monitoring. However, in some contexts it ap-
Forest commons pears that local rulemaking in particular and other institutional arrangements in general may yield similar
Intrinsic motivation results.
Participation © 2016 Elsevier B.V. All rights reserved.
Monitoring
Enforcement

1. Introduction
Rule compliance is an important, if not essential, condition for the sus-
tainable governance of social-ecological systems (SESs) (Young, 1979;
Zaelke et al., 2005). Indeed illegal harvesting of natural resources is
often identified as a leading cause of environmental degradation
(Contreras-Hermosilla, 2002; World Bank, 2004; Fromentin and Powers,
2005; Agnew et al., 2009); and thus a better understanding of the ways
in which policies might be designed to encourage compliance is needed
to confront the mounting environmental problems that human societies
face. Fortunately by synthesizing empirical studies of rule compliance
(Kuperan and Sutinen, 1999; Nielsen, 2003; Ramcilovic-Suominen
and Epstein, 2012) and common-pool resource (CPR) management
(Ostrom, 1990; Cox et al., 2010) scholars have identified a number of in-
stitutional conditions that appear to contribute to compliance and cooper-
ation. Most notably a number of studies suggest that prospects for
compliance and cooperation tend to increase when stakeholders are
(i) able to participate in rulemaking processes to influence the content
of rules that affect them; and (ii) organize monitoring and sanctioning
in order to enforce those rules. However, important debates remain
with regards to the relative priority of these factors and the ways in
which they might interact to influence behavior. While some have argued
for the primacy, if not sufficiency of enforcement as a strategy to generate
incentives for actors to follow rules (Becker, 1968; Hardin, 1968); others
have suggested that participatory processes can lead to the expression of
intrinsic motivation to follow rules only to be negated by external
E-mail address: graham.epstein@uwaterloo.ca.
interventions such as enforcement (Frey, 1997; Ryan and Deci, 2000;
Bowles, 2008).
A recent experimental study however contradicted predictions re-
garding the superiority of either participatory rulemaking or enforcement
by using a factorial design to highlight the synergistic effects of the two
approaches (DeCaro et al., 2015). Indeed, groups subject to a combined
voting and enforcement treatment exhibited lower rates of resource de-
pletion, higher average individual returns and higher levels of compliance
than groups who were subject to voting or enforcement alone. Therefore
this study seeks to build upon these findings by considering the relation-
ship between local rulemaking, local monitoring and compliance in the
context of fuelwood appropriation in state-owned forest commons
using data collected as part of the International Forestry Resources and In-
stitutions (IFRI) program (Wertime et al., 2007). The results suggest that
user groups are most likely to exhibit high levels of compliance with rules
for fuelwood appropriation when local rulemaking is combined with local
monitoring, while controlling for a number of potentially intervening fac-
tors; providing support for the findings presented in DeCaro et al. (2015).
The remainder of this chapter is structured in the following way.
Section 1.1 briefly outlines the concept of compliance and the develop-
ment of compliance theory from its origins in economics, to the more
socialized and political models of the present day. It then continues by
emphasizing the potential role of local rulemaking and local monitor-
ing; before turning to the question of when local rulemaking might con-
tribute to compliance. Section 2 presents the approach taken in this
paper, and the results are reported in Section 3. Finally, Sections 4 and
5 conclude this paper by discussing the limitations of this research
and implications for compliance theory.

http://dx.doi.org/10.1016/j.ecolecon.2016.09.012
0921-8009/© 2016 Elsevier B.V. All rights reserved.
 G. Epstein / Ecological Economics 131 (2017) 312–321
1.1. Rule Compliance and Human Motivation
Rule compliance refers to the behavior of actors relative to behav-
ioral prescriptions (Young, 1979). The concept of compliance is sim-
ilar, but not identical to the concept of cooperation used in CPR
experiments. They are similar in that the short-term economic in-
centives tend to favor overharvesting (or rule violations); but they
differ in that compliance, as opposed to cooperation does not neces-
sarily lead to better outcomes for the participants.1 For instance, if
governments ignore scientific recommendations when determining
quotas (MacKenzie et al., 2009) or policymakers lack sufficient
knowledge to develop effective rules (Romme and Don, 1989;
Acheson, 2006); then even high levels of compliance may lead to
highly suboptimal outcomes. Nonetheless, in field settings where it
is difficult, if not impossible, to estimate social optima the decision
to comply with or violate rules is perhaps the most important and
certainly the most proximate behavioral link to the sustainability of
resources.
Contemporary compliance theory is informed by behavioral
models from across the social sciences with contributions from eco-
nomics, social psychology, political science and sociology. While
these approaches vary with respect to the terminology they use to
describe human motivation and the assumptions they make about
how information is processed to make decisions; perhaps their
most significant differences rests in the extent to which priority is
given to extrinsic versus intrinsic sources of motivation. Classically
trained economists such as Becker (1968) give clear priority to ex-
trinsic motivation in arguing that crime is simply the result of a situ-
ation in which the expected net economic benefits of illegal activity
exceed the expected sanctions. Motivation in this context is consid-
ered extrinsic because of the emphasis on using externally imposed
punishments in order to compel individuals to obey commands
(Deci and Ryan, 2000).
Most social scientists concede that extrinsic factors can have an
influence on behavior and compliance. However, there is also a
growing consensus that human beings exhibit higher levels of coop-
eration and compliance than would be expected based upon extrin-
sic factors alone (Andreoni et al., 1998; King and Sutinen, 2010). As
a result a number of social science disciplines now argue that the
problem of securing cooperation and compliance rests, at least in
part, on the ability of governance systems to promote acceptance of
institutions and intrinsic motivation (Frey, 1997; Bowles, 2008;
DeCaro and Stokes, 2013). Institutional acceptance refers to the ex-
tent to which actors accept rules as legitimate often because they
are aligned with their interests and values (DeCaro and Stokes,
2013). Intrinsic motivation, meanwhile, refers to the extent to
which individuals personally endorse or internalize a behavior (i.e.
compliance with a rule) such that they behave in a particular way
even in the absence of observation (Kerr et al., 1997). In general it
has been argued that institutional acceptance and intrinsic motiva-
tion can be promoted by developing governance systems that satisfy
innate needs for self-determination (Ryan and Deci, 2000; Frey and
Jegen, 2001), procedural justice (Kuperan and Sutinen, 1999;
Nielsen, 2003; Tyler, 2003, 2006), and by providing a sense of securi-
ty and predictability for the actors involved (Shinada and Yamagishi,
2007; Ostrom, 2009; DeCaro et al., 2015).
Contemporary compliance theory has therefore developed to recog-
nize the potential for promoting compliance using both extrinsic and in-
trinsic sources of motivation. Scholars emphasizing extrinsic motivation
tend to promote investments in monitoring and sanctioning systems in
order to increase the perceived costs of rule violation (Becker, 1968).
1
Although cooperation will tend to lead to “better outcomes” for actors that hold rights
to harvest a common-pool resource, communities and societies might be “worse off” in ag-
gregate if some actors are excluded from using a common-pool resource.
313
Although there are some exceptions, most field studies provide at
least partial support for the hypothesis that rule violations decline
with increased monitoring and enforcement2 (Kuperan and Sutinen,
1998; Scholz and Lubell, 1998; Hatcher et al., 2000; Viteri and Chávez,
2007; Ramcilovic-Suominen and Epstein, 2015). In contrast, scholars fo-
cusing on intrinsic motivation tend to encourage the development of
participatory models of decision-making that appeal to the moral senti-
ments and social-psychological needs of human beings (Bowles, 2008;
DeCaro and Stokes, 2013). Much like enforcement, field studies of rule
compliance offer fairly broad, but not complete support for the hypoth-
esis that stakeholder participation3 in decision-making contributes to
increased compliance (Frey, 1997; Scholz and Lubell, 1998; Hatcher
et al., 2000; Viteri and Chávez, 2007; Madrigal-Ballestero et al., 2013).
There is now fairly widespread agreement among compliance
scholars that participatory rulemaking processes and/or monitoring
and sanctioning can be used as strategies to address compliance prob-
lems (Kuperan and Sutinen, 1999; Nielsen, 2003). Important debates
have, however, emerged concerning the relationship between partici-
patory rulemaking and enforcement and how they interact to influence
behavior. First, there is a large and growing body of research demon-
strating that the introduction of external monetary rewards or punish-
ments can in some cases crowd-out intrinsic motivation, often
undermining the very goals they are meant to achieve4 (Gneezy and
Rustichini, 2000a,b; Ryan and Deci, 2000; Bowles, 2008). Experimental
studies of CPR dilemmas have, for instance, found that the introduction
of sanctioning mechanisms often leads to worse or at least similar
outcomes as treatments that simply allow participants to communi-
cate (Ostrom et al., 1992; Cardenas et al., 2000; Cardenas, 2004;
Janssen et al., 2010); indicating the potential for tradeoffs between
motivation generated by participatory decision-making and enforce-
ment (DeCaro et al., 2015). However, others have argued that some
types of external interventions such as monitoring and sanctioning
might actually enhance intrinsic motivation provided that it is per-
ceived as supporting the self-determination of actors (Ostrom,
2000; Frey and Jegen, 2001; DeCaro et al., 2015); and providing a
measure of security against exploitation by free-riders (Shinada
and Yamagishi, 2007; Ostrom, 2009).
Communities and other groups of individuals rarely correspond to a
single narrow model of human motivation (Ostrom, 1998, 2000). In-
stead human beings exhibit considerable heterogeneity in terms of
their propensity to cooperate with others. However, a large fraction of
most groups can be described as conditional cooperators5 who have a
general tendency to respond in-kind to the behavior of other group
members (Fischbacher et al., 2001; Frey and Meier, 2004; Kocher
et al., 2008). As a consequence although conditional cooperators
might be intrinsically motivated to comply as a result of participatory
rulemaking; their compliance behavior is likely to depend crucially on
2
Specific measures used to operationalize this concept include self-reported subjective
probabilities of detection (Kuperan and Sutinen, 1998; Hatcher et al., 2000; Eggert and
Lokina, 2010; Ramcilovic-Suominen and Epstein, 2015), observation of enforcement officers
(Viteri and Chávez, 2007), and opportunity to cheat (Scholz and Lubell, 1998) which is an es-
timate of the ability of individuals to avoid detection for certain line items on tax returns.
3
Specific measures used to operationalize this concept vary across studies. Frey (1997)
consider the constitutional design of Swiss cantons (i.e. referenda, citizen initiatives, citi-
zen meetings) to develop a measure of opportunities for political participation. Viteri
and Chávez (2007) and Jenny et al. (2007) use a direct measure of participation in fisheries
cooperatives and rulemaking for a shared energy system. The remaining studies focus on
subjective measures of political efficacy by asking whether actors feel that their views are
represented or considered in the design of policies.
4
As an example The paper by Gneezy and Rustichini, 2000b is entitled “Pay enough or
don't pay at all” which effectively summarizes the main findings of their research. They
found that groups offered small monetary rewards for performance on IQ tests and collec-
tion of donations performed worse than students offered no compensation at all, and that
groups offered larger rewards tended to outperform those offered smaller rewards.
5
Conditional cooperators are typically defined as individuals who monotonically increase
their contributions in response to the contributions of others. Although contributions in-
crease it is worth noting that most studies have found that the contributions of conditional
cooperators fall short of matching the contributions of others (Kocher et al., 2008).
314 G. Epstein / Ecological Economics 131 (2017) 312–321
whether governance systems provide them with the sense of security
that is necessary (at least in theory) for their cooperation. In theory a
group composed entirely of conditional cooperators would be able to
sustain compliance indefinitely once initiated. However, most (if not
all) groups also include a non-trivial fraction of free-riders that never
voluntarily cooperate, regardless of the cooperation of others. Monitor-
ing and sanctioning mechanisms might therefore serve at least two im-
portant roles to support compliance in heterogeneous groups (Shinada
and Yamagishi, 2007). First, for free-riders it provides a direct disincen-
tive against rule violation that can compel obedience with rules provid-
ed that monitoring is effective and sanctions are sufficiently high (or
graduated). However, the existence of mechanisms to monitor, and if
necessary punish, free-riders also provides conditional cooperators
with a measure of security necessary for their cooperation (Shinada
and Yamagishi, 2007).
Although scholars have speculated about the ways in which partici-
patory rulemaking and enforcement might interact to influence motiva-
tion and behavior, these hypotheses were not explicitly tested until
recent experimental work by DeCaro et al. (2015). Their research
showed that groups exhibited higher levels of compliance in treatments
that combined voting with enforcement than voting or enforcement
alone. They also found that participants in the combined voting and en-
forcement treatment reported higher levels of procedural justice, self-
determination and security than participants in voting or enforcement
treatments; highlighting the contributions of intrinsic motivation as a
key source of incentives for compliance. Their research therefore offers
a significant advance in our understanding of human motivation, com-
pliance and the ways in which policies might be designed to improve
the management of natural resources. However despite their ability to
make strong internally valid causal inferences; the external validity of
experiments is often questioned for neglecting the complexity of field
settings (Poteete et al., 2010). This is particularly salient with regards
to the relationship between enforcement and behavior where field ex-
periments have found that enforcement might crowd-out incentives
for cooperation in groups with high levels of underlying trust (Vollan,
2008). Therefore, this study seeks to assess whether the synergistic ef-
fects of participatory rulemaking and enforcement observed in the lab
extends to the field while controlling for potentially intervening factors.
2. Data and Methods
The data used in this study is drawn from the International Forest
Resources and Institutions (IFRI) database (Wertime et al., 2007). IFRI
is an interdisciplinary and collaborative program that has to date col-
lected data from over 400 forests and 600 forest user groups on four
continents. The database broadly includes information about a forest,
the people that use it, and the rules that groups use to govern their in-
teractions with a forest and each other. Given its breadth IFRI is one of
a small set of large-n datasets of the commons with which scholars
can develop, test and refine hypotheses concerning small-scale environ-
mental governance (Poteete et al., 2010), although it is generally accept-
ed that the cases do not represent a random sample (Chhatre and
Agrawal, 2009; Coleman, 2009).
The sample used in this study includes unique observations of user
groups where fuelwood resources are one of the three most important
products harvested and used by that group in state-owned forests. A
user group is defined as a group of individuals that use a forest and
share similar sets of rights and responsibilities with regards to that for-
est and its resources (Wertime et al., 2007). For instance if one set of in-
dividuals is granted rights to appropriate forest resources such as
firewood, while another set holds both appropriation and management
rights then these would be coded as two distinct user groups. State-
owned forests were chosen as the vast majority of the world's forests
(and forests in the IFRI database) are owned by governments
(Agrawal et al., 2008), and because all user groups in community-
owned forests have de-facto management rights. The sample consists
of 93 unique observations,6 from 8 countries in Africa, Asia, and Latin
America between 1993 and 2008.
2.1. Variables
The variables included in this analysis are described and summarized
in Table 1. First the dependent variable, compliance, is a dichotomous var-
iable which distinguishes between groups that comply with rules most or
nearly all of the time (1), and those that comply half of the time or less (0).
The primary goal of this paper is to test for differences in the likelihood of
high levels of rule compliance under different combinations of local
rulemaking and local monitoring. Local rulemaking is a dichotomous var-
iable which records whether local user groups or earlier generations are
primarily responsible for the content of operational rules for fuelwood re-
sources. Local monitoring is also a dichotomous variable which records
whether members of a user group regularly organize to monitor the use
of forest resources. These variables are effects coded in statistical models
(i.e. local rulemaking =1; no local rulemaking = −1; local
monitoring = 1, no local monitoring = −1) in order to measure the av-
erage impacts of alternative institutional arrangements on the likelihood
of rule compliance. Thirty-five of the ninety-three user groups combine
local rulemaking with local monitoring, while 22 groups have created
rules but do not regularly organize to monitor the use of forest resources
and 8 groups organize to monitor compliance with rules they did not cre-
ate. Finally, the remaining 22 groups are neither primarily responsible for
rules nor do they organize to monitor.
A number of additional variables are also included in statistical models
to account for potentially intervening factors. First, models include a col-
lective action index which is a composite of two ordinal measures that re-
cord the extent to which members of group work together to harvest and
process resources outside of the forest (i.e. agricultural goods), scaled to
range between zero and one. In general levels of cooperative economic
production would be expected to correlate with levels of trust and social
capital in a community (Henrich et al., 2001; Ostrom, 2009); influencing
both the likelihood of self-organization and compliance. Furthermore CPR
field experiments undertaken in Namibia and South Africa suggest that
the effects of monitoring and sanctioning mechanisms might vary de-
pending upon the existing levels of trust within a group (Vollan, 2008).
Poor records the fraction of households in a user group that are con-
sidered poor. Levels of poverty would be expected to directly contribute
to non-compliance (Ehrlich, 1973), and might affect the willingness or
ability of groups to contribute limited resources to managing forest re-
sources. Density is a logarithmic transformation of the number of
households in a user group per hectare of forest. The likelihood of rule
compliance would generally be expected to decline with increased den-
sity as competition for fuelwood resources increases, alongside the costs
of creating and enforcing rules (Olson, 1965). Other group is a dichoto-
mous variable which indicates whether a forest is used by one or more
additional user group. The presence of other groups is likely to reduce
the likelihood of compliance by reducing the per-capita supply of re-
sources, and increasing monitoring costs (Nagendra and Ghate, 2005;
Basurto and Ostrom, 2009). Finally, market is a measure of the distance
to the nearest market in which forest products are sold; and is expected
to correspond to the potential for economic gains from legal and illegal
fuelwood harvesting. The final three variables, commercial users, com-
mercial value and external monitoring are used in the Appendix A to as-
sess the robustness of the results and are discussed in Section 3.1.
A correlation matrix of these variables can be found in Table 4 in the
Appendix A. Although there are a number of potentially interesting
6
310 user groups identified fuelwood as one of three most important forest products
they harvest in the forest. User groups in the United States (18) and forests owned by in-
dividuals or communities (40) were excluded. A further 112 cases were omitted as a result
of listwise deletion, while the remaining 47 cases were excluded as repeat observations of
user groups using different forests (random drops) or the same forest at a different time
(earlier observations were omitted).
 G. Epstein / Ecological Economics 131 (2017) 312–321 315

Table 1
Summary statistics.

Variable Description Obs. Mean Std. Dev. Min Max

Compliance Binary. Group complies with rules most of time or almost always 93 0.699 0.461 0 1
Local rulemaking Binary. Local users are the most important source for rules-in-use 93 0.613 0.490 0 1
Local monitoring Binary. User group regularly monitors use of forest resources. 93 0.462 0.501 0 1
Local rulemaking X Interaction between local rulemaking and local monitoring 93 0.376 0.487 0 1
Local monitoring
Collective action index Composite of two ordinal measures of user group cooperation on harvesting 93 0.158 0.250 0 1
and processing of resources outside of the forest. Re-scaled to vary between zero and one.
Poor Fraction of households in user group considered poor 93 0.188 0.249 0 1
Density Natural logarithm of number of households per hectare of forest 93 −0.688 2.136 −5.825 4.220
Other group Binary. Forest is used by one or more other user groups. 93 0.634 0.484 0 1
Market Distance in kilometers to nearest market 93 5.523 6.457 0 31
Commercial users Fraction of user group that derives commercial benefits from forest 93 0.166 0.322 0 1
Commercial value Ordinal. Foresters assessment of the relative commercial value of the forest 93 1.376 1.031 0 3
External monitoring Binary. External actors monitor the use of forest resources 73 0.877 0.331 0 1

associations among the variables, perhaps the most interesting relate to number of scalar model statistics. Because some forests are found in
local rulemaking and local monitoring. There is a positive and statistical- close proximity to one another, standard errors are clustered by the lo-
ly significant correlation (p b 0.10) between local rulemaking and local cation of forests (oversite in the IFRI database). Model fit was assessed
monitoring, both of which are inversely related to the presence of other using count r-squared values and by visually analyzing residuals.
user groups and the distance to market. External monitoring is nega- Count r-squared records the fraction of observations that are correctly
tively correlated to local rulemaking (p b 0.10), but not local monitoring. assigned by a model, and therefore shows that the model is successful
in classifying approximately 80% of the 93 cases. Standardized Pearson
2.2. Statistical Analysis residuals are presented in Table 2 in the Appendix A. All but one of the
residuals falls within the conventionally accepted range of ± 2.5
A logistic regression is used to estimate the likelihood that members of (Andersson and Gibson, 2007; Coleman, 2009). The only observation
a user group comply with operational rules for the use of fuelwood re- outside of this range is the also the only case in which the combination
sources as a function of a vector of independent variables. Logistic regres- of local rulemaking, local monitoring and low levels of compliance is ob-
sion addresses a number of limitations of the linear probability model, served. This case has experienced a fairly rapid inflow of migrants from
most notably assumptions related to the calculation of standard errors other regions which might pose challenges for promoting institutional
and the linearity of parameters (Long, 1997). The binary logistic model es- acceptance among new arrivals, increase monitoring costs (Agrawal
timates the log odds of an outcome such as high levels of compliance as a and Goyal, 2001) and has in a number of other cases led to the decline
function of a vector of independent variables as shown in Eq. (1). As a re- of community-based governance systems (Acheson, 2006; Basurto
sult, the likelihood of an outcome, depend upon the value of a focal pa- and Ostrom, 2009).
rameter such as local rulemaking, but also the value of all other The results presented in Table 2 provide support for the hypothesis of
variables in the model. a synergistic relationship between local rulemaking and local monitoring.
Table 3 reports the odds that a group with a particular combination of
! local rulemaking and local monitoring will have high rather than low
pcomply
ln ¼ β0 þ β1 x1 þ … þ βk xk : ð1Þ levels of compliance. It shows that groups combining local rulemaking
1−pcomply
with monitoring are approximately 61 times more likely to exhibit high
levels of compliance, while for local rulemaking there are 3 to 1 odds of
high levels of compliance. In contrast the odds are about even when
3. Results groups neither create nor enforce rules, and favor low levels of compli-
ance when groups monitor compliance with externally imposed rules.
The results are presented in Table 2 which reports estimated coeffi- Fig. 1 takes a slightly different approach and plots the predicted
cients, clustered robust standard errors and p-values, in addition to a probability of high levels of compliance with fuelwood appropriation
rules in order to compare the probability of compliance under different
Table 2 institutional arrangements. As can be seen groups that combine local
Results. rulemaking with local monitoring have the highest predicted probabil-
β SE p 90% C.I. ity of compliance (Fig. 1A), a difference which is statistically significant
at the 10% level while holding all other values at their sample mean.
Local rulemaking 1.617 0.467 0.001 0.848 2.385
Local monitoring 0.387 0.484 0.424 -0.409 1.183
Groups who are responsible for both the creation and enforcement of
Local rulemaking X local monitoring 1.154 0.418 0.006 0.467 1.842 rules are approximately 5 times more likely to comply with rules than
groups that monitor compliance with externally imposed rules; and
Controls
Collective action index 1.169 0.385 0.002 0.537 1.802
1.3 times more likely than groups that created rules but do not regularly
Poor −0.292 0.359 0.417 −0.883 0.299 monitor the use of fuelwood resources. However, the remaining plots
Density −0.091 0.304 0.765 −0.591 0.409 (Figure 1BCD) which show how the predicted probability of compliance
Other group −1.080 0.736 0.142 −2.291 0.131
Distance to market 0.616 0.271 0.023 0.170 1.063
Constant 1.641 0.615 0.008 0.630 2.653 Table 3
N 93 Odds ratios.
Count R2 0.796
Arrangements Odds ratio
Chi-square 35.26
p-value 0.000 Neither 1.034
Mean VIF 1.39 Local rulemaking only 3.207
Max VIF 1.77 Local monitoring only 0.216
Both 61.032
Note: Standard errors are clustered by the location of the forest.
316 G. Epstein / Ecological Economics 131 (2017) 312–321
Fig. 1. Predicted probability of high levels of rule compliance. (A) Probability of compliance under different combinations of local rulemaking and local monitoring. All other variables are
held at the sample mean. (B, C, D) Comparison of the probability of compliance for the combination of local rulemaking and local monitoring; and (B) neither local rulemaking nor local
monitoring, (C) local rulemaking without monitoring, (D) local monitoring without rulemaking; as the collective actions index varies from minimum to maximum value. Plots show 90%
confidence intervals; all other variables are held at the sample mean.
changes as the collective action index varies from its minimum to max-
imum value suggest that specific institutional arrangements for fuel-
wood appropriation are less important for groups who regularly
cooperate to harvest and process resources. In fact, local rulemaking
by itself appears to be sufficient to promote compliance in groups
with fairly low levels of collective activity, although at values that ex-
ceed the sample average.
3.1. Robustness of Results
A number of additional statistical models can be found in the
Appendix A and were estimated in order to assess the robustness of re-
sults to the choice of units for clustering, control variables and potential
outliers. Table 5 clusters standard errors over country rather than forest
location, while Tables 6 and 7 replace distance to market with the fraction
of the user group that derives commercial benefits from the forest and a
forester's subjective assessment of the commercial value of a forest;
none of which have an impact on the sign or significance (p b 0.10) of
the coefficients for local rulemaking, local monitoring or the interaction
term.
Tables 8, 9 and 10 use the linear probability model to consider the
potential impacts of outliers and omitted variables on the results. The
linear probability model was used because the logistic model cannot
be estimated when an outcome is separated by one or a combination
of predictors (Long, 1997); a situation which arises when the outlier is
removed and when external monitoring is added as a control variable.
Table 8 replicates the model presented in Table 3 using the same vari-
ables and observations. Table 9 excludes one observation with a partic-
ularly large residual (Fig. 2), while Table 10 includes external
monitoring7 as a control variable, omitting a number of observations
7
All cases in which external monitors were absent had high levels of rule compliance.
The logistic model is unable to estimate models in which outcomes are perfectly predicted
by the values of a variable.
due to missing data. The results from all of these additional models
are consistent in terms of the sign and significance (p b 0.10) of the co-
efficients for local rulemaking, local monitoring and the interaction.
While the linear probability models have a number of well-known lim-
itations; it can also provide valuable insights for some questions
(Caudill, 1988). In this case it allows for a general assessment of two
major threats, outliers and omitted variables, to the internal validity of
the results presented in Table 3.
4. Discussion
As a whole the results presented in Section 3 provide support for the
hypothesis of a synergistic relationship between local rulemaking and
local monitoring. Estimates from several models and methods are con-
sistent in suggesting that the likelihood of high levels of rule compliance
in state-owned forest commons is highest when (i) user groups are pri-
marily responsible for the content of rules for fuelwood appropriation
and (ii) make investments to monitor compliance with those rules.
However, although these differences are statistically significant for the
average group in the sample (Fig. 1A); the results also show that in
some contexts that the impact of the combination of local rulemaking
and local monitoring on the likelihood of compliance is indistinguish-
able from other institutional arrangements (Fig. 1B, C, D). Most notably
it would appear that in some situations that local rulemaking might
generate sufficient motivation to promote high levels of compliance,
even in the absence of local monitoring and the associated security it
might provide. This analysis represents a first attempt to explicitly test
whether synergies between participatory rulemaking and enforcement
observed in laboratory experiments (DeCaro et al., 2015) apply to the
behavior of resource users in the field. Although this paper makes sever-
al potential contributions to our understanding of the relationship be-
tween participatory rulemaking, monitoring and compliance, it is
important to take note of some of its limitations.
 G. Epstein / Ecological Economics 131 (2017) 312–321
First levels of compliance with fuelwood appropriation rules are
not associated with better forest conditions as assessed by local for-
est experts.8 Therefore, although local rulemaking and monitoring in
state-owned forest commons may promote compliance, it does not
necessarily imply that governments can rely upon local users to create
and enforce rules in order to meet their environmental goals. Second,
given that the sample is comprised entirely of groups using fuelwood
in state-owned forest commons; the relationships observed in this
study might not apply to other forest products or forests that are
owned by private individuals or communities. In fact, it is possible
that communities would be more willing to comply with rules and en-
gage in sustainable patterns of use if they were granted formal owner-
ship of a forest and its resources (Pagdee et al., 2006; Chhatre and
Agrawal, 2009).
Third, this analysis which draws upon data regarding the behavior
and attributes of groups does not necessarily apply to the behavior of in-
dividuals (Robinson, 2009). Indeed, it would be highly problematic if
compliance depended upon whether every individual in a group was
personally responsible for the content of operational rules; and was in-
volved in monitoring compliance with those rules. Instead at the indi-
vidual level it is likely more important that individuals feel that they
are treated fairly by governance systems, providing opportunities for
“people like them” to have a say in rulemaking processes (Scholz and
Lubell, 1998; Tyler, 2006), and using monitoring mechanisms to sup-
port their collective will (DeCaro et al., 2015).
Fourth, because of the cross-sectional nature of the study and the
absence of suitable instruments it is possible that the results are bi-
ased by endogeneity between local rulemaking, local monitoring
and compliance. For instance, groups that have been successful in
promoting compliance might be granted more formal or informal au-
thority to create and modify rules in state-owned forest commons;
while high levels of voluntary compliance could increase the likeli-
hood of user group monitoring by reducing the costs associated
with its provision.
Fifth and finally because of (i) important questions concerning the
fungibility of external and self- monitoring (Vyrastekova and Van
Soest, 2003; Coleman, 2009), and (ii) limitations of logistic regression
and other methods; this paper is unable to provide a strong empirical
test of the synergies between local rulemaking and local monitoring
while controlling for external monitoring.
4.1. Rulemaking, Monitoring and Compliance
The results presented in Section 3 show that local rulemaking has a
positive and statistically significant impact on rule compliance; a result
which is consistent with the predictions of crowding theory (Frey and
Jegen, 2001). This corresponds closely to similar findings from the liter-
ature on tax compliance (Frey, 1997; Scholz and Lubell, 1998) and re-
source management (Hatcher et al., 2000; Viteri and Chávez, 2007;
Madrigal-Ballestero et al., 2013) which have found relationships be-
tween several different measures of political participation and compli-
ance. Furthermore it is possible and perhaps even likely that the
absence of a relationship reported in some studies (Kuperan and
Sutinen, 1999; Jenny et al., 2007; Eggert and Lokina, 2010) is the result
of models which include measures of both participation and the per-
ceived fairness of rules, which is often seen as a core outcome of partic-
ipation itself (Nielsen, 2003). As a result these models likely
(i) underestimate the total effects of participation by including inter-
vening variables (King et al., 2001; Ray, 2003) and (ii) suffer from
8
65% of forests in which users comply with fuelwood appropriation rules most or all of
the time have typical or better levels of species diversity; compared to 64% when users
rarely or sometimes comply. 62% of forests in which users comply with fuelwood appro-
priation rules most or all of the time have typical or better levels of vegetation density;
compared to 68% when users rarely or sometimes comply.
317
problems associated with collinearity9 which limit the power of statisti-
cal tests.
The relationship between compliance and local monitoring is, howev-
er, somewhat more complex Whereas local rulemaking is consistently as-
sociated with increases in the likelihood of compliance; the effects of
monitoring appear to depend upon how rules are developed. When
rules are imposed by external actors, local monitoring does not have a sig-
nificant effect on compliance and has the lowest predicted probability of
compliance (Fig. 1A). This is suggestive of a crowding-out effect from
monitoring (Frey and Jegen, 2001; Bowles, 2008), and is consistent with
the results of CPR experiments using external sanctions (Ostrom et al.,
1992; Cardenas et al., 2000) and self-enforcement of imposed rules
(DeCaro et al., 2015). In contrast local monitoring has a positive and statis-
tically significant impact on compliance when combined with local
rulemaking, a result which appears to contradict the hypothesis that mon-
itoring, in-and-of-itself, crowds-out intrinsic motivation (Deci and Ryan,
2000; Bowles, 2008). Instead it appears that local monitoring of rules cre-
ated by members of the user group enhances motivation to comply with
rules for fuelwood appropriation. DeCaro et al. (2015) argue that partici-
patory rulemaking legitimizes the use of enforcement, while enforcement
encourages actors that might disagree with rules (i.e. voted for a different
rule) to eventually accept them and provides security against exploitation
by free-riders. Together these findings highlight the need for further in-
vestigation of different monitoring and sanctioning mechanisms (i.e. ex-
ternal vs. self-monitoring) that explicitly considers how the attributes of
rulemaking processes influence perceptions of enforcement, and impact
motivation and behavior in different social and ecological contexts.
These findings are also partially consistent with models of conditional
cooperation where long-term compliance would be predicted to depend
upon the existence of mechanisms to observe and if necessary sanction
free-riders. Given the prevalence of free-riders in human societies
(Fischbacher et al., 2001; Frey and Meier, 2004; Kocher et al., 2008), mon-
itoring and sanctioning mechanisms would be expected to be an impor-
tant, if not essential condition for long-term cooperation and compliance
(Rustagi et al., 2010). Whereas the results support the hypothesis that
groups who combine rulemaking with monitoring are the most likely
to comply with rules, they also highlight the potential for achieving sim-
ilar levels of compliance without local monitoring, particularly for groups
that frequently work together to cooperatively harvest and process re-
sources. This is consistent with some CPR experiments where repeated
communication treatments have yielded similar outcomes as communi-
cation and sanctioning (Ostrom et al., 1994; Janssen et al., 2010), but
poses an interesting question as to how groups might address free-rider
problems in the absence of enforcement.
One potentially interesting possibility is that cooperative harvesting of
resources, a common strategy for fuelwood extraction (Wickramasinghe,
1997; Kant and Cooke, 1998; Tabuti et al., 2003), might be used to pro-
duce monitoring as a byproduct of appropriation eliminating the need
for organized patrols. Similar strategies have been used by irrigators
where the combination of timing and rotation rules for opening and clos-
ing gates to their respective fields creates a situation in which two or
more actors have the opportunity and incentive to monitor the behavior
of the other (Ostrom, 1990; Tang, 1992). Alternatively if resource users
can be easily observed as they enter or exit the forest to harvest fuelwood
resources, informal monitoring might act as a substitute for organized pa-
trols (Wade, 1994). There is also the possibility that some groups might
be able to promote compliance even in the absence of monitoring as a re-
sult of frequent interactions to address social problems, contributing to
higher levels of intrinsic motivation and trust.
9
The tetrachoric correlation between local rulemaking and perceived fairness of rules
in this study is 0.705 and significant at the 1% level; it is worth noting, however, that the
same argument might apply to the combination of local rulemaking and local monitoring.
The tetrachoric correlation between local rulemaking and local monitoring is 0.582 and
significant at the 1% level. However, given that I am explicitly seeking to test for synergies
between local rulemaking and local monitoring, the incorporation of both terms is neces-
sary, but likely underestimates the total effects of local rulemaking.
318 G. Epstein / Ecological Economics 131 (2017) 312–321

Finally, the results suggest that the positive effects of local monitoring
(when combined with local rulemaking) on the likelihood of compliance
are independent of the levels of collective action in that group. Therefore
it provides some, albeit weak, evidence against the hypothesis that local
monitoring (when combined with local rulemaking) crowds-out intrinsic
motivation for groups with high levels of trust (Vollan, 2008). The evi-
dence is weak because of the absence of an explicit test which would re-
quire an additional interaction term, potential concerns about differences
in the ways in which collective action and trust are measured in the re-
spective studies, and the distribution of the collective action index
which is somewhat biased towards groups with low levels of collective
activity. Nonetheless, the models suggest that groups that cooperatively
harvest and process resources outside of the forest frequently (likely cor-
related to levels of trust) would have a higher likelihood of compliance if
they combined local rulemaking with local monitoring. However, given
the above noted limitations much more work is needed to better under-
stand the ways in which levels of trust in particular and contexts in gen-
eral influence the performance of different institutional arrangements.
5. Conclusion
This paper set out to test the hypothesis that local rulemaking and
monitoring have a synergistic effect on compliance building upon recent
experimental work by DeCaro et al. (2015). The results are generally sup-
portive of this hypothesis, but also highlight the potential for achieving
high levels of compliance without enforcement by allowing groups to in-
fluence the content of the rules that affect them. Collectively these
findings provide important supporting evidence to the above noted ex-
perimental study but much more work is needed to investigate the rela-
tionship between institutions, motivation, and compliance. Although this
paper has shown that compliance tends to increase when local users are
involved in the design and enforcement of rules; the absence of a relation-
ship between compliance and forest conditions potentially implies a
tradeoff between promoting compliance and the sustainability of re-
source stocks. If communities systematically adopt rules that are less strict
in order to promote compliance, then even long-term compliance might
ultimately lead to the depletion of resource stocks. Therefore, greater at-
tention is needed to better understand the conditions in which local
rulemaking and local monitoring might contribute to compliance, while
also ensuring the long term sustainability of forests.
Acknowledgements
This paper has benefited greatly from the insightful comments from a
number of faculty and students at the Ostrom Workshop in Political The-
ory and Policy Analysis at Indiana University and two anonymous re-
viewers. In particular I would like to thank Dr. Michael McGinnis, Dr.
Daniel Cole, Dr. Eric Coleman, Dr. Burney Fischer, Dr. Sergio Fernandez
and Julie England who contributed invaluable insights as I developed
this research. This work would not have been possible if it was not for
the sustained efforts of a large number of scholars and research centers
associated with the International Forestry Resources and Institutions
Program.

Appendix A

Table 4
Pairwise correlation matrix.

Local Local Local rulemaking X Collective Poor Density Other Market Commercial Commercial External
rulemaking monitoring Local monitoring action index group users value monitoring

Local rulemaking 1.000

Local monitoring 0.383 1.000
Local rulemaking × 0.617 0.838 1.000
Local monitoring
Collective action index 0.001 0.062 −0.032 1.000
Poor 0.036 0.342 0.211 −0.068 1.000
Density 0.057 0.182 0.203 0.004 −0.053 1.000
Other group −0.374 −0.326 −0.470 0.063 −0.069 −0.508 1.000
Market −0.191 −0.238 −0.212 −0.004 −0.105 −0.068 0.171 1.000
Commercial users −0.207 0.021 −0.200 0.081 0.184 −0.376 0.305 −0.150 1.000
Commercial value −0.010 0.101 0.148 0.140 0.056 0.103 −0.113 0.174 −0.198 1.000
External monitoring −0.313 −0.121 −0.208 −0.074 0.012 −0.074 0.130 0.085 0.082 −0.013 1.00

Note: n = 93 for all correlations with the exception of external monitoring where n = 73.

Table 5
Results.

β SE p 90% C.I.

Institutional Arrangements
Local rulemaking 1.617 0.702 0.021 0.463 2.771
Local monitoring 0.387 0.453 0.393 −0.359 1.132
Local rulemaking X local monitoring 1.154 0.529 0.029 0.285 2.024

Controls
Collective action index 1.169 0.421 0.006 0.476 1.862
Poor −0.292 0.433 0.500 −1.005 0.421
Density −0.091 0.180 0.614 −0.387 0.205
Other group −1.080 0.397 0.007 −1.734 −0.427
Distance to market 0.616 0.245 0.012 0.214 1.019
Constant 1.641 0.145 0.000 1.402 1.880
N 93
Count R2 0.796
Chi-square –
p-value –
Mean VIF 1.39
Max VIF 1.77

Note: Standard errors are clustered by country rather than forest location.
 G. Epstein / Ecological Economics 131 (2017) 312–321 319

Table 6
Logistic Regression Results.

β SE p 90% C.I.

Institutional Arrangements
Local rulemaking 1.318 0.483 0.006 0.524 2.111
Local monitoring 0.613 0.580 0.291 −0.341 1.568
Local rulemaking X local monitoring 1.006 0.468 0.032 0.236 1.776

Controls
Collective action index 1.141 0.410 0.005 0.466 1.815
Poor −0.342 0.449 0.446 −1.080 0.397
Density −0.265 0.375 0.480 −0.882 0.352
Other group −1.064 0.777 0.171 −2.342 0.213
Commercial users −0.684 0.294 0.020 −1.168 −0.200
Constant 1.869 0.718 0.009 0.688 3.050
N 93
Count R2 0.785
Chi-square 29.97
p-value 0.000
Mean VIF 1.49
Max VIF 1.76

Note: Distance to market is replaced with commercial users, standard errors clustered by the location of the forest.

Table 7
Logistic Regression Results.

β SE P 90% C.I.

Local rulemaking 1.455 0.420 0.001 0.764 2.147

Local monitoring 0.381 0.502 0.448 −0.445 1.206
Local rulemaking X local monitoring 1.196 0.423 0.005 0.500 1.892

Controls
Collective action index 1.057 0.371 0.004 0.448 1.667
Poor −0.356 0.371 0.338 −0.966 0.255
Density −0.070 0.308 0.821 −0.577 0.438
Other group −1.157 0.724 0.110 −2.348 0.034
Commercial Value −0.093 0.304 0.761 −0.593 0.408
Constant 1.826 0.662 0.006 0.738 2.914
N 0.000
Count R2 0.763
Chi-square 36.78
p-value 0.000
Mean VIF 1.40
Max VIF 1.76

Note: Distance to market is replaced with the commercial value of the forest; standard errors clustered by forest location.

Table 8
Linear regression results using full sample.

β SE P 90% C.I.

Local rulemaking 0.206 0.059 0.001 0.107 0.305

Local monitoring 0.022 0.050 0.653 −0.061 0.106
Local rulemaking X local monitoring 0.123 0.048 0.013 0.043 0.203

Controls
Collective action index 0.093 0.033 0.007 0.038 0.149
Poor −0.033 0.032 0.318 −0.086 0.021
Density −0.017 0.056 0.760 −0.111 0.076
Other group −0.174 0.090 0.059 −0.325 −0.023
Market 0.089 0.029 0.003 0.041 0.138
Constant 0.721 0.090 0.000 0.571 0.870
N 93
R2 0.369
F 11.50
p-value 0.000
Mean VIF 1.39
Max VIF 1.77

Note: Standard errors are clustered by forest location.

320 G. Epstein / Ecological Economics 131 (2017) 312–321

Table 9
Linear regression results excluding potential outlier.

β SE P 90% C.I.

Local rulemaking 0.217 0.058 0.000 0.120 0.313

Local monitoring 0.040 0.047 0.395 −0.038 0.119
Local rulemaking X local monitoring 0.132 0.047 0.007 0.053 0.210

Controls
Collective action index 0.085 0.034 0.014 0.029 0.141
Poor −0.047 0.029 0.108 −0.095 0.001
Density −0.026 0.055 0.642 −0.117 0.066
Other group −0.145 0.085 0.095 −0.287 −0.002
Market 0.093 0.030 0.003 0.043 0.143
Constant 0.709 0.088 0.000 0.562 0.856
N 92
R2 0.399
F 12.50
p-value 0.000
Mean VIF 1.40
Max VIF 1.78

Note: Standard errors are clustered by forest location. The case with a large residual in Fig. 2 is excluded for this model.

Fig. 2. Standard Pearson Residuals for model presented in Table 3.

Table 10
Linear regression results including external monitoring.

β SE P 90% C.I.

Local rulemaking 0.146 0.060 0.017 0.047 0.246

Local monitoring 0.049 0.052 0.349 −0.038 0.137
Local rulemaking X local monitoring 0.096 0.049 0.055 0.014 0.178

Controls
Collective action index 0.090 0.038 0.023 0.025 0.154
Poor −0.063 0.042 0.136 −0.134 0.007
Density −0.022 0.055 0.695 −0.114 0.071
Other group −0.234 0.093 0.014 −0.389 −0.079
Market 0.116 0.035 0.001 0.058 0.174
External monitoring −0.216 0.088 0.018 −0.364 −0.068
Constant 0.969 0.108 0.000 0.788 1.150
N 73
R2 0.389
F 8.32
p-value 0.000
Mean VIF 1.35
Max VIF 1.66

Note: Standard errors are clustered by forest location. External monitoring is included as a control variable.

References
Acheson, J., 2006. Institutional failure in resource management. Annu. Rev. Anthropol. 35,
117–134.
Agnew, D.J., Pearce, J., Pramod, G., Peatman, T., Watson, R., Beddington, J.R., Pitcher, T.J.,
2009. Estimating the worldwide extent of illegal fishing. PLoS One 4, e4570.
Agrawal, A., Goyal, S., 2001. Group size and collective action. Comp. Polit. Stud. 34, 63–93.
Agrawal, A., Chhatre, A., Hardin, R., 2008. Changing governance of the world's forests. Sci-
ence 320, 1460–1462.
Andersson, K., Gibson, C.C., 2007. Decentralized governance and environmental change:
local institutional moderation of deforestation in Bolivia. J. Policy Anal. Manage. 26,
99–123.
 G. Epstein / Ecological Economics 131 (2017) 312–321
Andreoni, J., Erard, B., Feinstein, J., 1998. Tax compliance. J. Econ. Lit. 36, 818–860.
Basurto, X., Ostrom, E., 2009. Beyond the tragedy of the commons. Econ. Fonti Energy Am-
bient. 52, 35–60.
Becker, G.S., 1968. Crime and punishment: an economic approach. J. Polit. Econ. 76,
169–217.
Bowles, S., 2008. Policies designed for self-interested citizens may undermine “the moral
sentiments”: evidence from economic experiments. Science 320, 1605–1609.
Cardenas, J.-C., 2004. Norms from outside and from inside: an experimental analysis on
the governance of local ecosystems. Forest Policy Econ. 6, 229–241.
Cardenas, J.C., Stranlund, J., Willis, C., 2000. Local environmental control and institutional
crowding-out. World Dev. 28, 1719–1733.
Caudill, S.B., 1988. Practitioners corner: an advantage of the linear probability model over
probit or logit. Oxf. Bull. Econ. Stat. 50, 425–427.
Chhatre, A., Agrawal, A., 2009. Trade-offs and synergies between carbon storage and live-
lihood benefits from forest commons. Proc. Natl. Acad. Sci. 106, 17667–17670.
Coleman, E.A., 2009. Institutional factors affecting biophysical outcomes in forest manage-
ment. J. Policy Anal. Manage. 28, 122–146.
Contreras-Hermosilla, A., 2002. Law Compliance in the Forestry Sector. An Overview.
World Bank Institute, Washington.
Cox, M., Arnold, G., Villamayor Tomas, S., 2010. A review of design principles for commu-
nity-based natural resource management. Ecol. Soc. 15, 38.
DeCaro, D.A., Stokes, M.K., 2013. Public participation and institutional fit: a social–
psychological perspective. Ecol. Soc. 18, 40.
DeCaro, D.A., Janssen, M.A., Lee, A., 2015. Synergistic effects of voting and enforcement on
internalized motivation to cooperate in a resource dilemma. Judgm. Decis. Mak. 10,
511–537.
Deci, E.L., Ryan, R.M., 2000. The “what” and “why” of goal pursuits: human needs and the
self-determination of behavior. Psychol. Inq. 11, 227–268.
Eggert, H., Lokina, R., 2010. Regulatory compliance in Lake Victoria fisheries. Environ. Dev.
Econ. 15, 197–217.
Ehrlich, I., 1973. Participation in illegitimate activities: a theoretical and empirical inves-
tigation. J. Polit. Econ. 81, 521–565.
Fischbacher, U., Gächter, S., Fehr, E., 2001. Are people conditionally cooperative? Evidence
from a public goods experiment. Econ. Lett. 71, 397–404.
Frey, B.S., 1997. A constitution for knaves crowds out civic virtues. Econ. J. 107,
1043–1053.
Frey, B.S., Jegen, R., 2001. Motivation crowding theory. J. Econ. Surv. 15, 589–611.
Frey, B.S., Meier, S., 2004. Social comparisons and pro-social behavior: testing “conditional
cooperation” in a field experiment. Am. Econ. Rev. 94, 1717–1722.
Fromentin, J.M., Powers, J.E., 2005. Atlantic bluefin tuna: population dynamics, ecology,
fisheries and management. Fish Fish. 6, 281–306.
Gneezy, U., Rustichini, A., 2000a. A fine is a price. J. Leg. Stud. 29, 1–17.
Gneezy, U., Rustichini, A., 2000b. Pay enough or don't pay at all. Q. J. Econ. 115, 791–810.
Hardin, G., 1968. The tragedy of the commons. Science 162, 1243–1248.
Hatcher, A., Jaffry, S., Thébaud, O., Bennett, E., 2000. Normative and social influences af-
fecting compliance with fishery regulations. Land Econ. 76, 448–461.
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.
Am. Econ. Rev. 91, 73–78.
Janssen, M.A., Holahan, R., Lee, A., Ostrom, E., 2010. Lab experiments for the study of so-
cial-ecological systems. Science 328, 613–617.
Jenny, A., Hechavarria Fuentes, F., Mosler, H.-J., 2007. Psychological factors determin-
ing individual compliance with rules for common pool resource management:
the case of a Cuban community sharing a solar energy system. Hum. Ecol. 35,
239–250.
Kant, S., Cooke, R., 1998. Conflict Resolution: A Critical Factor in the Success of Joint Forest
Management in India.
Kerr, N.L., Garst, J., Lewandowski, D.A., Harris, S.E., 1997. That still, small voice: commit-
ment to cooperate as an internalized versus a social norm. Personal. Soc. Psychol.
Bull. 23, 1300–1311.
King, D.M., Sutinen, J.G., 2010. Rational noncompliance and the liquidation of Northeast
groundfish resources. Mar. Policy 34, 7–21.
King, G., Keohane, R.O., Verba, S., 2001. Designing Social Inquiry: Scientific Inference in
Qualitative Research. Princeton University Press.
Kocher, M.G., Cherry, T., Kroll, S., Netzer, R.J., Sutter, M., 2008. Conditional cooperation on
three continents. Econ. Lett. 101, 175–178.
Kuperan, K., Sutinen, J.G., 1998. Blue water crime: deterrence, legitimacy, and compliance
in fisheries. Law Soc. Rev. 32, 309–338.
Kuperan, K., Sutinen, J.G., 1999. A socio-economic theory of regulatory compliance. Int.
J. Soc. Econ. 26, 174–193.
Long, J.S., 1997. Regression Models for Categorical and Limited Dependent Variables. Sage,
Thousand Oaks, CA.
MacKenzie, B.R., Mosegaard, H., Rosenberg, A.A., 2009. Impending collapse of bluefin tuna
in the Northeast Atlantic and Mediterranean. Conserv. Lett. 2, 26–35.
321
Madrigal-Ballestero, R., Schlüter, A., Claudia Lopez, M., 2013. What makes them follow the
rules? Empirical evidence from turtle egg harvesters in Costa Rica. Mar. Policy 37,
270–277.
Nagendra, H., Ghate, R., 2005. Role of monitoring in institutional performance: Forest
Management in Maharashtra, India. Conserv. Soc. 3, 509–532.
Nielsen, J.R., 2003. An analytical framework for studying: compliance and legitimacy in
fisheries management. Mar. Policy 27, 425–432.
Olson, M., 1965. The Logic of Collective Action: Public Goods and the Theory of Groups.
Harvard University Press.
Ostrom, E., 1990. Governing the Commons. Cambridge University Press, New York.
Ostrom, E., 1998. A behavioral approach to the rational choice theory of collective action:
presidential address, American Political Science Association, 1997. Am. Polit. Sci. Rev.
92, 1–22.
Ostrom, E., 2000. Crowding out citizenship. Scand. Polit. Stud. 23, 3–16.
Ostrom, E., 2009. Building trust to solve commons dilemmas: taking small steps to test an
evolving theory of collective action. In: Levin, S.A. (Ed.), Games, Groups, and the Glob-
al Good. Springer, Berlin Heidelberg, pp. 207–228.
Ostrom, E., Walker, J., Gardner, R., 1992. Covenants with and without a sword: self-gover-
nance is possible. Am. Polit. Sci. Rev. 86, 404–417.
Ostrom, E., Gardner, R., Walker, J., 1994. Rules, games and common-pool resources. Uni-
versity of Michigan Press, Ann Arbor.
Pagdee, A., Kim, Y.-s., Daugherty, P.J., 2006. What makes community forest management
successful: a meta-study from community forests throughout the world. Soc. Nat.
Resour. 19, 33–52.
Poteete, A.R., Janssen, M.A., Ostrom, E., 2010. Working Together. Princeton University
Press, Princeton, NJ.
Ramcilovic-Suominen, S., Epstein, G., 2012. Towards an analytical framework for forest
law compliance. Int. For. Rev. 14, 326–336.
Ramcilovic-Suominen, S., Epstein, G., 2015. The impacts of deterrence, social norms and
legitimacy on forest rule compliance in Ghana. Forest Policy Econ. 55, 10–20.
Ray, J.L., 2003. Explaining interstate conflict and war: what should be controlled for?
Confl. Manag. Peace Sci. 20, 1–31.
Robinson, W., 2009. Ecological correlations and the behavior of individuals. Int.
J. Epidemiol. 38, 337–341.
Romme, W.H., Don, G.D., 1989. Historical perspective on the Yellowstone fires of 1988.
Bioscience 39, 695–699.
Rustagi, D., Engel, S., Kosfeld, M., 2010. Conditional cooperation and costly monitoring ex-
plain success in forest commons management. Science 330, 961–965.
Ryan, R.M., Deci, E.L., 2000. Self-determination theory and the facilitation of intrinsic mo-
tivation, social development, and well-being. Am. Psychol. 55, 68–78.
Scholz, J.T., Lubell, M., 1998. Trust and taxpaying: testing the heuristic approach to collec-
tive action. Am. J. Polit. Sci. 42, 398–417.
Shinada, M., Yamagishi, T., 2007. Punishing free riders: direct and indirect promotion of
cooperation. Evol. Hum. Behav. 28, 330–339.
Tabuti, J., Dhillion, S., Lye, K., 2003. Firewood use in Bulamogi County, Uganda: species se-
lection, harvesting and consumption patterns. Biomass Bioenergy 25, 581–596.
Tang, S.Y., 1992. Institutions and Collective Action: Self-Governance in Irrigation. ICS
Press, San Francisco.
Tyler, T.R., 2003. Procedural justice, legitimacy, and the effective rule of law. Crime Justice
283–357.
Tyler, T.R., 2006. Why People Obey the Law. Princeton University Press, Princeton, NJ.
Viteri, C., Chávez, C., 2007. Legitimacy, local participation, and compliance in the
Galápagos Marine Reserve. Ocean Coast. Manag. 50, 253–274.
Vollan, B., 2008. Socio-ecological explanations for crowding-out effects from economic
field experiments in southern Africa. Ecol. Econ. 67, 560–573.
Vyrastekova, J., Van Soest, D., 2003. Centralized common-pool management and local
community participation. Land Econ. 79, 500–514.
Wade, R., 1994. Village Republics. ICS Press, San Francisco, CA.
Wertime, M., Ostrom, E., Gibson, C., Lehoucq, F., 2007. International Forestry Resources
and Institutions (IFRI) Research Program: Field Manual. Version 13. Center for the
Study of Institutions, Population and Environmental Change, Indiana University,
Bloomington.
Wickramasinghe, A., 1997. Anthropogenic factors and forest management in Sri Lanka.
Appl. Geogr. 17, 87–110.
World Bank, 2004. Sustaining Forests: A Development Strategy. The World Bank,
Washington, DC.
Young, O.R., 1979. Compliance and Public Authority: A Theory With International Appli-
cations. The Johns Hopkins University Press, Baltimore.
Zaelke, D., Stilwell, M., Young, O., 2005. Compliance, rule of law and good governance. In:
Zaelke, D., Kaniaru, D., Kruzikova, E. (Eds.), Making Law Work: Environmental Com-
pliance and Sustainable Development. Cameron May Ltd., London, pp. 30–51.