The Durability of Collective Action: When do Intentional Communities Survive Crises?
Amy R. Poteete University of New Orleans
Draft – cite at your own risk. (Please check for revisions first) Comments welcome: firstname.lastname@example.org
Paper prepared for the annual meeting of the Southern Political Science Association, New Orleans, January 8 – 10, 2004.
Durability of Collective Action
The Durability of Collective Action: When do Intentional Communities Survive Crises? Amy R. Poteete1 University of New Orleans How can variation in the ability of groups to sustain collective action be explained? Responses of intentional communities to challenges may shed light on variation in the durability of cooperation. Intentional communities encompass a wide variety of groups formed in an effort to create community, often through group living or cooperative working arrangements. These attempts to fashion alternatives to mainstream socioeconomic arrangements depend critically upon sustained contributions by members. Ideological commitment brings many individuals into such communities, but they face numerous challenges once established, ranging from financial pressures to resolution of conflicts. Such challenges may prompt clarification of organizational arrangements or lead to the dissolution of the community. This paper evaluates the relative importance of factors expected to influence responses to crises: group age, group size, institutional definition, institutional flexibility, trial membership periods, ease of exit, and organizational change in response to the crisis. Data are drawn from a long-term study of five intentional communities in southern Indiana. All five sites encompass forests owned and managed jointly by community members. Otherwise, the communities are quite diverse, ranging from a small set of friends sharing land through spiritual and recreational communities to a larger suburban residential community. Financial challenges confronted all five communities at various points. Internal disputes in three communities spilled into the court system. Responses to these challenges varied, as has the relative success in overcoming them. None of the five has completely collapsed, although two communities saw membership numbers plummet at least once. The variation in purpose and experience represented by these communities offers a unique opportunity to analyze the responses of groups to major challenges.
According to both theory and empirical research, sustaining collective action once achieved is easier than achieving it in the first place (Axelrod 1984; Baland and Platteau 2000; Olson 1971; Ostrom 1999). Nonetheless, the durability of collective action varies considerably. Some groups disband after achieving their initial goals. Others build on past success by adding new members and activities. Yet crises can overwhelm even those groups with a record of past successes. What factors influence the ability of groups to survive crises?
The hospitality of the five communities discussed in this study made this study possible I extend my sincere appreciation to the members of the research teams for their diligence and enthusiasm in the field. and enjoyable. The FAO, Ford Foundation, MacArthur Foundation, and NSF supported data collection. My work with IFRI was made possible through a grant from the Ford Foundation.
Durability of Collective Action
Past studies of the durability of collective action focus on characteristics of group members and of the groups themselves, especially their institutional design. Relevant variables can be culled from research on the collective management of shared natural resources, communal organizations, and organizational ecology. Ostrom’s (1999) review of studies of collective management of natural resources found that members of enduring institutions were characterized by (i) sufficient dependence on the resource to make investment in management worthwhile, (ii) common understanding of the resource system, (iii) relatively low discount rates, (iv) a shared interest in the survival of the resource among the more powerful or wealthy members, (v) trust, and (vi) prior organizational experience. Her review also found that long-enduring institutions enjoyed autonomy from government intervention, at least in the area of natural resource management. Survival of communal organizations has been linked to high costs of entry and exit, the existence of trial membership periods and formal socialization processes, egalitarian organization, and legitimate leadership (Kitts 2001; Shenker 1986).2 Organizational sociologists expect organizational age, the spatial distribution of complementary and competing organizations, and fashions in organizational forms to influence organizational mortality (Kitts 2001). These streams of research are concerned with the overall longevity of collective endeavors. Yet the risks to organizational survival change over time. Organizational sociologists hypothesize that the risk of organizational mortality is greatest for the newest organizations. Organizations that survive the first few years tend to have more robust structures, facilitating their ability to survive normal challenges.3 Less attention has been given to the relative importance of factors known to be associated with survival of crises.4 Which factors associated with longevity in general contribute to survival of crises? Should durability be understood simply as the ability to survive crises? To answer these questions, I draw upon data from an on-going study of five intentional communities in southern Indiana. Intentional communities refer to groups
See Hirshman (1970) for the classic argument of how exit and entry costs affect organizational performance. 3 With advanced age, organizations may face renewed risks from institutional ossification (Kitts 2001; Olson 1982). Kitts (2001) found that the risk increased again only after 100 years. The current study lacks the temporal depth needed to explore this possibility. 4 But see McKean (1992).
Durability of Collective Action formed in an effort to “create” community (Brown 2002; Fellowship for Intentional Societies nd.; Shenker 1986). Intentional communities are quite diverse, with no single definition even among practitioners (Fellowship for Intentional Societies nd). Many communities live in shared buildings, or on communally owned land. Others have separate residences but cooperate in the management and use of shared property. Collective economic endeavors are common but not universal. Regardless of the details of residential and economic organization, intentional communities attempt to create
communities that encompass multiple dimensions of members’ lives. Although a sense of community can be established in the short-term only to fade with the passage of time, these groups seek to create communities that endure at least for the lifetime of the founding members, and usually into subsequent generations. It might not make sense to expect groups with short-term goals to endure. Questions concerning longevity and the survival of crises are entirely relevant for intentional communities, given the long-term nature of their self-defined goals. Evaluating the Success of Collective Action How should we evaluate the success of collective action? Groups mobilize to meet diverse goals. Some seek to mobilize resources to create something (e.g., building an irrigation canal). Others mobilize resources to defend existing assets (e.g., guarding a forest against poaching). Once a collective good is achieved or defended from destruction, maintenance may be necessary. To evaluate collective action in terms of selfdefined goals requires criteria for drawing comparisons across different types of collective action. It is possible to compare the mobilization of resources in terms of monetary contributions; comparisons of labor contributions are less straightforward. Moreover, it is not clear whether the quantities contributed are as relevant as the quality of achievements. Not all goals require the same contributions, neither qualitatively nor quantitatively.5 How can differences in efficiency in resource use be taken into consideration? Survival rates offer an alternative indicator of collective success. Duration can be measured on a common scale. Nonetheless, it is problematic to treat duration as a measure of success without considering the goals of the group. For groups with short5
Cf., Sandler (1992, ch. 2) on different “technologies” for providing collective goods.
Durability of Collective Action term goals, a limited lifespan need not indicate failure. Even if the comparison is limited to groups with long-term goals, such as intentional communities, longevity alone says little about the quality of group performance. A group may survive for years yet become decreasingly active or stray from its original goals. Longevity loses its meaning if divorced from more qualitative assessments of performance. Responses to crises or challenges6 reflect group performance in a more meaningful manner. From this perspective, absolute collapse becomes an extreme point on a continuum. Some groups will survive crises with little change in their organization, capacity to survive subsequent crises, or performance in day-to-day activities. The most successful adapt to challenges by improving their organization, thereby increasing their ability to survive future crises and potentially enhancing their day-to-day performance. Survival of crises should improve a group’s ability to reproduce itself in terms of membership, thus enhancing its longevity. The endpoints of the continuum – collapse and strengthening through survival – are clear. Intermediate outcomes cannot be ranked so clearly. Crises may prompt groups
to broaden or narrow their scope of collective action. Not all adaptations made to survive crises, however, will be beneficial for performance during ordinary times or please core members. For instance, broadening the scope of collective action may improve the ability to recruit new members or raise new resources but involve a loss of focus or deterioration in the quality of performance in the original areas of activity. Members motivated by the original goals might prefer to disband rather than alter the group’s goals. With due attention given to possibility that groups survive in weakened form, responses to crises serve as useful indicators of group performance. This paper focuses on variables expected to influence responses to crises: group characteristics, the recruitment and retention of members, exit costs, and organizational changes in response to past crises. I give some attention to the organizational experience of members, although the data do not support a systematic analysis of this factor. When organizations are formed, groups develop institutions that define the terms of membership, especially the distribution of rights and responsibilities, including
“Challenge,” then, describes the general phenomenon and “crisis” a subset. “Crisis” conveys a greater sense of urgency and is understood as a more severe form of challenge. Since the focus of this paper is on existential challenges, I favor the term “crisis.”
Durability of Collective Action
decision-making authority. I focus on institutional definition, institutional flexibility, and the influence institutions have on member values. Lack of institutional definition can give rise to conflicts, or allow conflicts to fester, by failing to provide clear guidance for how the group should respond to situations with serious implications. Conflicts are especially likely if alternative responses either imply vastly different distributional outcomes or involve choices between organizational goals. Lack of institutional definition reflects the costs of institutional development as well as the failure to recognize the need for clarity on particular issues. Conflict resolution becomes particularly problematic if property rights, rights of expulsion, and authority over resource management are poorly defined. Experience with near disasters may raise awareness of risks and encourages greater institutional definition (McKean 1992, 73). Institutional flexibility facilitates adaptation to rapidly changing conditions by allowing groups to change their behavior without first having to change their rules. Flexibility also aids conflict resolution, by allowing room for error (Ostrom 1990). Institutional flexibility and institutional definition seem inherently contradictory goals, and in some situations they may be. Since institutional definition and flexibility are continuous measures, groups need not choose one extreme or the other. Likewise, high levels of definition in some areas can coexist with considerable flexibility in others. Groups must decide how to strike a balance between definition and flexibility. Flexibility is most valuable in rules concerning day-to-day activities. Institutional flexibility or lack of institutional definition is most likely to threaten organizational survival when group members disagree on group goals. Lack of agreement on goals reflects value heterogeneity. Institutions, in turn, influence value heterogeneity (Gibson and Koontz 1998). Studies of utopian and intentional communities have long stressed the importance of carefully initiating members (Mariampolski 1977; Shenker 1986). Much of the attention given to the need for commitment (Kanter 1968, 1972) can be understood as a need to recruit members with similar values and enhance similarity in values through a socialization process. In the absence of a screening mechanism, individuals may join to take advantage of collective benefits without contributing their fair share. Even without cynicism among potential members, inattention to socialization allows initial misperceptions about community goals to survive. As membership expands,
Durability of Collective Action the original members may find more recent members with different goals redirecting the community – and realize that they can do nothing about it. Or, differences in goals may proliferate, possibility interfering with the group’s ability to reach collective decisions.
Higher costs of leaving a community encourage members to work to overcome problems (cf., Hirshman 1970). Intentional and utopian communities have historically required that recruits endure a trial period before gaining full membership rights and responsibilities. Raising the costs of entry and exit also helps screen out those with a weak commitment to community goals or a desire to free-ride on collective benefits. Utopian communities in particular often require new members to give up their assets, either by donating them to the group or simply not making use of them while a member. The collapse of the Owenite experiment at New Harmony, Indiana has been attributed in part to the lack of attention given to the recruitment and socialization of members (Mariampolski 1977). Robert Owen traversed the country promising prosperity, good housing, superior childcare and education, medical care and other individual benefits. His speeches and print advertisements gave less emphasis to the work required to achieve these goals. Within a year, New Harmony attracted over a thousand recruits, outstripping the available housing stock. Many recruits did not understand the plans for collective production; the leadership managed labor resources inefficiently. A group of educational experts recruited from Philadelphia brought a competing set of utopian goals, setting the stage for ideological divisions. Within two years, in 1827, New Harmony collapsed, bankrupt and riven by ideological conflicts. An earlier study of two of the five communities discussed in this paper found that institutional arrangements played an important role in promoting value homogeneity, and thus successful collective action (Gibson and Koontz 1998). The Maple community set high costs of entry and restrictions on exit. They held regular meetings and circulated community information on a regular basis. New members enter into a membership agreement and received a membership packet that explicitly spells out member rights and responsibilities. According to Gibson and Koontz, these arrangements promote homogeneity in values, thereby contributing to stability in membership and success in collective action. In contrast, individuals faced relatively low costs to join the Oak community and could easily leave the group, resulting in a less stable membership.
Durability of Collective Action Meetings, although regularly scheduled, were poorly attended and minutes were not circulated. Differences over the management of the community’s resources eventually
spilled over into court battles. Gibson and Koontz’s analysis reinforces the importance of arrangements for initiating new members, and suggests that institutions that enhance interaction and communication among longstanding members also increase homogeneity in values. Institutions can facilitate or hinder collective action, but do not determine outcomes. The attributes participants bring to interactions are also important. In particular, past organizational experience creates a mental portfolio of options that members refer to when deciding how best to address collective problems. From experiences with failed groups, individuals know what does not work. Even failed groups typically have moments of success. Individuals draw on these experiences when designing new institutions or altering existing institutions. Experience with organizational successes and failures also inculcate a sense that there is no single best way of doing things, encouraging a willingness to experiment. Groups that consist of members with more experience with other organizations can be expected to adapt to challenges more readily, because of their greater acceptance of the need for change as well as the availability of a stock of earlier organizational experiences to draw upon. Responses to Crises among Communities in Southern Indiana Research teams based at Indiana University collected data on five communities in southern Indiana as part of a larger study concerning collective management of forests, the International Forestry Resources and Institutions (IFRI) research program. IFRI is an interdisciplinary program for the study of forests, the people who use forest resources, and their institutions for forest management.7 The methodology involves collection of biophysical, demographic, and institutional data. To allow over-time comparisons, research teams conduct repeat studies of each site every five years. The Center for the Study of Institutions, Population, and Environmental Change and the Workshop in Political Theory and Policy Analysis at Indiana University coordinate the IFRI program and host the US center. The research team identified six communities in the vicinity of Indiana University that met the criterion of owning shared forest land. Over the past ten
For more information about IFRI, see Ostrom and Wertime (2000) and Poteete and Ostrom (forthcoming).
Durability of Collective Action years, the research team has conducted two rounds of field work five years apart at each of the five communities that agreed to participate in the study. These five communities are quite diverse, as are intentional communities more generally. Each owns and manages some forested land collectively. For two communities, the acquisition of communal land satisfied a desire for a collective gathering place. For another, the communal land serves as a buffer against development that protects the privacy of the members. A fourth community also sees their land as a buffer against development, but gives greater priority to the preservation of nature and spiritual attachment to the land. The fifth maintains its property as a shared recreational
asset. These communities range in size from four members (the group seeking privacy) to just under 1500 (the group organized around recreational assets). To protect the privacy of our respondents, I use pseudonyms assigned in previous articles: Box Elder, Maple, Oak, Tulip Poplar, and Twin Oaks (Gibson and Koontz 1998; Poteete and Welch forthcoming). Table One summarizes key characteristics of the communities. [Table One about here] Some crises are inherent to the collective ownership and management of land. Each community had to raise money to purchase land, make payments to keep the land, develop physical infrastructure, and maintain their property. Failure to raise adequate resources threatens the ability to keep or maintain shared land, a defining characteristic of these communities. Most likely, these groups would disband if they lost their land. Decisions about recruitment involve a trade-off between increasing the pool of resources the community can draw upon and watering down the intimacy of the community. Communities differed in the balance struck. Some communities changed their approach to recruitment over the years; three set upper bounds on expansion. Each group also confronted the challenge of deciding how to organize itself. The desire to own land collectively made the issue of legal definition unavoidable. The communities varied in the extent to which they went beyond definition of a legal structure for property ownership and worked out organizational details for day-to-day operations. No community can avoid conflicts. Community organizations varied in their capacity to resolve conflicts internally; three had to turn to the courts to settle disputes. In addition to crises that all communities face, some also encountered random shocks.
Durability of Collective Action Crises often led to loss of membership and considerable internal conflict among the remaining members. One community regrouped and emerged stronger. Three
overcame immediate challenges without significant reorganization. Unresolved conflicts afflicted the fifth community for several years; its ultimate survival was in doubt at the time of the last study. Tables Two and Three summarize responses to crises by each community. Short historical sketches follow. [Tables Two and Three about here] Box Elder8 American society offers space for diversity of religious beliefs and lifestyles despite an overall conservatism. The existence of communities based on nature spirituality in the Midwest typifies this tension. In the early 1980s, some friends got the idea of founding communities where people could gather to celebrate a connection with nature and enjoy the company of likeminded individuals. In 1983, a small group began working toward this goal by organizing festivals in public parks. Distorted reports in the media suggested that festival participants engaged in sinister activities. Group members realized that they needed their own property if they wanted their gatherings to be undisturbed. The group purchased an old farm in 1987 with funds raised from festivals. Initially, there were concerns about the ability to raise sufficient funds to pay for the land, and about security given the hostility generated by the earliest festivals. Two decisions were made. First, the group organized as a non-profit membership organization. The organization held the land title. To help raise funds, the organization sold a limited number of parcels of land (about thirty) to individual members on an informal basis. Second, to provide additional protection for the community, the plots were arrayed along the borders with the road and neighboring properties. The thought was that any troublemakers would be deterred by the need to cross private property to reach the festival grounds. Raising funds turned out to be less problematic than expected. The festivals generated enough funds to pay off the debt for the land within two years. Local harassment was more serious, sometimes reaching the level of intimidation. Some minor
Site summaries draw on the author’s field notes for three of the communities, the IFRI archives at Indiana University, Gibson and Koontz (1998), and Poteete and Welch (forthcoming).
Durability of Collective Action vandalism and “accidental” trespassing by hunters occurred. Neighbors frequently call
the police to complain about noise during festivals. Box Elder launched a public relations campaign to explain their beliefs and demonstrate good citizenship within the broader community. They also constructed a shelter that dampens the sound of the drumming. The level of toleration increased over time. The private plots soon became problematic themselves. The idea of private property implies that the property holder can develop that land with limits set only by the law, can prevent others from using that land, and has no obligation to help with maintenance of community areas. These notions threatened the goals of creating a communal space and a sense of community endeavors. The issue reached crisis proportions between 1990 and 1992. A landholder stopped contributing to community work-days or other collective endeavors and did not maintain his own site in accordance with community expectations.9 The community evicted the landholder, who then took the case before the court. The court sided with Box Elder, probably because of the informal nature of the land sales. In the wake of these events, the community decided to buy back plots. Some original landholders had already sold plots back to the community upon deciding to leave the community. Recognition of the negative implications of having private areas embedded in the communal space prompted a shift from passively accepting opportunities to reclaim land to actively encouraging buy-backs. Landholders were not forced to sell their plots, but only a handful of private parcels remained at the time of the most recent study. Box Elder is a large organization with several hundred members. It is easy to join when attending a festival. Not surprisingly, many members have a very weak attachment to the organization and there is relatively high turnover. A smaller core of active members makes decisions about group activities and resource management (cf., Michels 1959). The need to rely on a small set of active members to make the organization function sits uneasily with the goal of operating as an equitable community.
Explanations of the conflict were somewhat vague on the details. It appears that there were concerns about sanitation and safety.
Durability of Collective Action Box Elder’s formal organization acknowledges different levels of commitment among members while promoting equity, at least in opportunity. A tiered arrangement allows people to choose among five levels of involvement. Membership rights expand with commitments to contribute to collective needs. The basic form of membership
grants rights to attend events and camp on the property in exchange for fees. Members at intermediate levels gain more extensive rights to use the facilities, receive discounts on regular fees, and can substitute community work for cash payment of fees. In exchange, these members commit to work for the community for designated number of hours each month. A council of decision-makers is selected from members who have committed to the highest level of monthly labor for at least one year and one day.10 The council makes decisions by consensus. Although only those with a long-term and intensive commitment are eligible for participation in decision-making, the most active members promote the idea that the organization is both equitable and flexible. The possibility of moving among levels of membership offers flexibility for individuals who have changing demands on their time and is portrayed as having no implications for status within the group. The hierarchical nature of the arrangement is acknowledged in its common description as “a pyramid with the top cut off.” It may be necessary to rely on a small set of active members, but reliance on consensus among councilors prevents the concentration of power in the hands of one or a very few individuals. Of course, consensus can be reached through assent or abstention as well as active agreement, and persistent individuals may wear down opponents. There were hints during interviews that a few individuals were disproportionately influential for these reasons. At the same time, reliance on festivals and basic membership fees provides a very strong check on decision-making. The ease of leaving the group or avoiding the festivals ensures rapid feedback to decisions that substantively alter the organization’s operations or sense of purpose. These organizational arrangements have served Box Elder well for more than two decades. It is not known how much prior organizational experience the founders could draw upon. The group had several years to develop its organizational framework between
Current councilors select new councilors. The number of councilors is not set and has varied over time. The advantages of encouraging higher levels of commitment through expansion of the council must be weighed against the increased difficulty of reaching consensus in larger groups.
Durability of Collective Action
1983 when it first coalesced and the purchase of land in 1987. The organization of several festivals provided opportunities to experiment with different organizational forms, and to recognize the administrative requirements of large-scale endeavors. By the time IFRI studied the community, its leadership included individuals with legal, accounting, and computer skills. It is not clear whether the community is particularly effective at recruiting and tapping the abilities of those with relevant skills, or whether the most active members developed relevant skills to address perceived needs. In either case, Box Elder manages its operations relatively smoothly. Reliance on membership dues and festival participation leaves Box Elder vulnerable to conditions beyond its control. Electronic newsletters always include appeals for participation in community work-days and contributions of both labor and material goods for special projects. Tours of the property reveal that the infrastructure is in decent shape, but could be maintained somewhat better. Recruiting members and mobilizing resources are on-going challenges, if not currently crises. Barring random shocks of unusual severity, Box Elder should continue to survive for the foreseeable future. Maple In 1973, an individual advertised for people with an interest in acquiring land collectively. Respondents to the ad began to meet regularly to develop plans for their community. Suitable land was identified and purchased in 1975. The Maple community was formally established the following year when the first members moved onto the property. The first members had a number of different goals: creation of community, attainment of self-sufficiency, stewardship of the land, and living with nature. Maple’s founders spent three years developing plans before officially launching their community. After considering a number of alternative forms of organization, they organized as a share-holding corporation to ensure truly collective ownership and decision-making. Funds raised through the sale of shares were used to make the down payment. Each shareholder contributed to paying off the mortgage. Shareholders pay an on-going monthly maintenance fee and participate in community work-days. This arrangement distributes financial responsibilities equitably among members. Corporate law guarantees that all shareholders have an equal vote on management decisions. Members adapted legal requirements to suit their own needs. The required
Durability of Collective Action annual meeting takes the form of a party celebrating the anniversary of their founding. One-year term limits on most council positions mean that they rotate among the membership; each community member has served on the council in several capacities. Legally, major decisions require the support of two-thirds of the shareholders, but members of the Maple community are loath to act without a consensus. Decisions are often worked out through informal conversations over months and even years before
being taken up by the council. In the absence of widespread support, an issue is unlikely to be brought to a vote. Initially, between 40 and 50 shares were purchased and around 145 individuals moved onto the property.11 Within a few weeks, the number of shareholders dropped to around 30. Most members moved into a community building with apartments after one year. That building burned down in 1980, forcing several members to live once again in teepees. Between 1976 and 1981, the number of shareholders declined to about 10. The fire was a flashpoint, but probably not the sole reason for declining membership. There were two underlying crises during these early years: the severity of conditions during the first tough year on the land and an initial lack of agreement about community goals. Share prices were originally quite low because costs were shared among a relatively large group of people. But conditions were harsh. Residents lived in teepees, did not have running water, and had to work hard to make their land habitable. Living on land under these conditions required a high level of commitment, especially once winter set in. Many who left during the early days lost their investment; the small size of the financial investment did little to discourage departures. The remaining members saw their own costs increase, as land payments and maintenance expenses had to be split among a shrinking pool of members. The remaining members probably felt a sense of urgency about retaining members and recruiting new members to help meet financial needs. In interviews, current members described the founders as having diverse concerns, and as somewhat naïve and ideological. Reports of conflicts over selfsufficiency versus taking jobs in town, and over the separation of utensils for vegetarians, give some sense of the intellectual climate. More serious conflicts arose over residential arrangements. Maple was founded as both an experiment in community living and an
Since each household may consist of several individuals, group size exceeds the number of shares.
Durability of Collective Action
effort to live in harmony with the land. Both goals are subject to divergent interpretations. Community living initially involved shared meals prepared in a community kitchen and the construction of a community building with apartments. Some individuals, however, wanted separate dwellings. Destruction of the community building by fire forced the issue. Instead of rebuilding the community building, separate household residences were developed. The move to separate household residences had implications for development of the property. The land was zoned for agricultural use when purchased, limiting the number of residential units to one for every five acres. Rather than fragment the forest with roads and buildings, the community preferred to develop residences mostly along a single strip. To do so, they had to overcome opposition to rezoning among neighbors, some of whom apparently feared larger-scale development than was actually planned. The group also changed its corporate form to reflect the abandonment of aspirations for commercial development. With this move, the group gave up any pretense of attempting to sustain themselves through land-based economic activities, completing a transformation from a collective economic endeavor to a residential community. The Maple community made these legal and organizational changes in conjunction with a reconsideration of membership goals. The dramatic drop in membership through 1980, from 40 or 50 shares to around 10, implied a steep increase in per member costs. The need to recruit new members seemed obvious, how many new members to recruit was less obvious. By 1981, Maple decided to limit the number of shares to fifteen. The self-imposed restriction on size allowed a combination of lowdensity settlement with high conservation of natural areas. Smaller size facilitated the maintenance of close interactions, an essential aspect of community that also promotes socialization. In exchange, Maple gave up potential decreases in per member costs. Recruitment procedures guarantee that new members understand the commitment they are taking on and share the central values of existing members. A candidate for membership must attend all community meetings - including council meetings, workdays, and community dinners - for at least three months before being allowed to join. In practice, people typically associate with the community for several years before they
Durability of Collective Action consider joining.12 Members expressed doubt that anybody would join with only three months of association. The legal structure gives the community as a whole the right of
first refusal when shares come up for sale. These restrictions on transfers of shares raise the cost of exiting the community, encouraging members to work together to address perceived problems. At the time of fieldwork, there was little turnover in shares. Membership had stabilized for more than fifteen years. Discussions of the most recent changes of membership suggested that new members were recruited quickly when shares became available. Moreover, financial demands are no longer so pressing. Members who have fully paid for their shares now receive dividends. The major impending challenge concerned the implications of aging and options for generational change in membership. Community members had not figured out how to address these concerns, but had begun a conversation about how best to do so. Considering that they are all able-bodied, they can debate options for years before being forced to make decisions. Overall, the Maple community is a success. The community developed institutions that were sufficiently detailed to guide them in dealing with difficult situations but sufficiently flexible to allow adaptation to changing needs. Most members could draw on the lessons of past organizational experiences. The founding members came from three distinctive backgrounds: former members of a failed commune, former members of an unsuccessful land-buying cooperative, and individuals with no prior experience with collective living. Current members are disproportionately from the first two sets. Some other members participate in other, non-residential, cooperative activities. The initial need to mobilize resources apparently resulted in overly lax attention to the need for a common vision for the group’s goals. In response to challenges, the group reorganized significantly. More selective procedures for recruitment and the introduction of an initiation period prior to membership ensure that more recent members agree on the main goals of the group while allowing plenty of room for diversity of views. Oak In 1968, a wealthy couple decided to found a less materialistic and less individualistic community. The Oak community never confronted the challenge of
The formal rules were probably more binding during the recruitment drive of the early 1980s.
Durability of Collective Action mobilizing resources to acquire land. The founders not only provided land, but also actively recruited members. In the early 1970s, the group formally organized itself as a
community church. Individuals gain access to community resources, especially its land, through membership to the church. In the early years, membership fluctuated dramatically. The fluctuations had a distinctive seasonality, peaking during summer months when many members camped on the property. During this period, there were sometimes as many as 300 members, but fewer than fifty at other times. The first crisis came when the community lost its patrons. The founders grew disenchanted with the project. They were disappointed with their failure to achieve a truly post-materialistic and group-oriented community. They donated the land to the church but no longer took active leadership roles. The loss of charismatic and enthusiastic leaders forced the community to take responsibility for its maintenance and growth. This meant devising new methods of mobilizing resources. Tied up with the various options were divergent notions about the proper direction for the community. Differences manifested as leadership struggles and factionalism. A particularly divisive conflict emerged in the early 1980s over management of community land. Some members wanted to sell plots cut out of the community’s communal land to individual members. Others felt that privatization would destroy the distinctive nature of the community and encourage individualistic behavior. The community ultimately decided to allow long-standing members to purchase parcels of community property. Members of the anti-privatization faction disputed the legality of selling off church property, but the courts allowed privatization to proceed. After five years of membership, individuals became eligible to purchase individual titles to parcels of community land at below-market prices. The community retains the right to buy back the property before it is offered for sale on the open market, but there are no restrictions on use. The fears of the anti-privatization faction proved to be well justified. Several people dropped out of the church after purchasing land. One individual clear-cut his parcel, thereby violating the community’s commitment to living with and protecting nature. Members of Oak found that they had no legal recourse. Leadership struggles, plus conflicts with former members over land use practices, continued through the 1990s. At the time of the most recent fieldwork, membership had a
Durability of Collective Action dropped to fourteen and the community had placed a moratorium on admitting new
members. They were clearly attempting to regroup. Despite their history of conflict, the remaining members resisted development of more formal institutions. Reliance on “likemindedness” to guide behavior appears to be a valued feature of the community, even if it had failed to protect other community values in the past. Under these circumstances, survival of the Oak community seemed uncertain. Twin Oaks Two friends interested in living in a more rural setting founded the Twin Oaks community. In 1973, the two purchased land at a favorable price for one person’s father. The pair held their property as tenants in common, a form of title that grants equal rights to all property holders. Any titleholder may make developments on the property, even without the consent of the other(s). The land did not have access to existing roads or connections to the water, sewage, and electrical networks. They could hike to the property and camp, but development of permanent residences required road access. In 1976, the pair bought a twenty-acre parcel that linked their land to the closest county road: one holds individual title to the fifteen acres that actually connect the road to the original property; the other holds title to the remaining five acres. To raise money for this purchase and development of a gravel driveway, they sold two parcels of twenty acres each to friends who shared their love of living in the woods. The new community members received individual titles to their land. A buy-back clause allowed the original members to repurchase the land at the purchase price plus six percent a year for the first six years – a below-market rate of appreciation. In addition, the founders sold parcels sandwiched between portions of the original property. The newcomers depended upon the original members to access the road. Both conditions raised the costs of leaving the community. In 1979, one of the new members decided to move to town. The founding members lacked the cash needed to take advantage of the buy-back clause, despite the below-market rate of appreciation. The departing member had no qualms about selling the property to an outsider. An informal personal loan from the other new member to the founders averted this threat to the community’s integrity. Some years later, the founders recognized this favor by granting title to the affected plot to the third member.
Durability of Collective Action This community consists of a small, informal group of friends. Group composition changed over the years as members married, divorced, or had live-in
partners, but never expanded to more than six people. A variety of property arrangements define legal rights to particular plots of land.13 Each household resides on private property. In practice, community members recognize zones of privacy in the immediate vicinity of residences, but share the entire area of undeveloped woodlands, regardless of legal title. Members walk their dogs and collect mushrooms throughout the territory. They value the sense of privacy associated with having most of their holdings undeveloped. No formal rules specify the size of private zones around residences or define rights of access or extraction. Nobody pushes for formalization and, in fact, at least one member denies the value of formal rules. The resulting ambiguities leave room for conflicts. For example, community members hold diverse opinions about hunting. All agree that the deer population has soared and probably suppresses saplings. One household allowed friends to hunt on the property in the recent past. The others have mixed feelings about hunting, in part out of distaste for killing deer, but mostly out of a concern for personal safety. All agree that everybody must be informed of the presence of hunters – or any other visitors. Community members have discussed, but not agreed upon, a requirement that hunters be accompanied at all times by a community member. In the absence of agreement, any household can permit hunting, despite the displeasure of any other household. Members also differ on the best way to address the lack of groundwater. Some households developed water collection ponds and installed cisterns. Others purchase tanks of drinking water. Cisterns require considerable maintenance, are not entirely reliable (e.g., during droughts), and can produce water with unpleasant aesthetic qualities (e.g., cloudy, unpleasant odor). Although they would have to bear the full cost of extending the existing pipe network, one household would like to connect to the county water supply. The others either feel ambivalently about this project or have no interest in it at all. With no consensus, nothing is done. If willing to bear the costs alone, either founding household could pipe water across the shared property, but must request right of
The original members redefined titles to the shared property to include their spouses.
Durability of Collective Action way to have pipes cross private parcels. The third household could only bring in piped water with the consent of at least one founding member. The other households are
unlikely to deny right of way. Nonetheless, lack of agreement magnifies the challenge of mobilizing sufficient resources, at least delaying high-cost developments of this nature. This community has confronted only a couple of crises, both related to the mobilization of resources. In the first instance, the group tapped additional resources by expanding their membership. There were no formal procedures to screen potential members or initiate them into the community. Rather, they supplemented the trust of friendship with deed restrictions and strategic location of plots to protect their interest in maintaining a larger block of undeveloped land. Later, when one newcomer decided to leave, both forms of protection were needed to keep the land within the community. Their success depended upon the fact that one member had the resources necessary to take advantage of the legal restrictions. No other situation has been serious enough to constitute a crisis. Community members are sufficiently likeminded that, although tensions sometimes simmer beneath the surface, they have not experienced any major fallings-out. Tulip Poplar A developer established Tulip Poplar in 1969. Tulip Poplar is a residential development built around man-made lakes. Members of the Tulip Poplar community include non-resident property holders, weekenders, and year-round residents. The centrality of the lakes to community identity and to the management activities of the neighborhood association gives this development a greater than usual sense of shared interest. As is typical with residential developments, the developer made the initial land purchase, carved out communal assets – including forested land as well as the lakes, and established a neighborhood association to manage communal property. Purchase of a private plot brings the right to use communal assets and be represented in the neighborhood association. Members did not have to organize themselves initially or mobilize resources to purchase or create their shared assets. They have had to sustain their organization and maintain their shared assets. A member handbook spells out rules for use of common assets. Rules related to use of the lakes are well defined; rules for use
Durability of Collective Action of forested commons leave many details to the common sense of community members. Committees within the neighborhood association take responsibility for specific maintenance tasks, such as keeping the roads clear of debris and overgrown plants. Tulip Poplar’s one major conflict related to maintenance of the lakes. Although
the lakes provide a common attraction for joining the community, members do not have identical interests in the lakes. Spatial distribution of plots relative to the lakes influences the benefits gained from each lake. Different forms of use, ranging from scenic value through fishing to boating, influence preferences about how the lakes should be maintained (e.g., desirable water level). On one occasion, disagreements about repairs for one lake spilled into the courts. The community accepted the court settlement without further ramifications. In particular, the conflict did not result in significant factionalism or a decline in membership. Tulip Poplar has more members, but a narrower realm of communal activity, than any other community in this study. The entrepreneurial interventions of the developer preempted the most serious crises confronted by most groups, those associated with mobilizing resources to purchase land, developing infrastructure, and designing institutions for self-governance. The institutions created by the developer proved incapable of resolving serious differences concerning shared assets, but stable enough to survive the conflict and the resultant intervention of government authorities. Despite – or maybe because of – the more superficial form of community pursued, this group survived a crisis that might have decimated a less resilient group. Comparative findings Each crisis that each community confronted represents a separate unit of analysis. Overall, this study includes twenty-six crises: five standard crises that each community had to overcome, plus the fire that destroyed Maple’s community building. By breaking the cases up so that each crisis is recognized as a separate observation, it becomes possible to capture fluctuations in each community’s size and different responses to particular crises. Thus, although the Maple community made significant changes in response to the destruction of its community building and the need to recruit new members in the early 1980s, it apparently made no significant institutional changes when confronted with earlier conflicts and problems of infrastructure development. Likewise,
Durability of Collective Action the Oak community handled successive crises and conflicts poorly at least through the
late 1990s, but its response to the most recent divisions remained uncertain at the time of the last study. The data set includes only two interval variables (community age and size, both with minimum values) and a relatively small number of observations for standard quantitative methods. I coded the non-interval variables either dichotomously (trial membership, formal socialization, adequacy of day-to-day performance14) or as ordinal categories (institutional definition, institutional flexibility, ease of exit, prospects for community reproduction, and change in capacity for survival). The histories of the cases as outlined above guided coding decisions. In some instances, when the exact date of a crisis was not reported, I estimated the age and size of the group based on indications of the general period during which the crisis unfolded (e.g., early 1980s as 1980). Table Four presents a sampling of conditions expected to influence responses to crisis. Because I lack data on prior organizational experience for most communities, I mention it in Table Four but do not include it in my analysis. [Table Four about here] Prospects for organizational change, community reproduction, and change in capacity for survival are all ordinal variables; ordered probit is the appropriate form of multivariate analysis. Organizational change is modeled as a function of age, size, institutional definition, institutional flexibility, the use of a trial membership period, the existence of a formal socialization process, and ease of exit. This model captures the influence of group characteristics (age and size) and institutional arrangements on immediate organizational responses to crises. The immediate response to crisis in terms of organizational change should influence a community’s prospects for both reproduction and survival of future crises. Thus, prospects for community reproduction and change in the capacity to survive are modeled as functions of age, size, institutional definition, institutional flexibility, the use of a trial membership period, the existence of a formal socialization process, ease of exit, and organizational change in response to the crisis. Table Five and Six reports results of the ordered probit analyses of organizational change in response to crisis and change in capacity for survival. The model of prospects
There was too little variation in the adequacy of day-to-day performance to use this variable in analysis.
Durability of Collective Action for community reproduction did not converge. Increases in group size and ease of exit from a community lower the probability that a community will adopt organizational changes in response to crisis. Organizational adaptation increases the probability of
improvements in the capacity for survival following a crisis, as expected. Improvements in capacity are also more likely for groups with more members, less well-defined institutions, less flexible institutions, trial membership periods, and high exit costs. Ease of exit emerges as the variable most strongly associated with poor responses to crises by these five communities. It has a strong direct effect on immediate adaptation and a strong indirect effect on capacity for survival of future crises. These patterns reinforce findings from previous case studies. Group size also influences responses to crises, but in an unexpectedly contradictory manner. Although smaller groups are more likely to adapt in response to crises, larger groups are more likely to emerge stronger from crises. It may be that larger groups have less organizational flexibility, but that larger numbers translate into resources that help communities survive crises. There may also be endogeneity problems, since crises often prompt drops in membership. The positive relationship between trial membership and improvements in capacity was expected. More puzzling are the negative relationships of both institutional definition and institutional flexibility with improvements in capacity. If a balance must be struck between institutional definition and flexibility, there may be non-linear relationships or interaction effects that the current models do not capture. Despite the high level of diversity across cases, all five communities had survived for at least twenty years at the time of the most recent field visit. Even the weakest of the five, Oak, had entered its fourth decade. In line with expectations from organizational ecology, these communities, having overcome the initial challenges of acquiring land, recruiting members, and developing the initial infrastructure, had relatively good prospects for continued survival over the medium-term. Based on the experiences of these five communities, the minimal conditions for surviving initial crises are not formidable. During their formative years, these groups had fluctuating memberships, some had very few members while others were quite large, some had very flexible and rather poorly defined institutions, and a number had relatively minimal conditions for
Durability of Collective Action
entry and exit. Despite theories identifying these conditions as problematic, it seems that none preordains a community’s demise. Organizational ecologists expect the risk of collapse to increase again as organizations begin to experience obsolescence. Kitts (2001) predicted obsolescence to occur only after 100 years, whereas the oldest community in this sample is a relatively young 35 years old. The differences in perceived prospects for survival among these communities suggest that mid-life challenges exist, but are not uniform. The 100-year life span may apply to relatively large organizations that operate in a manner comparable to firms in a competitive market. Box Elder and Tulip Poplar fit this description, and it is easy to imagine that they will continue to survive for several generations. Issues related to old-age confront smaller organizations with less membership turn-over earlier. Organizational problems, on the other hand, sometimes remain unresolved at mid-life, as in the Oak community. Weak organizations survive beyond the founding crises only under special circumstances, as in the presence of an external sponsor. For smaller intentional communities, organizational old-age may set in around fifty or sixty years.15 Barring significant organizational changes in the medium-term, the life-spans of Maple and Twin Oaks are bound up with the life stages of their current members. Both communities feature relatively small size, lack generational diversity, and decided not to expand their membership. Neither community has agreed upon arrangements for generational change. Maple has slightly more members with slightly greater generational diversity. If departures of current members are at least somewhat staggered, Maple’s current arrangements for recruitment may support incorporation of a new generation. Several members have adult offspring with some interest in the community; generational change may occur at least partially through inheritance. Members of Twin Oaks have neither offspring nor clear plans for their estates; it is not at all clear whether this community will survive beyond the current members. Aging will become a more pressing issue for Maple and Twin Oaks over the next ten to fifteen years. Oak survived its founding period on the basis of external assistance rather than its own organizational strength. Its patrons secured land, recruited members, and helped
The expected life span is based on the observation that intentional communities tend to be founded by people in their twenties.
Durability of Collective Action provide the initial infrastructure. The community survived these crises despite low institutional definition, high institutional flexibility, and considerable membership
volatility. Only since the withdrawal of its sponsors has the Oak community survived on its own merits. It has not fared terribly well. As a result of on-going factionalism, its membership declined, it lost property, and its assets are not being maintained. Oak’s continued survival depends on the success of efforts to regroup underway at the time of the last study. These findings are suggestive, but far from conclusive. The study includes five of six communities within the vicinity of Indiana University that own some shared forested land.16 Although this amounts to fairly comprehensive coverage of a particular type of community within a particular region, it is not a random sample. Only a small number of communities are represented, with several data points drawn from each. Most of the communities were founded between the late 1960s and the mid-1970s. It is impossible to know whether characteristics linked to the time, place, or mode of land ownership correspond with reactions to crises. The analysis cannot be generalized beyond this nonrepresentative sample. Moreover, this is a censored sample; it does not include any instances of complete collapse. History engages in a non-random sampling process. Any community that failed to meet initial challenges did not survive long enough to be included in our sample. It is not possible to sample failed communities directly unless the research team studies new communities in the process of forming, some of which subsequently fail. These would have to be contemporary efforts at organization, and there may be issues of temporal bias if the challenges of organizing change with the zeitgeist. Previous communities that failed during their founding periods could be studied if data were drawn from archival materials and interviews with former community members. Again, selection bias is likely since failed communities differ in the extent to which they are known to current researchers, leave a public record, or have former members who can be identified and are willing to be interviewed. The IFRI research team is aware of the inherent biases in this sample; ideas about ways to supplement existing data have been discussed. Meanwhile, the findings of the
The sixth community refused to participate in the study.
Durability of Collective Action current study represent an advance beyond analyzing a single case study or comparing two or three communities. The temporal depth of the data, based on two visits to each community, strengthens my confidence in the validity of the findings. The consistency between my findings and those of previous scholars – especially concerning the
importance of exit costs – suggests that some patterns may be general. On the other hand, the hints of divergent types of community organizations with different life stages points to a new direction for research. If future studies confirm the existence of divergent types of communities with different life spans or stages, distinguishing among types will advance our ability to understand variation in organizational longevity.
Durability of Collective Action Table One: Characteristics of Five Communities in Southern Indiana Characteristic Purpose Box Elder Privacy Spirituality Gathering space Founding Land acquired Group size: range Group size: most recent (year) Total land area (ha) Communal area (ha) 18 90 213 110 128 80 109 304 449 1100 140 200 (2001) 1983 1987 200 – 500 197317 1975 About 25 – about 145 39 (2000) 14 (1999) 1450 (2003) 4 (2002) 1968 1968 14 - 300 1969 1969 25 - 1450 1970 1970 2-6 Maple Community Stewardship of land Oak Community Gathering space Tulip Poplar Recreation Residential development
Twin Oaks Privacy
The community was officially founded in 1976, when the first residents settled on the land. Planning began in 1973. 18 Forested land, with the exception of the Maple community, which has 200 hectares of forest and a 13 hectare meadow.
Durability of Collective Action Table Two: Crises and Responses by Five Southern Indiana Communities
Crisis Purchase of Land
Box Elder Festivals Recruitment Sale of land
Maple Sale of shares in organization Fees Workdays
Oak Grant from founder
Tulip Poplar Provided by developer
Twin Oaks Partnership Recruitment Sale of land
Infrastructure Festivals development Work-days
Provided by developer
Recruitment Costsharing or Individual
Friendships bolstered by land titles
Festivals Tiered system of costs
Open, low cost
Open, low cost
By purchase of individual plot Automatic membership in association
Trial period prior to
Probationary period of limited membership Current moratorium Flexible internal rules Court involvement
By purchase of individual plot Automatic membership in association
Friendship networks Buy-back clause in title deed
Demonstrated membership commitment Right of first
for leadership refusal on positions Conflict resolution Flexible internal rules Court involvement transfers Flexible internal rules
Formal internal Few, procedures Court involvement informal rules Iterated interaction
Durability of Collective Action Table Three: Responses to Crises by Five Southern Indiana Communities
Crisis Organizational Change Day-to-Day Performance Prospects for Reproduction Capacity to Survive Future Crises
Box Elder None to Minimal Adequate
Oak In process (significant)
Tulip Poplar Twin Oaks None None
Promising, but uncertain
Durability of Collective Action Table Four: Selected Conditions affecting Responses to Crises19
Condition Modal size20 Age at land purchase Institutional definition Changes in institutional definition Institutional flexibility Trial membership
Box Elder Large 4 years
Maple Moderate 2 years
Oak Moderate 0 years
Tulip Poplar Large 0 years
Twin Oaks Small 0 years
Yes – basic membership
Yes – association prior to membership
Yes – sponsorship period
Increases with level of responsibility
Initially low, now extensive Low Most with relevant experience
Initially low, now extensive High ?
Moderate, ability to purchase Moderate ?
Informal, based on friendship Low ?
Ease of exit Past experiences of members
High Leadership with relevant skills
Unless otherwise specified, the summary reflects conditions during the most recent round of data collection. 20 Modal size refers to the number of members during most years in the group’s history. Small groups have no more than 30 members, moderate groups have more than 30 but fewer than 100 members, and large groups persistently have more than 100, and usually several hundred, members.
Durability of Collective Action Table Five: Ordered Probit of Organizational Change in Response to Crisis Coefficient21 -0.01574 -0.02368* -1.90504 -4.76221 1.57838 3.32628 1.55931* -10.20422 -8.69814 -7.95634 = 25 = -13.4727 = 27.80 = = 0.0002 0.5078
Variable Age Size Institutional Definition Institutional Flexibility Trial Membership Formal Socialization Ease of Exit Cut 1 Cut 2 Cut 3 N Log Likelihood LR chi-squared Prob > chi-squared Pseudo R-squared
Standard Error 0.07121 0.00996 1.52760 3.74405 0.93421 1.84665 0.62251 11.7521 11.5449 11.5834
Statistical significance at 0.05 level indicated by a single asterisk.
Durability of Collective Action
Durability of Collective Action Table Six: Ordered Probit of Change in Capacity of Community to Survive Coefficient22 -0.19669 0.00920* -5.56838* -16.27165* -6.64345* 3.73678 -4.74823* 3.02971* -60.07974 -51.61856
Variable Age Size Institutional Definition Institutional Flexibility Trial Membership Formal Socialization Ease of Exit Organizational Change Cut 1 Cut 2
Standard Error 0.11704 0.00418 2.39660 6.97760 3.39601 16.65094 2.18348 1.53588 30.6398 28.0344
N Log Likelihood LR chi-squared Prob > chi-squared Pseudo R-squared
= 26 = -5.92223 = 22.79 = = 0.0037 0.6580
Statistical significance at 0.05 level indicated by a single asterisk.
Durability of Collective Action
Durability of Collective Action Works Cited
Axelrod, Robert. 1984. The Evolution of Cooperation. New York: Basic Books. Baland, J-M. and Platteau, J-P. 2000 . Halting Degradation of Natural Resources: Is There a Role for Rural Communities? New York: Oxford University Press. Brown, Susan Love, ed. 2002. Intentional Community: An Anthropological Perspective. State University of New York, Albany, New York. Fellowship for Intentional Communities. Nd. Intentional Communities Web Site. Most recently accessed, December 2003. Gibson, Clark C. and Tomas Koontz. 1998.“When 'Community' is Not Enough: Institutions and Values in Community-Based Forest Management in Southern Indiana,” Human Ecology 26, no. 4: 621 - 647. Hirschman, Albert O. 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Cambridge: Harvard University Press. Kanter, Rosabeth Moss. 1968. “Commitment and Social Organization: A Study of Commitment Mechanisms in Utopian Communities,” American Sociological Review 33 (August): 499 – 515. Kanter, Rosabeth Moss. 1972. Commitment and Community: Communes and Utopias in Sociological Perspective. Cambridge: Harvard University Press. Kitts, James A. 2001. Community Ecology: Organizational Dynamics and the Mortality of American Communes, 1609 - 1965. Unpublished Ph.D. dissertation, Department of Sociology, Cornell University. Mariampolski, Hyman. (1977). The Dilemmas of Utopian Communities: A Study of the Owenite Community at New Harmony, Indiana. Unpublished Ph.D. dissertation, Department of Sociology, Purdue University. McKean, Margaret A. 1992. “Management of Traditional Common Lands (Iriaichi) in Japan,” in Daniel W. Bromley, ed., Making the Commons Work: Theory, Practice, and Policy. San Francisco: ICS Press. Michels, Robert. 1959. Political Parties. New York: Dover Publications. Olson, Mancur. 1982. The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities. New Haven: Yale University Press.
Durability of Collective Action Olson, Mancur. 1971. The Logic of Collection Action: Public Goods and the Theory of Groups. Cambridge, MA and London: Harvard University Press. Ostrom, Elinor. 1999. “Self-Governance and Forest Resources,” CIFOR Occasional Paper no. 20. Bogor, Indonesia: Center for International Forestry Research (CIFOR). Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. New York: Cambridge University Press. Ostrom, Elinor and Mary Beth Wertime. 2000. “International Forestry Resources and Institutions Research Strategy,” in Clark C. Gibson, Margaret A. McKean, and Elinor Ostrom, eds., People and Forests: Communities, Institutions, and Governance. Cambridge, MA: The MIT Press. Poteete, Amy R. and Elinor Ostrom. Forthcoming. “An Institutional Approach to the Study of Forest Resources,” in John Poulsen, ed., Human Impacts on Tropical Forest Biodiversity and Genetic Resources. New York: CABI Publisher.
Poteete, Amy R. and David Welch. Forthcoming. “Institutional Development in the Face of Complexity: Constructing Systems for Managing Forest Resources,” Human Ecology. Sandler, Todd. 1992. Collective Action: Theory and Applications. Ann Arbor: University of Michigan Press. Shenker, Barry. 1986. Intentional Communities: Ideology and Alienation in Communal Societies. London and Boston: Routledge & Kegan Paul.