Department of Political Science
University of Colorado-Boulder
Conflict Resolution Consortium
Campus Box 327
University of Colorado
Boulder, Colorado 80309
Working Paper #90-5
This paper was written with a small grant from the Conflict Resolution Consortium, University of Colorado. Funding for the Consortium and its Small Grants Program was provided by the William and Flora Hewlett Foundation.
The statements and ideas presented in this paper are those of the author(s) and do not necessarily represent the views of the Conflict Resolution Consortium, the University of Colorado, or the William and Flora Hewlett Foundation.
CONFLICT RESOLUTION CONSORTIUM
Funded by the William and Flora Hewlett Foundation and the University of Colorado, the Conflict Resolution Consortium is a coordinated program of research, education and application on three of the University's four campuses. The program unites researchers, educators, and practitioners from many fields for the purposes of theory-building, testing, and application in the field of conflict resolution. Current focus areas include international conflict; environmental and natural resource conflict; urban, rural, and inter- jurisdictional conflicts; and the evaluation of alternative dispute resolution mechanisms.
The Conflict Resolution Consortium working paper series includes a variety of papers written by our members as a part of their research. Usually these papers are in prelimi- nary draft stage and are being prepared for eventual publication in professional journals or books. Other papers record discussions from Conflict Resolution Consortium seminars and plenary presentations.
The purpose of the working paper series is to generate a dialogue about the work presented. Readers are encouraged to respond to the papers either by contacting the author directly or by contacting the Consortium office.
Additional copies of this or other working papers can be obtained from the Conflict Resolution Consortium, Campus Box 327, University of Colorado, Boulder, Colorado, 80309-0327. Phone (303) 492-1635.
Some of the central questions in the social-psychological study of social movements and collective action concern the circumstances and motivations of individual participants: who joins, who persists, and why. This research project is designed to focus on these questions as they pertain to participants in the Sanctuary Movement, a movement organized to protest U.S. policy toward Central American refugees.
One prominent method in the attempt to understand collective action at the individual level has been through rational choice models. The most plausible modification of rational choice theory to explain participation in this type of protest activity is the "public goods" model of collective political behavior developed by Muller and Opp (1986). This model addresses some of the obvious shortcomings in the classical rational choice model's ability to explain collective action in which material self-interest cannot plausibly be cited as the primary motivator.
The results of this study show that while the "public goods" model is currently the most applicable in studying this type of collective behavior, there remain some underlying problems in its formulation. Specifically, Muller and Opp argue that "a basic indicator of the public goods value of rebellious collective action is the extent to which an individual is hostile to or alienated from the political system in general." The findings here contradict this premise and argue instead for a more direct valuation of the public good based upon discontent with a particular policy, in this case U.S. policy toward Central American refugees.
The Sanctuary movement began in 1980 when the morally outraged congregations of two Tucson churches provided bail and sanctuary for the 13 surviving members of a group of 26 Salvadorans abandoned in the Arizona desert by a professional smuggler, or "coyote." At the same time, several congregations in the San Francisco/Berkeley area who were aware of the problems facing Central American refugees had decided to become places of sanctuary. Efforts were coordinated, and on March 24, 1982, the second anniversary of the assassination of Salvadoran Archbishop Oscar Romero, these churches as well as one in Washington, D. C. publicly declared themselves sanctuaries for Central American refugees. Since then the movement has grown to include over 400 sanctuary churches and 2000 support congregations throughout the United States. These sanctuaries are usually, but not always, churches or synagogues of many different religious denominations, including Catholic, Presbyterian, Methodist, Jewish, Quaker and Mennonite.
Members of this movement provide sanctuary and support for Salvadorans and Guatemalans who have fled their strife-ridden countries. Hundreds of thousands of Central Americans have come to the United States since the early 1980s in the wake of massive counter-insurgency campaigns carried out by military governments in the region. The most intensive of these have adopted slash-and-burn and strategic hamlet techniques, such as those used in Vietnam, which take a heavy toll on rural civilian noncombatants. Since current governmental interpretation of immigration policy categorizes the vast majority as economic rather than political refugees, movement members argue that they are in danger of deportation and deadly persecution in their home countries. Sanctuary Movement participants argue that it is their moral, as well as legal, obligation to protect refugees from the unjust policy of the U.S. government. Through conventional forms of protest as well as unconventional forms, such as illegally transporting and harboring refugees, movement participants put themselves in direct conflict with government policy.
Sanctuary is a distinctive type of social movement. Its members are committed to direct action on behalf of political refugees, and accept potentially substantial costs, and in many instances the risk of prosecution, because of normative commitments rather than the pursuit of material or political gains for themselves or their group. Gamson characterizes this type of group as "universalistic," where the major beneficiaries of the group's actions and its constituent members are not identical. (Gamson, 1975: 61-62.) While some discussion of motivations at the group level is included, the questions addressed in this work focus primarily on the individual motivations of participants in universalistic movements, specifically in this case the Sanctuary Movement. What calculations do members con- sider in making their decision to join such a group?
In order to test the "public goods" model, personal and survey interviews were conducted with members of the Sanctuary Movement across the country. Two pilot surveys were conducted in Denver and Boulder, Colorado, the results of which were used to develop the survey instrument distributed in Denver, Boulder, Austin, Tucson, Minneapolis, Washington, D.C., Chicago, Los Angeles, San Francisco and Berkeley. The personal interviews, approximately 100, have provided con- siderable information concerning the background and differ- ent considerations facing congregations at the time of their decisions to declare sanctuary. This research was conducted between June 1987 and April 1990.
The responses used to test the Muller and Opp model will be drawn from the 104 written surveys which have been returned to date. The survey questions focus directly on the participant's perceptions of his or her goals and motivations for joining the movement, feelings of efficacy, the political context in which he or she is acting, and dynamics within the group. Generally, the research is designed to examine: why actors participate in universalistic move-
 The potential legal penalty for assisting undocumented refugees is five years imprisonment and a $10,000 fine for each refugee. These penalties may be compounded by a conspiracy charge which carries a maximum penalty of ten years in prison. While only a few individuals have actually been sent to prison for Sanctuary activity, the Tucson trials of 1984 as well as several other unsuccessful attempts by the government to bring charges against individuals involved in Sanctuary make these potential costs a serious factor in the decision calculation of movement participants.
 Due to the sensitive nature of the questions in this survey some difficulties were encountered in collecting data. Respondents were asked about illegal activities which they had taken on behalf of refugees as well as legal penalties for those actions. The response rate, twenty-three percent is below average for a mail survey. Although survey responses were anonymous, hesitance on the part of respondents is understandable in light of various attempts by government agencies to infiltrate this movement. Four hundred fifty surveys were distributed in all, three hundred to participants in the Sanctuary Movement and one hundred fifty to non-participants serving as controls. ments; what types of cost/benefit calculations, if any, they consider in making their decision to join such a group; and ultimately the overall dynamics and influences which determine the type and extent of their participation.
Understanding the dynamics of collective action has long been a source of fascination, and heated debate, in political science, sociology, economics and anthropology. It has been studied as both a group-level and an individual-level phenomenon. At the group level, resource mobilization and relative deprivation theories have provided the most compelling explanations and the greatest fuel for debate. Resource mobilization theories argue that the critical factor in collective action is a group's ability to mobilize the resources necessary to take effective action (Tilly: 1978; Jenkins: 1983; Oberschall: 1973; Gamson: 1975). From this perspective, the ability to act comes from a change in the material or non-material resources of a group, which may include financing, material support, communications or even expertise.
Others have expanded on this to illustrate further the notion of resources, including the opportunity to act, a perception within the group that action has the potential to succeed, and the ability of the group to mobilize consensus, raising expectations that others will likewise act (McAdam: 1982; Zald and McCarthy et al.: 1979; Klandermas: 1984). "More often, when events and organizers mobilize people, it is because they build solidarity, raise consciousness of common interests, and create opportunities for collective action" (Fireman and Gamson: 1979). The problem is not finding shared grievances, it is the capacity to act on them collectively. Many of the resource mobilization theories came as a reaction to more social-psychological explanations of collective action and rebellion which argued that the fundamental impetus for action lie in a frustration-aggression mechanism.
Theories of relative deprivation argued that the critical factor in why men rebel could be found in reactions to the thwarting of goal realization. This may be caused by a period of rising expectation, which could be achieved, followed by a sharp downturn in the ability to attain these goals (Davies: 1962; 1969), the degree to which social conditions create a gap between "want formation" and "want satisfaction" (Feierabend and Feierabend: 1969) or where there exists a perception of unequal potential among groups or individuals to reach goals which they feel they can or should be able to achieve (Gurr: 1970).
While these theories provide significant insight, some have argued that their aggregate nature wash over or mask motivations at the individual level. An alternative method, one which focussed on individual-level analysis, is the rational choice approach. A number of theorists approach questions of collective action with assorted versions of the Downsian model of rational choice whereby people are assumed to act in calculated, efficient ways to achieve ends that are in their personal interest (Downs: 1957). The Downsian model implies economically efficient, material self-interest as the primary motivator of action. Oberschall argues that "the simple assumption of rationality in economic theory is sufficient in this theoretical effort" (Oberschall: 1973, p. 118).
Yet Mancur Olson, in his seminal work The Logic of Collective Action (1968) argues that members of a collectivity will not act unless they receive incentives or sanctions, since the benefits from a given public good will be theirs regardless of whether they participate (incur the costs) or do not participate in the activity necessary to bring about that public good. (Hence the problem of "free riders.") Selective incentives are therefore necessary to induce action.
Almost since its inception, however, many scholars have taken exception to classical rational choice's limitation of incentives to material, economic ones. Tullock (1971), for example, distinguishes between material and nonmaterial incentives, arguing that selective incentives may include such rewards as power, status, entertainment or fulfilling a sense of duty. Others have included the expected reactions of "significant others," whose opinions affect one's perception of the costs and benefits of action (Freeman: 1979; Tolbert, 1981).
While these theories provided greater explanatory power for many movements which could not be adequately explained by classical rational choice models, there were still a good number of movements which required even further refinement of the model. In contrast to the conventional private interest model of rational choice, Muller and Opp developed a public goods theory of rebellious collective action which contends that non-material, non-private (public goods) incentives are the critical motivating factors in particular types of protest (1986). In their study of anti-nuclear protestors in Hamburg, West Germany and New York City, they examine rational choice models, including various selective incentives, and conclude that "...in contrast to conventional rational choice models of rebellion, the public goods variable, not selective incentives appear to be the most important incentives for performing rebellious political behavior" and that "expected costs do not appear to be a disincentive, and they may even have some weak incentive value" (p. 483). As will be discussed later in this article, the survey questionnaire is designed to elicit information which will allow the testing of this model in the case of the Sanctuary Movement.
As others have done in the past (LeBon: 1903 (1968); Hoffer: 1951; Kornhauser: 1959), Muller and Opp base their model on the premise of protestors' alienation from the political system. Indeed they define their key variable, the public goods variable, by the degree to which an individual is alienated from the political system and thus considers the attainment of goals achieved through protest to be a public good (p. 477). In previous work, Muller and Jukan argue that while a variety of grievances may contribute to participa- tion in civil disturbances, "apparently, alienation from the political system is the only common denominator among them" (1983). This basic assumption, however, has itself been a basis of contention. Studies of the civil rights movement (Morris: 1981), political protest movements (Marsh: 1979; Barnes and Kasse: 1979), environmental, anti-abortion, and anti-nuclear movements (Leahy and Mazur: 1978) have all found evidence contrary to the alienation hypothesis. As will be shown, this study supports findings against the necessity of alienation in the mobilization of protest action.
What limited literature there is concerning the Sanctuary Movement in the United States focuses primarily on group-level analysis. In order to gain the type of information necessary to test an individual-level model such as Muller and Opp's, an entirely new data set had to be collected.
A survey questionnaire was designed to answer the basic questions posed. Specifically, the survey targets the participant's personal motivations and goals within the Sanctuary Movement as well as his/her perceptions of the political context in which he/she is acting. Caution must always be taken when using an instrument such as a survey for research. Because of the vulnerability to selective perception, aggrandizement or subjectivity bias, surveys must be designed with care to minimize the potential for misrepresentative responses. This is certainly no less true in the case of Sanctuary. In addition, the sensitive nature of the questions being asked in this particular survey, eliciting discussion of illegal activities and knowledge of potential legal penalties, requires that special measures be taken in the effort to secure the most honest and representative responses possible.
With this in mind, the survey used was designed so that each of the major elements of the Muller and Opp model could be  For an exception see Davidson, Miriam, Convictions of the Heart: Jim Corbett and the Sanctuary Movement (University of Arizona Press, Tucson: 1988). addressed through several different questions. In the effort to tap the degree to which a given respondent is alienated from the political system, for example, three separate questions were used in different parts of the survey. This method cuts down on the possibility of biased or inconsistent answers leading to erroneous conclusions. Moreover, careful steps have been taken to ensure the anonymity of every respondent, further encouraging the maximum degree of honesty in responses.
One problem which has arisen in the effort to test this model has been the determination of a control group against which participants in the Sanctuary Movement can be compared. In fact, this has presented something of a dilemma, since each of the possible groups which may be considered as constituting a "control" presents its own methodological problems.
In the effort to isolate theoretically, as well as methodologically, who exactly should serve as a control group several options were considered. Should the entire population of this country who do not participate in Sanctuary be considered a control group? Certainly not, for this would imply that participation in Sanctuary is a choice which has confronted everyone in the United States, and moreover that there is a dichotomous choice between participation and non- participation. Given all of the possible activities in which people participate, such a sledgehammer approach would tell us little about the differences which make participants in this movement unique.
Another possibility is to look at religious congregations of the same denominations who do not participate in Sanctuary. This would seem a more plausible alternative since much of the justification for participation rests on religious conviction, morality and value structures which are largely shaped by religious culture. Since the majority of participants in Sanctuary are affiliated with churches, it would seem that a common value structure would provide a reasonably strong basis of comparison. One drawback with this option, however, is that in using non-participating congregations of the same religion as the control group, we are assuming a priori that it is in fact the religious or congregational context which is the determining element for the decision in the first place. Isn't it possible that activists in this movement bring their consciousness and drive for action from somewhere outside the church, and that the congregation provides a network and support structure where by these ideas can be put into practice? Questions included in the survey which address the original motivations of participants would seem to indicate that in fact a large part of the basis for action does come from the value system taught within the church context. A sound methodology, however, should assume nothing away beforehand. As with the previous example, using these congregations as controls also imposes a false dichotomy of choice between participation in Sanctuary in particular and non-participation. Many churches participate in other activities, such as care for the homeless or shelters for abused women. In short, these congregations may well have nothing against Sanctuary, but one congregation cannot do it all.
For this study it was determined that the best option for a control group comprised those individuals who are congregational members of declared sanctuaries, yet have chosen not to be participants in the Sanctuary Movement. Control surveys were therefore sent to sanctuary congregations along with participant surveys. Several problems arose in this effort. First, many of the congregations who did vote to become sanctuaries did so through secret ballot, so that each individual might vote his conscience without pressure from the group. Consequently, pastors are reluctant to ask church members how they voted for fear of alienating any member or raising a divisive issue. Within those congregations who did not have secret ballots, many of the members who were against participating in Sanctuary left the church after sanctuary was declared. Alternatively, several instances were recounted of individuals who were originally against participation, but who have changed their minds. Finally, since many of these congregations declared sanctuary as many as nine years ago, pastors often cannot remember who voted which way.
Despite the problems involved in isolating the "perfect" control group, it is still important to have some basis of comparison in order to gain insight into the particular motivations which drive those who do in fact participate in the Sanctuary Movement. Rather than using a strict comparison of participants versus non-participants, therefore, an ordinal scale was developed based upon the degree to which individuals declared themselves to be involved in movement activities. The scale ranged from 1 to 5: 1=leaders; 2=very involved; 3=somewhat involved; 4=occasionally involved; 5=non-participant. The distribution of the respondents was: 1 (37%), 2 (25%), 3 (20%), 4 (8%), 5 (10%).
Muller and Opp's full rational choice model of rebellious collective action, including public goods and soft selective incentives is written as:
E(R)=(pg + pi)V + E(F) + E(A) + E(O) - E(Cr)
 The results were somewhat disappointing. Of 150 control surveys sent, only 10 were returned. where E(R) is the expected utility of rebellious collective action; V is the value of rebellious collective action in terms of public goods, pg is the expected group influence on the provision of the public good; pi is the expected individual influence on the provision of the public good; E(F) is the expected entertainment (fun) value of participation in the rebellious collective action; E(A) is the expected social affiliation; E(O) is the expected value of conforming to behavioral norms of important others; and E(Cr) is the expected private costs to the individual of participating in rebellious collective action.
Muller and Opp argue that, contrary to Olson's logic, individuals do not necessarily make cost/benefit calculations based solely upon the objective potential of individual efficacy (pi), assumed in standard rational choice models to be negligible. Rather, they base calculations of individual efficacy upon the realization of individual contributions to the success of the group: collective rationality rather than individual rationality. "It follows that collectively rational individuals might estimate pi to be significantly greater than zero because they recognize that free riding is collectively irrational in the case of rebellious collective action, where group size and cohesion can be a factor of critical importance. And if pi cannot be assumed necessary (sic) to be negligible, then the public goods value of rebellious collective action becomes a relevant considera- tion in the decision calculus of a rational average citizen." (p. 473).
The results of this survey support Muller and Opp's findings as shown in Table 1. When asked, "How effective do you feel the efforts of a single individual such as yourself can be in influencing U.S. policy on Central American refugees?" the bulk of participants, 71.4% gave answers indicating a perception of minimal ('not effective') to only moderate (`somewhat effective') individual efficacy. Only 28.6% of the participants perceived themselves to have strong indi- vidual efficacy.
When asked, "How necessary do you feel acting with a group is for influencing U.S. policy on Central American refugees?" 87.1% responded that it was `critical to success,' 12.9% responded that it was `helpful but not necessary,' and none of the respondents answered that it was either `not very helpful' or `creates obstacles to success.' Clearly these individuals perceive group action as much more efficacious than individual action.
To test Muller and Opp's hypothesis, however, it is necessary to analyze how any given participant evaluates individ- ual action in relation to group action. Table 1a records the cross tabulation of the two questions as answered by each participant. For example, only 9.6% the respondents answered both that individual action was very effective and that group action was critical to success. In contrast, the largest group, 31.3%, answered both that individual action is somewhat effective and that group action is critical to success. The responses on Table 1a are divided between those who could be considered to be acting on individual rationality in terms of efficacy (high individual efficacy and high group efficacy) and those who could be considered to be acting on collective rationality (moderate to low individual efficacy and high group efficacy). There were no cases where the respondent indicated high individual efficacy and low group efficacy. The vast majority, 72.3%, can be said to be acting on collective rationality. These results support Muller and Opp's hypothesis that when individuals, in this case Sanctuary members, act on the basis of "collective rationality" the individual efficacy calculus is greater than zero (pi>0).
Moving to the selective incentives in the public goods equation, Muller and Opp find that neither the "entertainment" value nor the affiliation rewards are statistically significant in determining rebellious collective behavior. The results of the survey support their findings on both counts. Only one of the respondents answered that enter-tainment, or in his word "adventure", was a motivating factor in participation. Likewise, affiliation rewards did not prove to be a statistically significant motivator, yet it was one of the more frequently mentioned (along with a sense of duty and desire for justice). Thirty-five percent of the respondents indicated that affiliation with the particular church group and their activities was a motivator of action. The most common sentiment expressed in this regard was one of solidarity with other members of the church or the Sanctuary Movement.
As regards E(O), the value of conforming to the behavioral norms of important others, Muller and Opp found this to have a small, but statistically significant effect. Replicating the question in this survey, the findings of this survey are that E(O) is a significant predictor or motivator of action  This finding also concurs with a follow-up study done by Finkel, Muller and Opp (1989).  Respondents were asked to discuss their reasons for joining the Sanctuary Movement. From this open-ended question a nominal breakdown was obtained of selective incentives which included entertainment and affiliation rewards, sense of duty, pursuit of justice, following religious convictions and for non-participants fear of legal penalties. While affiliation, duty and justice were mentioned by 35, 26, and 23 percent of the respondents respectively, none constituted a significant motivator for the movement as a whole. This finding appears to contrasts with what Finkel, Muller and Opp expect in their "collective rationality" model (1989) but this may be an artifact of the coding scheme. yet the relationship is negative. From the coding scheme, degree of participation is ranked in ascending order, such that leaders are ranked '1' and non-participants '5'. The negative relationship found for important others thus indicates that the more involved an individual is in the movement, the greater the impact from expectations of important others. These findings support the influence of "psychic income," rewards gained from conforming to the expectations of those one views as important, found in other studies (Silver, 1974; Opp, 1986).
Addressing the final variable in the equation, Muller and Opp find that expected costs serve not as a deterrent, but as a mild incentive for rebellious collective action. "Surprisingly, the expected utility of negative sanctions, E(Cr), is estimated to have a significant negative instead of positive effect. Since costs generally are negatively valued, this means that respondents who attach a relatively high probability to the occurrence of negative sanctions are somewhat more likely to have a high rebellious behavior potential than those who attach a low probability to the occurrence of negative sanctions." (p. 482.)
Respondents in this survey were asked both whether they were aware of the legal penalties for their actions, and their perception of the probability of incurring those penalties on a six-point scale ranging from 'impossible' to 'certain'. To concur with Muller and Opp's hypothesis, we would have expected that the more involved an individual was in the Sanctuary Movement the greater the probability he would attach to incurring legal penalty. Instead, the results from this question supported a more traditional view of the effects of potential costs. That is, the more involved the participant was in the movement, the less likely he was to attach a high probability to suffering legal consequences. (Table 2) Tobit results show the relationship to be positive and highly significant.
It is interesting to note that the perception of legal penalties is one of the few clear differences found in this study between participants and non-participants in the Sanctuary Movement. While the majority of participants, ranging from leaders to .pa only occasionally involved, felt that prosecution was 'possible but not probable' non-participants perceived a minimum probability of `possible' with several choosing `certain.'
 Tobit is produced by TSP International statistical software and is analogous to Probit and Logit techniques. The advantage of using Tobit in this study is that it allows for ordinal and nominal data with which is not dichotomous and whose values are either zero or positive. Strong support for the cost-as-incentive hypothesis, however, was demonstrated through another series of questions. Regarding the 1984 Tucson trials, in which sixteen Sanctuary workers were indicted and eight convicted of transporting and harboring undocumented refugees as well as conspiracy, respondents were asked whether they felt these trials had an incentive or a disincentive effect on members and Sanctuary activity. The results from these questions are reported in Table 3.
The first question in Table 3 addresses the effect of the Tucson trials on individual motivations of participation. Forty-two percent of the participants responded that the trials had actually increased their participation in the movement. Fifty-eight percent responded that the trials had no effect on their participation. When asked directly 'How have the trials affected your feelings concerning the legal penalties which you may face?' the vast majority, 79.2%, said that they had 'no effect.' Ten point four percent said that the trials had indeed increased their fear of legal penalties while another 10.4% responded that the trials had decreased their fear.
As was to be expected non-participants as a group were proportionately more likely to say that the trials increased their fear of legal penalties: 30% gave this response. (Table 3a) Still, the majority of non-participants (40%) said that the trials had no effect. This is not surprising since non-participants perceived a higher probability of legal action to begin with.
These responses of movement participants suggest that for these individuals even the physical manifestation of costs in the form of trials (as opposed to the abstract potential) lacked the disincentive value that traditional rational choice models would predict. While some might argue that this result is a product of hindsight or selective perception based upon the fact that the defendants were not sent to jail, anecdotal evidence further bolsters the findings. Sanctuary Movement organizers commonly estimate that the number of churches declaring sanctuary doubled as a direct consequence of the Tucson trials (Davidson: 1988; Crittenden: 1988).
The third question in Table 3 addresses the effect of the Tucson trials on the group. Here 73.2% of the participants responded that they perceived the trials as an incentive to participation in general, 7.3% felt they had no effect and 8.5% felt that the trials proven to be a disincentive (mainly to those who were only marginally committed.) Another 7.3% felt that the trials had had mixed effect.
Again, it is clear that even the direct attempt of the U.S. government to impose legal penalties for Sanctuary activity did not prove a disincentive at either the group or the individual level. (Interestingly, even the majority of non- participants responded that the trials had either been an incentive or had no effect on the movement as a whole.)
The results of this study concur with Muller and Opp's findings of costs having mild incentive value for collective protest behavior. (See Equation 1, Table 7). Results of individual questions show strong support for the hypothesis while the tobit results in Equations 1 and 3 show significance at the 80% confidence level. The questions which remain to be explored are the dynamics of the incentive value. It may be that an individual's participation calcu- lations, be they formal or informal, are not significantly affected by abstract perceptions of potential costs, as much as by the method and context in which there is an attempt to impose those costs. In the case of Sanctuary, the government risks creating sympathy for movement members by making them appear as martyrs for a humanitarian cause.
The most interesting finding of the surveys is one which contradicts an underlying principle of Muller and Opp's public goods model. That is the degree to which the indi- vidual participating in collective rebellious behavior is alienated from the regime. "...(A) very basic indicator of the public goods value of rebellious collective action is the extent to which an individual is hostile to or alienated from the political system in general" (p. 477). They argue that an individual who is not alienated from the political system "would be unlikely to consider acts of rebellious behavior as being in the public interest--indeed, such a loyal citizen would be likely to regard change resulting from rebellious behavior as a public "bad." By contrast, a person who is fundamentally alienated from the existing government and political institutions would be likely to regard change resulting from rebellious behavior as a public good." (p. 477)
In the attempt to tap feelings of alienation, three differ- ent measures were employed in this survey. The results of the first two are reported in Tables 4 and 5. First, participants were asked to say whether their protest was directed `mainly,' `in part' or `little/none' against each of the following: U.S. government treatment of Central American refugees; U.S. policy in Central America; the Reagan
 The positive coefficient in Equation 1 refers to the incentive rather than disincentive value of costs. Muller and Opp were surprised to find that the coefficient in their equation was negative, which by their scaling procedure suggested a negation of the negative value of costs: an incentive value. Because of the scaling of the cost variable in this study, a positive coefficient is indicative of the incentive value of costs.
 Several respondents quipped, "Only two things become more valuable when stepped on: Persian rugs and the church." Administration; or the governmental system as a whole. The results indicate that the majority, 68.6%, responded that their protest action was `mainly' against U.S. policy toward Central American refugees, 62.4% said that it was 'mainly' against U.S. policy in Central America. Only 18.1% said that their protest was 'mainly' against the Reagan Administration and even less, 8.2% said that it was 'mainly' against the governmental system as a whole.
The majority of participants said that their protest was 'in part' against the Reagan Administration, yet 'little or none' against the governmental system as a whole. Clearly these individuals are more alienated from the Reagan Administration specifically than they are the institutions of government in this country. From this question a picture emerges of protest against specific policies, less or only in part, against the Reagan Administration, and little against the governmental system as a whole. Thus while these individuals are alienated from a specific policy or set of policies, they do not appear to be "fundamentally" alienated from the political .pa system. Tobit results which fail to show a significant relationship between participation and these measures of alienation confirm this assessment as shown in Table 4a.
This presents an interesting nuance which was not explored in the original Muller and Opp study: the degree to which protesters may be alienated from a specific administration yet not necessarily alienated from the governmental system. In order to flesh out this distinction, the respondents were asked, "How much of the time do you think you can trust the government in Washington to do what is right?" As is shown in Table 5, the respondents were asked to distinguish between the Reagan Administration and any administration.
None of the respondents, either participants or non-participants, answered that they could `always' trust either the Reagan Administration or any other administration. From the table, however, it is clear that participants are inclined to trust 'Any Administration' more than the Reagan Adminis- tration. While a majority of participants chose `some of the time' in both categories, for the 'Reagan Administration' this was a slim majority. Non-participants appear more inclined to trust the Reagan Administration than participants, yet even they are generally skeptical. The distribution for both participants and non-participants' evaluations of 'Any Administration' are similar yet participants do weigh more heavily on the mistrustful, alienated end. These responses do not exactly provide a rousing endorsement from either participants or non-participants in terms of trust in the word of the government. Tobit estimates do show participants in this movement to be more alienated from the system of government as a whole than non- participants, and even more alienated from the Reagan Administration in specific (Table 5a).
 The survey specifically asked about the respondents' feelings toward the Reagan, not Bush, Administration for two reasons: 1) the majority of Sanctuary activity took place in the early to mid-1980s, during President Reagan's term in office; and 2) most of these surveys were conducted while Reagan was still in office. Surveys taken after January 1990 clearly specified that questions still referred to the Reagan Administration.
Third, the respondents were asked the America and apple pie question. They were asked to choose which of two statements concerning our form of government they agree with more:
Seventy-eight percent of the respondents said they were proud of many things about our form of government while 21.7% responded that they couldn't find much to be proud of. Since this question is posed as an evaluation of the general form of government rather than of a specific administration, it indicates further that the majority of these respondents are not alienated from our system of government as a whole. As can be seen from Table 6, the cross tabulation of degree of participation and evaluation of our form of government yields no significant pattern of results. Taken together with the previous questions, a picture emerges of protesters who are clearly anti-policy rather than anti-system, yet who strongly distrust the current administration.
In terms of alienation from the political system, in Muller and Opp's terms "fundamentally alienated from the current government and political institutions" (emphasis added) participants in the Sanctuary Movement do not appear to qualify. Two of the three indicators of alienation from the governmental system yield insignificant results. Yet the composite public goods variable weighted by degree of alien- ation from the regime does come up significant in Equa- tion 1. In contrast to Muller and Opp's original findings, the coefficient is lower than that of the first public goods variable (individual efficacy* value of the public good) but they are close. This result seemed counterintuitive given the lack of evidence for the influence of alienation from the governmental system on protest.
 To make the composit variable responses were used from the question `How much of the time do you feel you can trust the government in Washington to do what is right?' (`Any Administration'). These responses appeared to correspond most closely to Muller and Opp's conjecture and provide the fairest test for their hypothesis.
 To gauge this variable, respondents were asked directly how important Sanctuary was to them in relation to other elements in their lives, such as family, job, other activities.
In order to determine whether the significance found for the public goods as determined by alienation was an anomaly created by the compounding of the individual efficacy varia- ble and the alienation variable, further tests were performed. As can be seen in Table 7 this appears to be the case. Equation 1, is the replication of Muller and Opp's public goods model using data from the Sanctuary Movement. Both piVn and piVs are shown to be significant, although piVs is the weaker of the two where in the original model it was the stronger predictor. The fact that both public goods variables contain within them the individual efficacy variable (pi) leads to the suspicion of multicollinearity, biasing parameter estimates. Running a correlation matrix on piVn and piVs revealed a relationship of .46 between these variables.
Equations 2 and 3 confirm that when run separately, the public goods variable weighted by alienation from the political system (piVs) drops out of significance, while the public goods variable weighted by the importance attached to the public good (piVn) remains a strong predictor of protest action. While these results confirm that the public goods variable is a stronger predic- tor of social protest than soft selective incentives, they contradict the basis upon which Muller and Opp define the public goods variable. While much protest may be catalyzed by opposition to particular policies or even particular administrations, alienation from the governmental system does not prove to be a "basic indicator of the public goods value of rebellious collective action" (Muller and Opp: 1986, p. 477). A more appropriate indicator is the importance attached by the individual to achieving the public good, or analogously, discontent with a particular government policy. It is at this point that an argument put forth by Barnes and Kasse becomes particularly useful. Barnes and Kasse argue that there is an "increased inclination of the citizenry to participate in [demonstrations and other acts previously seen as illegitimate.]...as a legitimate resource of demo- cratic citizenship..." (1979, p. 31).
This appears to be the case with members of Sanctuary. Table 8 shows the results of the survey question designed to test this assertion. A series of actions were listed which could be used to protest U.S policy toward Central American refu- gees. The respondents were asked to indicate which of these actions they felt were acceptable forms of protest.
 Multicollinearity results when two independent variables in an equation are linearly related. This condition has the effect of biasing parameter estimates, leading to unwarranted inferences from the data. See Kennedy (1985, p. 42).
From these results it is clear that Sanctuary is a non-violent movement, the majority of whose members find violent forms of protest to be unacceptable. When asked their feelings concerning protest action which could lead to a prison sentence or to physical danger, the vast majority (98.8% and 93% respectively) felt these to be either some- times acceptable or always acceptable. In other words, these individuals did not find such unconventional forms of protest to be unacceptable. In response to the actions most commonly associated with the Sanctuary Movement, transporting and harboring refugees as well as helping refugees cross the border, the clear majority found these to be always acceptable and the remainder called them sometimes acceptable. A few responded that helping refugees cross the border was an unacceptable form of protest action.
In support of Barnes and Kasse's argument, these results show that the actions which participants in Sanctuary perceive as legitimate are not limited to those deemed legitimate by the state. They view their particular form of unconventional protest as among the legitimate repertoire of actions available to members of a democratic citizenry.
One point which is crucial to understanding the members of the Sanctuary Movement is their view of exactly who is breaking the law. One of the basic premises of the Sanctu- ary Movement is that the U.S. government is illegally and immorally disregarding international and domestic law concerning refugees (Geneva Convention of 1949, United Nations High Commission on Refugees Statement of 1981, and the Congressional Refugee Act of 1980). Sanctuary workers maintain, therefore, that it is not they who are breaking the law, rather they are acting on their legal and moral obligation to uphold the law. Again, this does not appear to be the philosophy of a group of individuals who are fundamentally alienated from the political institutions of the country.
Evaluating the responses of non-participants does not reveal the same perception of acceptability toward protest action. (Table 9). Interestingly, these non-participants appear more sympathetic to the protest actions specific to the Sanctuary Movement than might be expected. Only 20% found that harboring, transporting and helping refugees cross the border was unacceptable activity. This may well reflect the fact that although these individuals are non-participants, they have remained members of Sanctuary congregations and thus are more accepting of the movement than a more detached control group might be. Still, they differ from participants in their views toward protest which could lead to prison sen- tences or physical danger. None of the non-participants found these actions to be acceptable. The majority (80% and 70% respectively) said that such actions were `sometimes acceptable', as was the case with participants, while the remainder felt that they were 'unacceptable.' One can only speculate as to why a higher percentage did not perceive such clearly unconventional (and in the government's eyes illegitimate) forms of action as unacceptable. The results do appear to support Barnes and Kasse's hypothesis, in that even if non-participants are unwilling to undertake such actions themselves, they perceive them as at least sometimes legitimate forms of action.
One final point resulting from the pilot survey provides a further dimension of interest as regards individuals' per- ceptions of legitimate actions available to a democratic citizenry. The respondents were asked what other types of actions they had considered taking to address the issue of Central American refugees. The options included convention- al action such as letters to representatives or letters to newspapers, as well as some less conventional actions such as protests and demonstrations or vigils. A blank space was also provided for respondents to volunteer alternative possibilities. They were then asked whether they felt that these actions were effective, somewhat effective, ineffective or they didn't try it. The results of this question are reported in Table 10.
Ninety-nine percent of participants in the Sanctuary Move- ment indicated that they had indeed tried at least one alternative course of action, 82% had tried three or more. While these alternatives were mostly considered only `somewhat effective' they were less often listed as `ineffective'. This suggests that turning to the illegal activities of transporting and harboring refugees was not simply an act of desperation taken after all else had failed. Rather, it seems to support the Barnes and Kasse hypothesis that the repertoire of what is considered legitimate action has expanded, in this case to include action considered illegal by the current administration.
The Muller and Opp public goods model of rebellious collec- tive behavior is an extremely important contribution to and modification of rational choice models of collective action. Clearly there is more to many acts of protest than self-interest and economically objective calculations of individual efficacy.
The most important modification which needs to be made, however, is the conception of protestors as necessarily alienated from the governmental system. As Barnes and Kasse have recognized, perceptions of legitimate action are changing, and so too is the profile of the activist. The respondents to this survey fit a clear pattern of white, upper-middle or middle class individuals predominantly between forty and sixty. While many have been involved in other social action groups (anti-nuclear, urban poor, homeless, civil rights) not all have been. They do not fit the ster- eotype of "sixties activists looking for a new cause." For many, this is the first social action protest groups they have been involved with.
It is also important to make clear that this model is not a test of whether participants are acting for "selfish" or "unselfish" reasons. As one respondent remarked, "everything has a selfish reason at the bottom," referring to the fact that the fulfillment of his moral and religious commitments make him feel satisfied and good. "But those same reasons call you to ameliorate the suffering of others." Russell Hardin comments in Collective Action "...there is presumably no doubt in most circles that people are often public-spirited," (p. 103)..."(T)he interesting issue for moral motivations is not whether they underlie any actions--surely they do--but whether they underlie any large-scale actions where narrow self-interest could not motivate cooperation." (p. 104.)
In the case of Sanctuary, the participants do not benefit materially by achieving either the short run goals of the movement (helping refugees in distress) or in the long run (changing policy toward refugees, and perhaps U.S. policy toward Central America). Thus, if narrow self-interest is defined as material reward, this group does not qualify. Even if narrow self-interest includes such "soft selective incentives" as personal gratification, or "psychic income," the results of the survey taken .pa here support Muller and Opp's findings that "the public goods variables, not selective incentives, appear to be the most important incentives for performing rebellious political behavior." (p. 483.)
This study also adds to the growing literature which finds costs to have a contrary effect to that expected in tradi- tional rational choice models. When do costs serve as a deterrent and when do they have the unintended consequence of providing incentives to protest? Hirschman (1982) argues that an impetus to joining collective action in which one believes stems from the fact that to do otherwise would ultimately lead to a reduction in that individual's own utility. In fact, Muller and Opp's perplexity over the mild incentive effects of what would otherwise be considered costs are addressed by Hirschman as misspecified. "...(T)he neat distinction between costs and benefits of action in the public interest vanishes, since striving, which should be entered on the cost side, turns out to be part of the benefits." Chong (1987) characterizes this sentiment as "fighting the good fight" which " is ennobling and uplifting and it carries its own rewards" (p. 6). Thus perhaps the time, energy and material resources donated to realizing the public good serve not as costs but as benefits in themselves, since not only the achievement of the goal, but the process itself is seen by participants as a public good.
The results of the national survey of the Sanctuary Movement serve to test some general understandings of the nature of collective protest as well as more specific hypotheses put forward in recent models. They show the importance of incorporating the public goods variable in a rational choice model of collective action, as Muller and Opp have done, and add support to other evidence which questions the nature of potential costs in the calculations of protesters, particularly when that protest is of a social or universalistic nature.
Most importantly, however, they contradict one of the underlying tenets of the public goods model, (and a common stereotype of activists) that the participants in rebellious collective action are necessarily alienated from the politi- cal system. It is important to distinguish between alienation from a given administration or set of policies and alienation from the political institutions of a country. In the case of the Sanctuary Movement it may even be argued that participants view their protest against the Reagan Administration as simultaneously in defense of their understanding of this country's basic political institutions which they feel are being violated. This perception lends credence to their view of unconventional forms of protest as legitimate tools of expression and participation available to citizens of a democracy.
The Sanctuary Movement, as well as Muller and Opp's case study of anti-nuclear protesters, is indicative of a partic- ular type of rebellious collective action. It demonstrates a case of relatively well-off, well educated individuals in a well developed society and economy who engage in comparatively mild forms of protest in order to effect change within a political system. The results of this study are not intended to be generalized across all types of rebellion and protest. It would be impractical, if not utterly ridiculous, to argue that armed insurgents in the Third World are not alienated from the regimes against which they are rebelling. Surely they are. So too would be the case for separatist movements.
This study does not argue that "free riding" is not a problem for most political action, or that material incentives are irrelevant in mobilizing support for many political groups. What it does do, rather, is to evaluate a case which is representative of a growing, perhaps resurging, type of political participation: direct action protest over moral and social issues facing a developed society. Much debate has raged over the degree to which political participation has changed based upon the post-materialist value structures found in advanced industrial democracies (Ingle- hart, Flanagan: 1987). Other case studies such as the pro- choice vs. anti-abortion, environmentalist, pro- and anti- school prayer movements, advocates of the homeless, and other social action groups illustrate the numerous cases in which individuals are motivated to join in collective action. Surely there are others who support the cause yet choose not to participate, probably more in number than those who do get involved.
The questions which this study and others like it, including Muller and Opp's, are primarily concerned with is what distinctive characteristics can be isolated to better understand the motivations of activists in universalistic and public goods movements in the context of industrialized democracies. The fact that the public goods variable emerges as the primary motivator in this model is undoubtedly the product of the specific nature of the protest movement. The same result would be surprising if analyzing urban riots in the 1960s or revolutionary movements.
Still, it is important to recognize differences in types of protest movements and develop models which most accurately reflect the motivational forces specific to them. While some of these may overlap, others will prove unique. Understanding the growing number of social protest movements of a public goods nature in the United States and other industrialized democracies is key to deciphering changes in political participation in those societies.
If, as the results of this study demonstrate, participants in direct action social protest movements cannot be assumed necessarily to be alienated from the political system in general, then the door is open for more constructive inter- pretations of the meaning of political participation within a democratic system. As more traditional institutions of participation, such as political parties and elections, lose their potency for large numbers of citizens, new channels take their place. While protest has historically been recognized as an alternative for the poor and disaffected (Piven and Cloward: 1979; Lipsky: 1968, Rude: 1964; Thomp- son: 1971) it has also been considered an illegitimate form of action.
Participants in the Sanctuary Movement have not chosen protest becasue they were excluded from all other means of access to the governmental system. Rather, they chose protest in addition to other, more conventional means of participation as a legitimate means to voice their discontent with, and effect change in, a particular policy. What is distinctive about these individuals is not the fact that they are alienated from the governmental system, but the degree to which they value the public good for which they are working and their view of what constitutes legitimate forms of action in order to realize that goal. Understanding such movements will provide considerable insight into the motivations catalyzing a significant amount of political participation in contemporary democratic society.
Barnes, Samuel and Max Kasse et al. 1979. Political Action: Mass Participation in Five Western Democracies. (London: Sage Publications).
Chong, Dennis. 1987. "Public Spirited Collective Action." Paper prepared for the APSA, Chicago, September.
Davidson, Miriam. 1988. Convictions of the Heart: Jim Corbett and the Sanctuary Movement. (Tucson: University of Arizona Press).
Davies, James. 1962. "Toward A Theory of Revolution," American Sociological Review, XXVII(February) pp. 5-19.
_____________. 1969. "The J-Curve of Rising and Declining Satisfactions as a Cause of Some Great Revolutions and Contained Rebellion," in H.D. Graham and T.R. Gurr, eds., Violence in America: Historical and Comparative Perspectives. (Washington, D.C.: National Commission on the Causes and Prevention of Violence).
De Nardo, James. 1985. Power in Numbers: The Political Strategy of Protest and Rebellion. (Princeton, N.J.: Princeton University Press).
Downs, Anthony. 1957. An Economic Theory of Democracy. (New York: Harper).
Feierabend, Ivo and F.R.L. Feierabend and Betty Nesvold. 1969. "Social Change and Political Violence: Cross-National Patterns," in Graham and Gurr, eds., Violence in America. (New York: Praeger) 632-687.
Finkel, S., E. Muller & K. Deiter-Opp. 1989. "Personal Influence, Collective Rationality, and Mass Political Action." APSR Vol. 83, No. 3 (September) pp. 885-903.
Finkel, Steven and James Rule. 1986. "Relative Deprivation Theories and Related Psychological Theories of Civil Violence: A Critical Review," in K. Lang, G. Lang and L. Kreisberg, eds. Research in Social Movements, Conflict and Change. Vol. 9 (Greenwich, Conn: JAI Press, Inc.).
Fireman, Bruce and William Gamson. 1979. "Utilitarian Logic in the Resource Mobilization Perspective." in Zald and McCarthy eds. The Dynamics of Social Movements. (Cambridge, Mass: Winthrop Publishers).
Freeman, Jo. 1979. "Resource Mobilization and Strategy: A Model For Analyzing Social Movement Organization Actions," in Zald and McCarthy, eds., The Dynamics of Social Movements. (Cambridge: Winthrop Press) 167-189.
Gamson, William. 1975. Strategy of Social Protest. (Homewood, Ill.: Dorsey).
Gurr, Ted Robert. 1970. Why Men Rebel. (Princeton, N.J.: Princeton University Press).
Hardin, Russell. 1982. Collective Action. (Baltimore: The Johns Hopkins University Press).
Hirschman, Albert O. 1982. Shifting Involvements. (Princeton: Princeton University Press).
Hoffer, Eric. 1951. The True Believer. (New York: Harper Press).
Inglehart, Ronald and Scott Flanagan. 1987. "Controversy: Value Change in Industrial Societies." APSR, Vol. 81 No. 4 (December) 1289-1319.
Jenkins, J. Craig. 1983. "Resource Mobilization Theory and the Study of Social Movements." Annual Review of Sociology, 9:527-53.
Kennedy, Peter. 1985. A Guide To Econometrics. (Cambridge, Mass.: The MIT Press).
Klandermans, Bert. 1984. "Mobilization and Participation: Social-Psychological Expansions of Resource Mobilization Theory." American Sociological Review, Vol. 49 (October) 583-600.
Kornhauser, William. 1959. The Politics of Mass Society. (Glencoe, Ill.: Free Press).
LeBon, Gustave. 1903 (1968). "The Mind Of The Crowds," in L. Genevie, ed., Collective Behavior and Social Movements. (Itasca, Ill.: Peacock Press).
Leahy, Peter and Allen Mazur. 1978. "A Comparison of Movements Opposed to Nuclear Power, Flouridation and Abortion," in L. Kreisberg, ed. Research in Social Movements, Conflicts and Change. Vol. 1 (Greenwhich, Conn.: JAI Press, Inc.) 143-154.
Lipsky, Michael. 1968. "Protest as a Political Resource." APSR, Vol. 62 pp. 1144-1158.
Marsh, Alan. 1977. Protest and Political Consciousness.
(Beverly Hills: Sage).
McAdam, Doug. 1982. Political Process and the Development of Black Insurgency, 1930-1970. (Chicago: The University of Chicago Press).
Morris, Aldon. 1981. "Black Southern Student Sit-In Movements: An Analysis of Internal Organization." American Sociological Review, 46: 744-67.
Muller, Edward N. and T.O. Jukan. 1983. "Discontent and Aggressive Political Behavior." British Journal of Political Science, 13:159-179.
Muller, Edward and Karl-Dieter Opp. 1986. "Rational Choice and Rebellious Collective Action." APSR, Vol. 80. No.2. June: 471-487.
Oberschall, Anthony. 1973. Social Conflict and Social Movements. (Englewood Cliffs, N.J.: Prentice-Hall).
Olson, Mancur. 1965.1971. The Logic of Collective Action: Public Goods and the Theory of Groups. (Cambridge,
Mass: Harvard University Press).
Piven, F.F. and Richard Cloward. 1979. Poor People's Move- ments. (New York: Vintage Books).
Rude, Eric. 1964. The Crowd In History: A Study of Popular Disturbances in France and England 1730-1848. (New
Tilly, Charles. 1978. From Mobilization to Revolution. (Reading, Mass: Addison-Wesley).
Thompson, E.P. 1971. "The Moral Economy of the English Crowd in the Eighteenth Century." Past and Present, Vol. 50 (February) 76-136.
Tolbert, Pamela. 1981. "Determinants of Social Movement Participation." Paper Presented at Annual American Sociological Association Meetings, Toronto.
Tullock, Gordon. 1971. "The Paradox of Revolution." Public Choice. 11:89-99.
Zald, Mayer and John McCarthy eds. 1979. The Dynamics of Social Movements. (Cambridge, Mass: Winthrop).
Copyright (C) May 1990 by Martha Liebler Gibson
All rights reserved.
Single copies of this paper may be reproduced for personal use with the following conditions:
- All information concerning copyrights, authorship, acknowledgement of grant support, and publication must not be deleted from printed or electronic copies.
- Any use of this material must be fully cited and in compliance with all copyright statutes and ethical fair use principles.
- The paper may be reproduced only in its entirety.
- This paper may not be reposted on any other electronic bulletin board or retrieval system without formal permission from the Consortium or the author.
- This paper is provided free of charge and may not be offered for sale by anyone other than the Consortium or the author(s).
Graphic images are not included in this file. For information on how to obtain graphics contact the CRC at the address below.
All correspondence related to this paper should be addressed to:
Conflict Resolution Consortium
Campus Box 327
University of Colorado
Boulder, CO 80309