CJPSR3

Wiley and Law and Society Association are collaborating with JSTOR to digitize, preserve and extend access to Law & Society Review.

http://www.jstor.org

Understandings of Justice: Institutional Legitimacy, Procedural Justice, and Political Tolerance Author(s): James L. Gibson Source: Law & Society Review, Vol. 23, No. 3 (1989), pp. 469-496 Published by: on behalf of the Wiley Law and Society Association Stable URL: http://www.jstor.org/stable/3053830 Accessed: 21-02-2016 15:38 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

UNDERSTANDINGS OF JUSTICE: INSTlITUTIONAL LEGITIMACY, PROCEDURAL JUSTICE, AND

POLITICAL TOLERANCE

JAMES L. GIBSON

This research examines the linkages among institutional legiti- macy, perceptions of procedural justice, and voluntary compliance with unpopular institutional decisions within the context of political intolerance and repression. Several questions are addressed, includ- ing: To what degree do judicial decisions contribute to the acceptance of unpopular political decisions? Do court decisions have a greater power to legitimize than the decisions of other political institutions? Are courts perceived as more procedurally fair than other political institutions? Do perceptions of procedural fairness-be it in a court or legislative institution-contribute to the efficacy of institutional decisions? The basic hypothesis of this research is that to the extent that an institution employs fair decisionmaking procedures, it is viewed as legitimate and citizens are more likely to comply with its decisions, even when they are unpopular. Based on an analysis of na- tional survey data, I conclude that, although perceptions of institu- tional procedure have little impact on compliance, institutional legiti- macy does seem to have some effect. The United States Supreme Court in particular seems to have some ability to elicit acceptance of public policies that are unpopular with the mass public. This effect is greatest among opinion leaders. I conclude with some observations about how these findings fit with the growing literature on proce- dural justice and with some thoughts about the implications of the findings for the protection of democratic liberty.

This article is a revised version of a paper delivered at the 1988 Annual Meeting of the Law and Society Association. It is based on research funded by the National Science Foundation (SES Grant No. 86-06642). I am deeply in- debted to Felice Levine at the Foundation for support for the project. A number of colleagues have contributed significantly to the development of the research, including Jonathan Casper, James Davis, Jennifer Hochschild, Stan- ley Presser, Lee Sigelman, Paul M. Sniderman, John L. Sullivan, and Tom Tyler. James P. Wenzel and Peter J. Van Koppen have made useful com- ments on draft versions of this paper. I am also indebted to the National Opin- ion Research Center - and especially to Dick Rubin - for their excellent ex- ecution of the survey. James P. Wenzel, Bernadette McKinney, Steven Shamberger, and Marilyn Yale provided quite helpful research assistance. This paper also makes use of General Social Survey data, made available to me through the Inter-University Consortium for Political and Social Research. Kathleen McGraw, Walter Murphy, and Tom Tyler provided most helpful comments on an earlier version of this article. I am also thankful for the ex- tremely thorough and thoughtful reviews provided by the referees.

LAW & SOCIETY REVIEW, Volume 23, Number 3 (1989)

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

470 UNDERSTANDINGS OF JUSTICE

I. INTRODUCTION

One enduring component of the scholarly folklore on the United States Supreme Court is that the Court has some sort of special ability to legitimize government policies and actions. Be- cause the Court is revered by the citizenry as the foremost guard- ian of the Constitution, its decisions are given uncommon sway. This special ability to legitimize is crucial to the political system because legitimacy engenders voluntary compliance with law by citizens (cf. Easton, 1965).

Dahl (1957) was one of the first social scientists to claim this power for the Court. After being forced by empirical evidence to conclude that the Supreme Court has rarely protected less power- ful political minorities from the abuses of the majority, Dahl as- serted that the institution's function throughout American history has been to legitimize public policy, not to upset it.' Not only do the decisions of the Supreme Court legitimize policy, but they also do so more effectively than the decisions of any other institution.2 Dahl's view quickly became the conventional wisdom among judi- cial scholars (see, e.g., Black, 1960; Bickel, 1962).

This question of legitimizing capacity is crucial because of its implications for voluntary compliance with law. Since legitimate policies, even unpopular ones, are more likely to evoke compli- ance, the legitimizing abilities of the Supreme Court perform an extremely important function for the American polity. Unpopular decisions by institutions considered to be legitimate are more likely to be accepted as authoritative than are those made by polit- ical institutions of lesser legitimacy (cf. Rasinski et al., 1985). Since coercion is rarely effective in modern, mass societies-and indeed it would be extremely costly if political systems were re- quired to enforce each of their policies coercively-this ability to legitimize is often considered essential to system stability and per- sistence.

As important as this problem may be, empirical research has not generated a great deal of support for the legitimation hypothe- sis.3 For instance, Murphy and Tanenhaus (1968, 1969) conclude that only a fairly small proportion of the population meets even the minimal criteria necessary for the Court to legitimize public policies. Jaros and Roper (1980), analyzing the responses of col- lege students, were able to observe compliant behavior but were

1 Dahl's thesis is controversial. For contrary points of view, see Funston (1975) and Casper (1976).

2 Adamany (1973: 807) summarizes this position succinctly in defining le- gitimacy as "an evaluative perception by the people that Supreme Court man- dates should be accepted because the justices, as guardians of the Constitution, act by legal right, because they exercise a traditional authority, and because they constitute an appropriate societal institution."

3 More generally, there is actually little systematic empirical research on the concept legitimacy. See, for example, McEwen and Maiman, 1986; and Rasinski et al., 1985. The most important exception is Tyler (forthcoming).

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 471

remarkably unsuccessful in accounting for it. Baas and Thomas (1984), also analyzing college students, found little ability of the Supreme Court to legitimize policies. Others have become more directly skeptical about the power of the Court to legitimize the policies of other institutions. Writing in 1973, Adamany (p. 807) concluded that "none who bottom their arguments on the Court's legitimacy-conferring capacity offer the slightest empirical basis for its reality. Perhaps the doubter should be content to move for a directed verdict, the proponents having failed utterly to adduce even a scintilla of evidence, much less a preponderance, in support of their plea."4 Thus research on the United States Supreme Court does not provide much confidence in the legitimation hy- pothesis.

On the other hand, new evidence suggests that courts in gen- eral may have special powers to legitimize. The growing literature on procedural justice emphasizes that decisions that are perceived as having been fairly made-even if they are unpopular-engender considerably more support than they otherwise would (see, e.g., Tyler, 1988, forthcoming; Lind and Tyler, 1988). As Tyler, Rasin- ski, and Griffin (1986; 976) note, "Citizens' compliance with the law and acceptance of new government policies is [sic] not based solely on reward and punishment considerations, but also on judg- ments about the fairness of rules and policies and of the process by which these policies were arrived at."5 Proper process contributes to acceptance of unpopular products. Perhaps the special abilities of the Supreme Court to legitimize public policy are associated with perceptions of its decisionmaking processes. Perhaps the edge that courts in general are thought to have in generating com- pliance is associated with their special concern with procedural fairness.

Indeed, it would not be surprising to find that courts are per- ceived as the most procedurally fair of all political institutions, since many people view them as insulated from "politics." That is, courts are thought to be non-political in the sense that they make decisions on the basis of merit and logic, shunning the favoritism, logrolling, and pork barrel so central to decisionmaking in other political institutions (see, e.g., Jaros and Mendelsohn, 1967; Eng- strom and Giles, 1972; Jaros and Roper, 1980). In courts, it is not who one knows that controls the decision, but rather it is the mer- its of one's case. As the literature on procedural justice would hy-

4 Commenting a decade later, Adamany and Grossman (1983: 430) as- serted: "It is no longer tenable to continue to continue to assert that the fail- ure of attempts to curb the Court or overturn its policies lies in some wide- spread public or elite support either for the justices' decisions or for the Court as an institution."

5 Max Weber (1947) termed this "formal legitimation." For a recent em- pirical analysis of this concept, see McEwen and Maiman, 1986.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

472 UNDERSTANDINGS OF JUSTICE

pothesize, to the extent that courts are perceived as fair, their deci- sions are more likely to be accepted as authoritative and final.

Consequently, there is reason to re-open the question of whether the courts have special powers of legitimation. The gen- eral hypothesis is that to the extent that an institution employs fair decisionmaking procedures, it is perceived as legitimate and citizens are more likely to comply with its decisions, even when they are unpopular.6 I test this hypothesis by considering several questions. To what degree do judicial decisions contribute to the acceptance of unpopular political decisions? Do court decisions have greater powers to legitimize than the decisions of other polit- ical institutions? Are courts perceived as more procedurally fair than other political institutions? Do perceptions of procedural fairness in either a judicial or legislative institution contribute to the efficacy of institutional decisions? To what degree does institu- tional legitimacy contribute to compliance?

II. JUSTICE, LEGITIMACY, AND POLITICAL INTOLERANCE

I shall consider the problems of justice, legitimacy, and com- pliance within the context of political tolerance and civil liberties. In particular, my concern is with the role that political institutions can play in encouraging citizens who are predisposed not to toler- ate unpopular political minorities not to act on their intolerance. The basic hypotheses under consideration are: (1) Institutional de- cisions can block the behavioral consequences of attitudinal intol- erance by getting citizens to accept non-repressive policy decisions7 as authoritative and final; (2) judicial institutions are more effec- tive at this than are legislative institutions; (3) institutions per- ceived as making decisions fairly are more effective at generating compliance than are institutions not perceived as fair; and (4) more legitimate institutions are more effective at bringing about compliance than are less legitimate institutions. Because the procedural justice literature rarely considers political problems of this nature (but see Tyler, et al., 1985; Rasinski, et al., 1985; and Tyler, forthcoming), it is perhaps useful to begin the inquiry with

6 Of course no one would argue that procedural perceptions are the sole determinant of compliance. The procedural hypothesis is instead that percep- tions of fair methods of decisionmaking have an independent impact on will- ingness to accept an institutional decision as final and binding. For a more complete consideration of the compliance problem, see Tyler (forthcoming).

7 This analysis is mainly concerned with the ability of institutional legiti- macy to neutralize citizen intolerance. Of course, courts that ratify repressive public policy may neutralize tolerant public opinion. My focus is on institu- tional decisions that are democratic but that face an intolerant mass public. This is justified because intolerance is so widespread among the mass public in the United States (see, e.g., Gibson, in press) and because there is some evi- dence that elites, including those who make decisions within political institu- tions, are more tolerant than the ordinary public (see, e.g., Stouffer, 1955; Mc- Closky and Brill, 1983). I do not gainsay that the processes, however, can work in both democratic and anti-democratic directions.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 473

a digression to an actual civil liberties dispute. I use this conflict to

give some idea of how institutional legitimacy and procedural fair- ness may influence acceptance of unpopular policy decisions. The event on which I shall focus concerns the right of the American Nazi Party to hold an anti-semitic demonstration in Skokie, Illi- nois, a predominantly Jewish suburb of Chicago.

The dispute in Skokie revolved around the efforts of Frank Collin and his National Socialist Party of America to hold a dem- onstration in 1977.8 When confronted with the request to demon- strate, the Village of Skokie promptly adopted three pieces of leg- islation designed to control assemblies. Among the requirements of the first ordinance were $350,000 liability and property damage insurance, and a prohibition on demonstrations that would libel

specific religious, racial, ethnic, national, or regional groups. The second ordinance forbade the dissemination of materials that would libel groups, and outlawed markings and clothing of sym- bolic significance. The last ordinance prohibited demonstrations in

military-style uniforms. All three ordinances were deemed decid-

edly unconstitutional by the federal trial and appellate courts (Col- lin v. Smith, 447 F. Supp. 676 (N.D. Ill. 1978); and Collin v. Smith, 578 F.2d 1197 (7th Cir. 1978)).

For the purposes of this article, the importance of the Skokie

dispute is not its implications for constitutional law but instead its lessons on compliance, tolerance, and repression. The citizens of the village strongly and actively opposed the demonstration, even to the point that some threatened violence if the Nazis were actu-

ally to demonstrate. Even the community leaders were united in

opposing the demonstration. Thus, there was a widespread intoler- ant and anti-democratic consensus among the elites and masses in Skokie.

At the same time, however, the dispute was brought to a rela-

tively democratic conclusion.9 The ordinances were invalidated, and Collin was allowed to demonstrate (for various reasons, he chose to demonstrate in Chicago instead). Thus, despite over-

whelming intolerance, the dispute was resolved without undue

compromise to basic civil liberties. Barnum (1982) has argued that four factors were particularly

important in the Skokie-Nazi dispute. First, the political leaders of Skokie acted undemocratically largely because they were forced to do so by their constituents. Their own inclinations were much more tolerant. Second, in this controversy (as in all political con- troversies: see Gibson and Bingham, 1985), there was a substantial

gap between attitudinal intolerance and behavioral intolerance.

8 For accounts of this conflict, see Gibson and Bingham, 1985; Barnum, 1982; and Hamlin, 1980.

9 My understanding of democracy very closely parallels that of Dahl (1971).

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

474 UNDERSTANDINGS OF JUSTICE

Many of those who opposed the demonstration simply did not act on their beliefs. Third, the policymaking process itself tended to defuse the emotional and symbolic impact of the issue. By delay- ing a decision and by defining it as a technical, legal dispute, the process of adjudication in the courts helped to decrease the sali- ence and emotional impact of the issue. Finally, there was residual support for democratic procedural values among the elites and masses in Skokie. Although many may have opposed the demon- stration, there was overwhelming support for allowing the dispute to be resolved through democratic procedures like adjudication. Barnum (1982: 504) argues that "the existence of judicial mecha- nisms for resolving civil liberties disputes must be accompanied by something like consensual support among the political elites for the propriety of resorting to the courts and the necessity of ac- cepting, without question, the policy determinations of the courts" (emphasis added). It is thus the legitimacy of the judiciary that contributed to acceptance of a policy decision that was abhorred by the overwhelming majority of the community.10 Consequently, there is some evidence that public understandings of institutional legitimacy and procedural justice tend to block or ameliorate the political impact of political intolerance. Commitments to demo- cratic procedures neutralize the politically damaging and anti-dem- ocratic impact of political intolerance. This hypothesis will be sub- jected to empirical testing.

III. DATA

My analysis is based on a national survey that was an exten- sion of the 1987 General Social Survey (GSS), a nearly annual sur- vey conducted by the National Opinion Research Center (NORC) with funding from the National Science Foundation. (For details on the sample see the Appendix). Respondents were randomly as- signed one of the two versions of the questionnaire. All respon- dents were asked a set of legislative questions (e.g., regarding a city council), while half the respondents answered a set of judicial questions on the "U.S. Supreme Court" and the remaining half an- swered the same set of questions concerning a "local judge." This reduced the number of cases available for analysis of the judicial questions but otherwise had no effect on the analysis due to the random assignment of the questionnaire versions.11

10 A better test of the hypothesis would have been provided had the Na- zis actually demonstrated in Skokie. Had they done so, some segment of the population would have tried to stop the demonstration through any means possible, including violence. Because the Nazis did not demonstrate, it is im- possible to estimate the proportion of the population that would not have al- lowed the demonstration to take place.

11 Copies of the relevant items are available from the author on request.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 475

IV. POLITICAL INTOLERANCE AND INSTITUTIONAL INTERVENTION

A. Attitudinal Intolerance

The first question to be addressed is how much political intol- erance in fact exists within the mass public. Although this seems likely a relatively simple research problem, measurement issues require a brief digression.

Political intolerance is "a willingness to restrict a disliked group's democratic rights based on the content of one's views" (Sullivan et al., 1985: 225). Sullivan, Piereson, and Marcus (1982) have argued that the concept makes little sense unless the target of intolerance is disliked. One cannot tolerate that with which one agrees: antipathy is a necessary precondition for tolerance. Conse- quently, researchers must insure that the target groups used in tol- erance surveys are disliked by each of the respondents. One way of insuring this is to use the "least-liked" measurement approach.

Before asking respondents whether groups should be toler- ated, the "least-liked" approach asks them to identify the politi- cally active groups that they dislike the most. This insures that all respondents are questioned about a greatly disliked political mi- nority, even if the specific group varies across individuals. Follow- ing the identification of this target, the respondent is asked whether members of the group should be allowed to engage in cer- tain types of political activity, such as giving speeches and running for public office. This approach is thought to generate a more valid measure of political tolerance.12

The subjects in the 1987 survey were first asked to identify and rank order the four groups in politics that they disliked the most. They were then asked the following question about the group they named as most disliked:13

Now I'm going to ask you about a situation involving the [MOST DISLIKED GROUP]. Suppose that the [NAME OF GROUP] wanted to hold a demonstration here in your neighborhood to advocate its political views. Would you strongly support, support, oppose, or strongly oppose a de- cision by the government to allow the demonstration to take place?

The responses reveal a considerable amount of intolerance in the United States: 45 percent strongly oppose allowing the demonstra-

12 A whole series of conceptual and operational issues surround this mea- surement approach (see, e.g., Gibson, 1986; see also Mueller, 1988; Sullivan and Marcus, 1988), but they are of little concern here. I have only used the tech- nique to insure that all respondents are asked to tolerate a threatening activity by a greatly disliked group. The approach is not used to generate a general measure of political tolerance.

13 The most disliked group was the Ku Klux Klan (named by 32%), fol- lowed by communists (24%), advocates of military government (11%), atheists (11%), and Nazis (10%).

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

476 UNDERSTANDINGS OF JUSTICE

tion; an additional 33 percent oppose it; while only 18 percent sup- port allowing it (the remainder are uncertain).14

The analysis that follows is concerned with the degree to which the intolerant attitudes of citizens are neutralized by the tolerant decisions of legitimate institutions. Consequently, I have little concern for the 18 percent of the sample that is initially tol- erant (or with the 4 percent that is uncertain), but rather focus ex- clusively on those who would not support a decision allowing the demonstration.

B. Behavioral Intolerance

The above item measures attitudinal intolerance. In the short term at least, few would argue that legitimate and fair political in- stitutions would affect the attitudes of citizens. The central hy- pothesis of this research is that the legitimacy of non-repressive political institutions neutralizes the behavioral implications of the intolerant attitudes of the mass public by getting citizens to accept the authoritative decisions of the institution, abide by them, and not to challenge them further. Thus, the first question that should be addressed concerns the likelihood of a behavioral response flowing from these attitudes. I consider this problem by asking the subjects what they are inclined to do in response to the proposed demonstration.15

It is useful to have a base-line measure of behavioral propensi- 14 I am not using these data to make any assertion about the absolute

level of tolerance and intolerance in the United States. Political intolerance is a continuum, and no single item can tell us how much exists. Indeed, the mea- sure I have used is a fairly difficult one in that it asks about the most disliked group engaging in a highly threatening activity with a great deal of perceived potential for violence, and it places the demonstration directly in the respon- dent's neighborhood. The severity of this stimulus as compared to the more typical "should a communist be allowed to speak in your community" is con- firmed in part by the considerably greater proportion of intolerant responses it generates. For a more complete discussion of contemporary intolerance in the United States, see Gibson (in press).

15 There is little doubt that a considerable chasm exists between simu- lated or hypothetical behavior and actual action within a real civil liberties dis- pute. Nonetheless, there are important reasons for studying hypothetical be- havior. First, hypothetical behavior may be thought of as a behavioral propensity. Reactions to any given dispute may diverge from the behavioral propensity due to contextual factors, but the propensity nonetheless repre- sents the central tendency (or most likely response, on average) of a larger distribution of actual behaviors. Second, it is exceedingly difficult to study ac- tual repressive political behavior because opportunities to repress one's fellow citizens do not emerge with the periodicity of elections, court appearances, and other chances for political activism. Thus, researchers can rarely mobilize a field study of these sporadic events. (For examples of studies that focus on particular civil liberties disputes, see Gibson and Bingham, 1985; Gibson, 1987; and Gibson and Tedin, 1988.) Although hypothetical behavior is not the same as actual behavior, it is useful to try to take a step beyond the simple focus on attitudes to determine if respondents themselves foresee behavioral conse- quences of their beliefs. For an excellent study of hypothetical behavior in a related area, see Muller (1979). On the use of hypothetical, or "role-playing," measures more generally, see Cooper (1976), and Forward, et al. (1976).

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 477

ties before considering the impact of institutional decisions. Table 1 reports data on the responses to several behavioral possibilities occasioned by the supposition that the demonstration was sched- uled to take place next week.

The first thing to note in the table is that roughly one-half of the subjects report that they would probably do something to try to get the government to stop the demonstration (47.5% is either somewhat or very unlikely to do nothing). Moreover, a relatively small but not inconsequential minority say they would go beyond conventional political activity to take direct actions against the group, including actions that might be illegal. It thus appears that levels of behavioral intolerance would be relatively high prior to any decision by an authoritative political institution.

C. The Effect of Institutional Intervention

The question of greatest interest is whether institutional inter- vention has any impact on willingness to allow the unpopular political minority to exercise its civil liberties. The subjects were asked to suppose that a decision to allow the demonstration had been made by the local legislature (the city council or county gov- ernment), and then by a court (for one-half the sample, the United States Supreme Court; for the other half, a local judge). The legit- imacy hypothesis predicts that the rulings of these institutions should increase the likelihood of accepting the decision as final and doing nothing further to challenge it. Table 2 reports data rel- evant to this question.16

Perhaps the most striking aspect of Table 2 is the overall simi- larity of the effects of the different institutional decisions on com- pliance. Although there are some minor differences, a decision by a local legislature generally has roughly the same impact on com- pliance as does a decision by a local judge or the Supreme Court. For instance, 52.5 percent report that they would be very or some- what likely to do nothing after a legislative decision. The compa- rable figure for a decision by a local judge is 53.5 percent. Slightly greater compliance is observed after a decision of the Supreme Court: 61.8 percent are unlikely to do anything further. Thus, while there is little difference between the impact of the local court and the local legislature, to the extent that the figures differ, the Supreme Court has a slight edge in generating acceptance of its decisions. Overall, the differences across institution are surpris- ingly slight.17

16 Note that the actual text of the question directed toward the Supreme Court was, "How likely is it that you would do nothing at the moment but base your vote in the next election on getting new justices on the Court?"

17 A different way of assessing institutional differences involves changing the unit of analysis from the respondent to the respondent-institution. In this format, the responses to the compliance questions are "stacked" on top of one another and thus do not distinguish between the type of institution, which is

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

Table 1: Intolerant Behavioral Propensities

Likelihood of Action (%)

Very Somewhat Somewhat Very Action Likely Likely Unlikely Unlikely Totala

Try to get the government to stop the demonstration Try to get people to go to the demonstration and stop it in

any way possible, even if it meant breaking the law Try to get the group not to hold its demonstration, even if

it meant doing some things against the group that are illegal

Do nothing to try to stop the demonstration from taking place

25.2 6.4

4.6

23.6

28.2 10.4

10.7

29.0

22.9 26.9

28.6

26.0

23.6 56.3

56.2

21.5

99.9 (933)b >

100.0 (936) U

100.1 (937) 0

0 T1

100.1 (928) C

C-I n 3A -Q

a Those who said "don't know" were omitted from the calculations. b N is in parentheses.

C! z T141 u tI II

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 479

Table 2. Intolerant Behavioral Propensities, After Institutional Intervention

Likelihood of Action (%)

Very or Somewhat Somewhat or Very

Action Likely Unlikely Totala

Try to get people to go to the demonstration and stop it in any way possible, even if it meant breaking the law

Local legislature Local judge U.S. Supreme Court

Try to get the institution's decision reversed by some other government body

Local legislature Local judge U.S. Supreme Court

Do nothing at the moment but vote against the [decisionmaker] at the next election

Local legislature Local judge U.S. Supreme Court

Do nothing to try to stop the demonstration from taking place

Local legislature Local judge U.S. Supreme Court

15.4 13.8 13.4

48.7 47.5 41.3

72.3 71.8 73.0

47.5 46.5 38.2

a Those who said "don't know" were omitted b N is in parentheses.

84.6 100.0 (936)b 86.2 100.0 (460) 86.6 100.0 (476)

51.3 100.0 (931) 52.5 100.0 (453) 58.7 100.0 (471)

27.7 100.0 (924) 28.2 100.0 (457) 27.0 100.0 (466)

52.5 100.0 (935) 53.5 100.0 (456) 61.8 100.0 (475)

from the calculations.

The data in Table 2 do not control for the response prior to the institutional decision and may therefore be misleading because, quite apart from any questions of institutional legitimacy, individu- als vary in the likelihood of taking action in political disputes. Consequently, Table 3 reports the likelihood of behavioral intoler- ance after the legislative decision, controlling for the initial behav- ioral propensity. The first hypothesis is that legislative action in- creases the probability of doing nothing. To simplify matters, the

then used as a variable to assess response differences. This approach yields statistically significant differences across institutions only on the "do nothing" and "try to reverse the policy" items. However, for both of these eta equals .08, indicating an extremely weak level of association. Thus, it is appropriate to conclude that institutional differences in compliance responses are trivial.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

480 UNDERSTANDINGS OF JUSTICE

Table 3. The Influence of Institutional Legitimacy on Behavioral Acceptance of Political Tolerance: Doing Nothing

Likelihood After Initial Likelihood of Acting (%) Institutional Decision Unlikely Likely

Local legislature Unlikely to act 71.0 31.7 Likely to act 29.0 68.3 Total 100.0 (486)a 100.0 (439)

Local judge Unlikely to act 68.0 36.4 Likely to act 32.0 63.6 Total 100.0 (242) 100.0 (206)

U.S. Supreme Court Unlikely to act 73.1 49.6 Likely to act 26.9 50.4 Total 100.0 (242) 100.0 (228)

a N is in parentheses.

data from the summary item on taking any action in response to the decision are presented.

Nearly one-third of those who were initially inclined to take action (very or somewhat unlikely to do nothing) report being less likely to act after the legislative decision. It seems that the legisla- tive decision significantly affected the chance of anti-democratic action. On the other hand, among those initially disinclined to act, the legislative decision seems to have made nearly 30 percent more predisposed to take action. Thus, while the legislative decision de- creased the likelihood of action among those initially predisposed to act, it increased the likelihood among those initially predisposed not to act.

The results are similar after a local court decision on the re- quest to demonstrate.18 Among those initially inclined to act, 36.4

percent became less likely to act after the local court decision (see Table 3). Similarly, among those who probably would not have ac- ted initially, 32.0 percent became more likely to act. Thus, the court decision increased the probability of compliance but at the same time tended to increase the likelihood of action among those who were initially disinclined to act.

Table 3 also reports the results of a similar question focused

18 It must be noted that the court questionsfollowed the questions on the local legislative decision. Thus, Table 3 shows the effect of both the legislative decision and the ruling of the court. It was simply not feasible to use a split ballot to randomize the order of the legislative and court stimuli. Moreover, the most common situation in real political conflict is for the legislative branch (e.g., the city council) to act first and for the legislative action to be reviewed by a court.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 481

on a decision by the Supreme Court.19 Although the findings are generally like those for the local court and local legislature, the ef- fect of the Court decision is substantially stronger. Nearly one- half of those initially inclined to do something to block the demon- stration report being less so inclined after a decision by the Supreme Court. The effect is not overwhelming-50.4 percent still intend not to comply with the Court's decision-and there is an- other group (26.9 percent) who would be mobilized to action by the decision, but there is nonetheless some evidence of the Court's ca- pacity to engender compliance with unpopular political decisions.20

Why would an institutional decision incite action among re- spondents who apparently were not so inclined in the first place? This is a bit of a puzzle, and perhaps an important one. First, it is important to rule out the possibility that the changes in both direc- tions are simply random. While random responses would produce approximately equal numbers of respondents showing greater and lesser willingness to take actions, they would not yield a pattern of consistent reactions to different institutions. In fact, there are moderately strong relationships between willingness to take action after the legislative and judicial decisions. For the legislative and lower court items, the correlation is .51; for the legislative and Supreme Court items, the correlation is .44 (based on collapsing those who did not change and those who were less likely to act). This suggests that some portion of the responses is due to the at- tributes of the individual, not of the institution.21

19 For the sake of simplicity, the data in Table 3 have been dichotomized. Using the original 4-category responses ranging from "very likely" to "very unlikely," among those initially predisposed to act, 42% became less likely to act after a legislative ruling, 47% became less likely to act after a local court decision, and 59% became less likely to act after a Supreme Court decision. From these data, it appears that institutional decisions are important, that there is a slightly greater impact on compliance of judicial decisions, and that the Supreme Court is the most effective of the 3 institutions.

20 A different way to isolate the effect of court action on acceptance of an unpopular political decision is to use as the base line the action probabilities after the legislative decision rather than the initial likelihood of acting. This reveals that nearly one-half of those who were initially quite inclined to act were less likely to do so after a local court ruling, while slightly more than one-half were less likely to act after a Supreme Court ruling. Among those who would probably take action, 29% were less likely to act after a local court decision (23% were more likely to act), while 23% were less likely to act after a Supreme Court decision (17% were more likely to act). Thus the Supreme Court is slightly but not greatly more effective at blocking action.

21 Another possible explanation is that some sort of social desirability ef- fect is causing the probabilities of action to increase among those initially predisposed to do nothing. Assuming that action is the socially desirable re- sponse, there could be an unwillingness to report repeatedly to the interviewer that no action would be taken. This tendency probably inflates the proportion claiming to be willing to take action and does so disproportionately among those initially disinclined to act.

On the other hand, there is no relationship between a widely used mea- sure of susceptibility to social desirability (Crowne and Marlowe, 1960) and willingness to take action before or after any of the institutional decisions, nor does level of self-esteem correlate with the behavioral responses. This sug-

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

482 UNDERSTANDINGS OF JUSTICE

Table 4. The Influence of Institutional Legitimacy on Behavioral Acceptance of Political Tolerance: Get People to Stop the Demonstration

Likelihood After Initial Likelihood of Acting (%) Institutional Decision Unlikely Likely

Local legislature Likely to act 95.1 32.0 Unlikely to act 4.9 68.0 Total 100.0 (777)a 100.0 (155)

Local judge Likely to act 94.7 45.5 Unlikely to act 5.3 54.5 Total 100.0 (376) 100.0 (79)

U.S. Supreme Court Likely to act 94.7 44.4 Unlikely to act 5.3 55.6 Total 100.0 (398) 100.0 (76)

a N is in parentheses.

It is possible that institutional action to allow the demonstra- tion elicits a response because the respondent perceives that the demonstration is more imminent. Perhaps some imagined, when first asked about taking action, that it was unimportant to do any- thing because the government would prohibit the demonstration. Once they were told that the government had decided to allow the

demonstration, they saw no further obstacles and decided they had to do something themselves. I cannot be certain about this inter-

pretation, because no data are available to test the hypothesis that the institutional action increased the perceived threat of the dem- onstration.

It is also useful to consider illegal and extra-legal action more

directly by focusing on the responses to the item asking whether the respondent would be likely to "try to get people to go to the demonstration and stop it in any way possible, even if it meant

breaking the law." The hypothesis is that legislative and judicial rulings reduce the likelihood of this sort of behavior. These data are shown in Table 4.

Relatively few respondents report any likelihood of taking di- rect action to stop the demonstration. Those so inclined, however, show some tendency for an institutional ruling to decrease the

gests that the conclusions are unaffected by problems of a socially desirable tendency toward political action.

Moreover, based on interviewer judgments, there is no evidence that these subjects were any less interested in the interviews, or had a lower understand- ing of the questions. Using more objective measures, the respondents did not have less political knowledge and their opinions on the demonstration were not more intense.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 483

chance of anti-democratic behavior. For instance, a legislative de- cision reduces the chances of direct action for 32 percent of those initially predisposed to act. Comparable figures for the local courts and Supreme Court are 45.5 percent and 44.4 percent, re- spectively. Unlike the findings regarding any sort of action, there is little difference between the local courts and the Supreme Court in ability to dissuade illegal action. Nor is there a similar mobiliza- tion effect, for only a handful of respondents would be stimulated by an institutional decision to engage in this sort of behavior.22

Thus, these data provide some evidence that institutional rul- ings affect the likelihood of compliance with tolerant but unpopu- lar policies, and that the combined effect of judicial institutions and legislative actions is slightly more efficacious on this score than legislative action alone. The effect, however, is modest. More importantly, there is no direct evidence that anything about the in- stitutional decision is related to perceptions of its fairness. Per- haps the observed effect is due to nothing more than pragmatic judgments about the likelihood that resistance to institutional de- cisions will produce results. Once a decision has been made, through whatever sorts of processes, people may simply judge that the odds of overturning it are slight. It is therefore necessary to consider the procedural justice hypothesis more directly.

D. Institutional Procedures

The literature on procedural justice suggests that willingness to accept and comply with an unpopular decision is in part a func- tion of perceptions of the fairness of the procedures used to make the decision (see, e.g., Adler, et al., 1983; McEwen and Maiman, 1981, 1984; Tyler, 1984, 1988; Tyler, et al., 1985; and Lind and Tyler, 1988). This suggests that perceptions of the institutional decision- making processes may contribute to compliance. More specifically, I hypothesize that to the extent the decisionmaking processes of these institutions are perceived as fair, the behavioral conse- quences of intolerant attitudes will be minimized.

Table 5 reports data on the perceptions of the decisionmaking process in the local government council (such as the city council and the board of alderman), in the local courts, and in the Supreme Court. First, it should be remembered that the numbers of cases for the judicial institutions is roughly one-half that of the local council due to the split-ballot format. Second, a significant portion of the respondents claims to be uncertain as to the nature of decisionmaking within these institutions. The table shows the respondents perceiving the decisions to be fair as a percentage of

22 Summarizing the uncollapsed data (not reported in the table), 38% of those very likely to engage in direct action to stop the demonstration were less so inclined after a legislative decision; while 53% and 50% were less so inclined after a local court and Supreme Court decision, respectively.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

484 UNDERSTANDINGS OF JUSTICE

Table 5. Perceptions of Procedural Fairness in Institutional Decisionmaking

Percent Perceiving Procedural Fairness Local Local U.S. Supreme

Decisional Attribute Legislature Court Court

Considers the views of all sides 35.0 26.5 26.2 to an issue before making a decision

Gives interested citizens an 68.2 54.7 49.5 opportunity to express their views before making a decision

Makes decisions only after 52.2 62.2 73.2 they assemble all the relevant information on an issue

Can be counted on to make 57.5 63.4 69.7 decisions in a fair way

all respondents answering the question (the uncertain respondents are included in the denominator). Third, most respondents per- ceive the decisionmaking processes within all three institutions to be relatively fair. Of those who are not uncertain, substantially more perceive the institutional decisionmaking to be fair rather than unfair. The most negative sentiment is on whether the local council considers the views of all sides to an issue before making a decision (35% believe that the council seldom does so).

Nonetheless, there are some important inter-institutional dif- ferences. The institution perceived to make decisions most fairly is the United States Supreme Court, in part because it makes deci- sions on the basis of full information. On the other hand, the Court is seen as less likely to give citizens the opportunity to ex- press their views, even if it does consider all sides to an issue.

Generally, the local legislature is perceived as most fair in terms of accessibility to citizens. Fully two-thirds perceive the lo- cal council as providing citizens an opportunity to be heard, even if

they are less sanguine about whether their views are actually con- sidered in making decisions. While the local legislature and local courts are evaluated similarly in overall judgments in decision- making fairness, local courts are perceived as being more likely to make a decision based on full information.

For all three institutions, the strongest correlate of general evaluations of fairness ("counted on to make decisions in a fair way") is the item concerning whether all relevant information is assembled before making a decision (Pearson correlation coeffi- cients, r, range from .50 to .59). Second most important in each in-

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 485

stance is the item on whether citizens are given an opportunity to express their views (r correlations range from .35 to .48). Those who perceive the local council as fair are generally likely to per- ceive local courts as fair (r = .43) and are somewhat more likely to perceive the United States Supreme Court as fair (r = .27). The crucial question for this analysis is whether these perceptions of decisionmaking fairness have any impact upon willingness to com- ply with the decisions of the institution. This hypothesis can be tested using the data shown in Table 6.

Perceptions of the fairness of the decisionmaking processes within these institutions have virtually no impact on willingness to accept the institutional decision as final and binding. I base this conclusion on the measures of the likelihood of challenging the action of the institution shown in Table 6. For instance, con- sider the likelihood of doing nothing after the institutional deci- sion. The correlation between perceptions of institutional fairness and the likelihood of doing nothing is .01 for the legislative deci- sion: -.01 for the local court decision; and .01 for the Supreme Court decision. Even when we control for the initial likelihood of acting, there is little relationship.23 It is clear that citizens report that they are no more willing to accept an unpopular decision if they perceive that it is fairly made than when they do not perceive that it is fair.

A different way of testing this hypothesis is simply to regress willingness to comply with the decision of the institution on per- ceived fairness in decisionmaking processes. With or without the control for initial behavioral propensities, the conclusions are the same: Perceptions of procedural fairness have little or no impact on willingness to comply. The maximum R2 observed is .04, which results from regressing acceptance of the decision of the local council on the four perceived attributes of its decisionmaking pro- cess. Perceptions of procedural justice in Supreme Court and local court decisionmaking have no impact whatsoever on willingness to accept the judicial decisions as final. Where institutional decision- making processes are perceived as being more fair, acceptance of the unpopular institutional decision is not enhanced.24 This analy- sis suggests that to the extent that there is any impact on compli- ance that flows from institutional ratification of unpopular deci-

23 Due to differences in question wording, this comparison can only be made for 2 sorts of activity: doing nothing to try to stop the demonstration and trying to get people to go to the demonstration and stop it in any way pos- sible, even if it meant breaking the law. The analysis is of the difference in the likelihoods after and before the respondent is given information about the decision of the institution. The only significant relationship observed is be- tween perceptions of local court fairness and the likelihood of doing something (p = .02), but it is not monotonic and not interpretable.

24 Shifting the unit of analysis to the respondent-institution merely con- firms this conclusion. When willingness to comply is regressed on the 4 fair perceptions and 2 dummy variables representing the 3 institutions, the R2 is .02.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

486 UNDERSTANDINGS OF JUSTICE

Table 6. Impact of Perceptions of Procedural Fairness on Compliancea

Procedural Perceptions

Opportunity Full Considers to Express Information Fair

Action All Sides Views Decisions Decisions

Local legislature Try to stop the

demonstration, even illegally

Try to get the government's decision reversed

Vote against the decisionmakers in the next election

Do nothing Local court

Try to stop the demonstration, even illegally

Try to get the government's decision reversed

Vote against the decisionmakers in the next election

Do nothing U.S. Supreme Court

Try to stop the demonstration, even illegally

Try to get the government's decision reversed

Vote against the decisionmakers in the next election

Do nothing

-.00

-.09

.02

.06

.01

-.17

.02

-.03

.05

-.11

-.04

.07

.13

-.07

.03

.02

.16 .09

-.05 -.03

-.10 .03

-.01 .01

.03 -.04

-.08

-.18

-.07

-.06

.08 -.03

-.01

.07

-.01

-.11

-.04

-.01

.03

-.03

.03

.01

.08 -.01

.01

.03

.00 .01

a Entries are Pearson correlation coefficients. All variables have been coded so that negative relationships are hypothesized.

sions, this impact is not associated with the perceived fairness of the institutional decisionmaking processes.25

25 It is most curious that to the extent that the perceived fairness of the institutional decisionmaking processes has any impact on compliance, it is the legislative institution that is effective, not the judicial institutions. This is curi- ous because the argument is commonly made that judicial institutions have

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 487

E. Legitimacy

I have shown that a decision by the Supreme Court has some impact on compliance but that the impact is not associated with perceptions of institutional fairness. Why then are citizens more likely to accept an unpopular political decision after the Supreme Court has rendered its opinion? One reason may be that the Court is perceived as a legitimate institution, irrespective of whether its decisionmaking procedures are perceived as fair (or for that mat- ter not perceived at all). As Rasinski, Tyler, and Fridkin (1985) and Tyler (forthcoming) have shown, perceptions of legitimacy have an independent impact on compliance. Indeed, the original Dahl (1957) hypothesis was grounded in assumptions about the le- gitimacy of the Supreme Court.

No direct measure of legitimacy is available here, and indeed it is not obvious that legitimacy can be directly measured in survey research. Instead, I employ an index of diffuse support26-a closely related concept (cf. Easton, 1975)-and I hypothesize that those who are more supportive of the Court as an institution are more likely to accept its decisions as authoritative.

The data provide no support for the hypothesis: The relation- ship between the diffuse support measure and willingness to ac- cept the Supreme Court's decision is a trivial .07, and is not in the hypothesized direction. Like procedural perceptions, it seems that these attitudes have little impact on compliance.

This relationship may be confounded by two factors. First, it does not control for the initial behavioral propensity, and, second, it may be influenced by those who are more likely to act after a Court decision. To consider whether the hypothesis suffers due to these data maladies, I have constructed a more sensitive measure of change in behavior that ranges from negative scores that indi- cate a greater likelihood of compliance (that is, inaction) after the decision to positive scores indicating a greater likelihood of non- compliance after the decision. So the conditional effects can be ob- served, two separate analyses are performed. The first compares those who are more likely to comply with those who were unaf- fected by the Court decision, while the second compares those more likely not to comply with those who did not change. Both

special powers of legitimation beyond those of legislative institutions. Here, however, is evidence (albeit slight) of exactly the opposite effect. Even more curiously, it is not the responsiveness of the local legislature to constituency opinion that is most influential but instead the belief that the local council considers all relevant information in making decisions. One might have ex- pected that this aspect of procedural fai ness would be the forte of judicial, not legislative, institutions. Because these relationships are so weak, it is perhaps prudent not to try to explain them.

26 The conceptualization, operationalization, and validation of this mea- sure are reported in Caldeira and Gibson (1989). The correlations between dif- fuse support and perceptions of procedural fairness range from .16 to .35.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

Table 7. The Impact of Perceptions of Procedural Fairness and Diffuse Support on Compliance: United States Supreme Court

Change in Behavior* More Likely to Comply More Likely Not to Comply

Attitude r b 3 r b 3

Procedural perceptions Considers all sides .07 .13 .15** -.11 -.09 -.11 Opportunity to express views -.04 -.05 -.06 -.04 -.04 -.05 Full information decisions .04 .06 .05 .04 .01 .01 Fair decisions -.01 -.04 -.04 .06 .07 .07

Diffuse support -.13 -.19 -.20** -.01 .05 .05 R2 .04 .02

* r = Pearson correlation coefficient, b = unstandardized regression coefficient, 1 = standardized regression coefficient. ** Significant at .01 or less

co

vi

z

En

t4

O

ci

0 tI

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 489

the procedural justice variables and diffuse support are employed as predictors. The results are shown in Table 7.

Diffuse support emerges as the best predictor of compliance: Those who are supportive of the Court are significantly more likely to comply with its decisions even when they are disagreea- ble. The other significant relationship-between a procedural per- ception and compliance-is not in the predicted direction: Those who believe that the Supreme Court seldom considers all sides to an issue are more likely to comply. This may reflect the percep- tion that it is futile to challenge the decisions of the Court because the justices are unlikely to take contrary views into account when making their decisions. None of the predictors accounts for those who are more likely to act after a decision by the Court. Indeed, the directions of most of the relationships are not as predicted.

Thus there is some evidence that the legitimacy of the Court, at least as reflected in levels of diffuse support, affects compliance with unpopular decisions. No such evidence exists for perceptions of procedural justice.

F. Opinion Leaders

It is sometimes suggested that while the impact of court deci- sions on the ordinary mass public is not great, their impact on opinion leaders (variously defined) is more substantial. It is these opinion leaders who in turn structure the perceptions of ordinary citizens through a "two-step" flow of information. Thus it is useful to consider whether the research findings reported here also char- acterize opinion leaders.

Opinion leadership is operationalized in terms of self-reports of the frequency with which others ask for the respondent's opin- ions.27 Responses were collected using a trichotomous response set; whether the respondents are very often, sometimes, or hardly ever asked for their opinions. I hypothesize that the effect of pro- cedural perceptions is more significant on opinion leaders.

Opinion leaders are only slightly more likely to be tolerant of the demonstration in the first place. Among those who are intol- erant (the group of concern throughout this article), opinion lead- ers do not differ from other citizens in their perceptions of the procedural elements of judicial or legislative decisionmaking. Ta- ble 8 reports the evidence on whether opinion leaders who per- ceive that the decisions of these institutions are fairly made are more likely to comply with unpopular policies.

There is some evidence that the procedural perceptions of the

27 Two other variables indicate opinion leadership: (1) how often the re- spondent attempts to persuade others how to vote and (2) attentiveness to the United States Supreme Court. None of the substantive findings reported here is dependent upon the choice of the specific indicator of opinion leadership. Additional details on the attributes of the opinion leaders are available from the author.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

490 UNDERSTANDINGS OF JUSTICE

Table 8. The Impact of Perceptions of Procedural Fairness on Procedural Compliance, by Opinion Leadershipa

Frequency of Multiple Correlation Between Perceptions Opinion Giving of Fairness and Compliance

Local legislature Often gives opinions .13 (79) Sometimes gives opinions .05 (374) Hardly ever gives opinions .04 (412)

Local court Often gives opinions .08 (41) Sometimes gives opinions .03 (201) Hardly ever gives opinions .04 (181)

U.S. Supreme court Often gives opinions .14 (38) Sometimes gives opinions .03 (170) Hardly ever gives opinions .01 (228)

a Entries are R2. The figures in parentheses are the numbers of cases on which the coefficients are based. Four procedural perceptions are used as predictors of compliance.

opinion leaders do affect their willingness to accept an institu- tional decision as binding. The most powerful effect is in terms of a decision of the Supreme Court or local legislature, followed by a decision of a lower court. When a decision is perceived as more fairly made, opinion leaders are more likely to report a willing- ness to comply with the unpopular institutional decision. Thus, we find some support for the procedural hypothesis in this seg- ment of the population.

There is also a considerably greater effect of legitimacy on compliance among opinion leaders. The addition of the diffuse support measure adds 7 percent to the explained variance within opinion leaders, while the increment within the other two groups is trivial. Willingness to comply among opinion leaders is thus contingent upon both perceptions of procedural justice and institu- tional support for the Supreme Court.

These conclusions are limited to some degree by the relatively small number of respondents who are identified as opinion leaders, which reflects the fairly small segment of the American popula- tion who see themselves in this role. Thus while the effect of pro- cedural perceptions might not be trivial, the proportion of the pop- ulation affected by such perceptions is quite small. Additional research is needed to test the stability of these findings for special populations.28

28 It is tempting to analyze the views of these opinion leaders further, but the small numbers of cases available for analysis makes it prudent not to ex- pand this line of analysis.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 491

V. CONCLUSION

This analysis shows that institutional intervention can play some role in insuring citizen compliance with unpopular policy de- cisions. To some extent, the antidemocratic implications of the political intolerance so commonly observed in the United States are neutralized when institutions make policy decisions to protect unpopular political minorities. The results suggest that the legiti- mizing capacity of judicial institutions exceeds that of legislative institutions, although not by much. The United States Supreme Court appears to be slightly more effective at stimulating compli- ance than lower courts.

At the same time, the procedural fairness hypothesis does not fare particularly well. Decisions that are perceived as more fairly made, within both the legislative and judicial contexts, do not nec- essarily engender greater compliance within the mass public. Even cross-institutional differences in perceptions of fairness can- not account for the variance in willingness to comply. The effect of procedural concerns is greater among opinion leaders, but this is a relatively small segment of the population, and the research has not been able to address the processes through which opinion lead- ers affect the views of others.

Research based on actual litigation demonstrates a signifi- cantly greater effect of perceptions of procedural fairness (see Lind and Tyler, 1988). Perhaps losing in an actual dispute creates such cognitive discomfort and dissonance that individuals are com- pelled to claim some sort of satisfaction with the decision, even if it is only in terms of how the decision was made. In hypothetical disputes such as the one on which I have focused, this dissonance is surely not generated, making it unnecessary to seek solace through satisfaction with procedure. Moreover, since few have di- rect experience from which to judge the fairness of local courts, the local legislature, or the Supreme Court, the procedural meas- ures are tapping only abstract views, which are notoriously poor predictors of behavior in actual controversies. Assessments of fair- ness grounded in experience are no doubt stronger predictors of compliance (cf. Tyler, forthcoming). These are just some of the ways in which the real and the hypothetical might diverge; there are many more that suggest caution in relying too heavily on the results of this analysis.

At the same time, it is not clear that we should generalize from litigant satisfaction with adjudicatory processes to the more general and diffuse world of politics. At a minimum, litigants are engaged in the dispute, with disengagement being a costly alterna- tive. Citizens involved in civil liberties disputes can easily and rel- atively costlessly withdraw from the conflicts and resume a more normal position of apathy and complacency. To withdraw from lit- igation means losing; to withdraw from a political dispute may ac-

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

492 UNDERSTANDINGS OF JUSTICE

tually mean winning. Moreover, citizens do not bear direct respon- sibility for carrying the dispute in politics. It is possible, indeed profitable, to become a free-rider. As a consequence, the process of disengagement generates few costs. These differences in the types of disputes may well mean that procedural concerns are of lesser significance in politics than in the courthouse.

It is also unclear whether we should generalize these findings beyond the late 1980s. It is possible that the period of sustained ju- dicial activism that the United States has witnessed since the 1960s has eroded any pre-existing abilities of courts to legitimize public policy. As Caldeira (1986) has shown, judicial activism seems to re- duce support for the judiciary (but see Lehne and Reynolds, 1978). Perhaps this activism has exhausted stores of judicial legitimacy that prior to the 1960s were in fact adequate to legitimize public policy. If so, the differences we observe in this analysis may be considerably muted compared to earlier periods of American his- tory.

The courts in the United States also have a special problem of legitimacy that makes them unlikely candidates for generating cit- izen compliance. The decisions of federal courts can always be challenged on accountability grounds. Because federal judges are appointed for life terms, they are not responsible to the electorate, and therefore their unpopular decisions are particularly vulnera- ble to charges of illegitimacy (cf. Cooper, 1988). At the same time, as the election of most state judges has made them more accounta- ble through more competitive and more partisan campaigns, their basic fairness and impartiality have become suspect. Especially with the ticklish issue of campaign contributions from litigants (ac- tual or potential), these courts have few reserves of legitimacy that they can share with other political institutions. Thus, whatever le- gitimizing ability the judiciary may have once had is likely to de- cline (cf. Caldeira and Gibson, 1989).

It seems that something other than procedures leads to accept- ance of unpopular political decisions. What is it about the deci- sions of institutions that increases the likelihood of compliance? Institutional legitimacy is part of the answer, but there are several other possibilities. Citizens may simply comply with the decisions of political institutions out of fear of sanctions. As the dispute over housing integration in Yonkers, New York, in 1988 revealed, many citizens will not abide by bitterly opposed decisions without direct coercion. Perhaps what has been ascribed to legitimacy is actually best attributable to the coercive potential of political insti- tutions (cf. Hyde, 1983).

More generally, compliance may be the result of rational cal- culations by citizens and may have little to do with a sense of obli- gation to countenance unpopular decisions. Coercion may be an important part of this calculus, but so too might estimates of the likelihood of reversing a political decision. Perhaps it is the final-

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 493

ity of the decisions of the Supreme Court that encourages citizens to comply, rather than any notion that the Court is an appropriate and proper institution or that its decisionmaking processes are fair.

Finally, one might question whether legitimacy is necessarily benchmarked by acceptance of unpopular decisions. Is the asser- tion by the attorney general of the United States that a Supreme Court decision is not the final and supreme law of the land a de- nial of institutional legitimacy? When a state legislator introduces a bill to require prayers in public schools, is this action a denial of Court legitimacy? Or when a presidential candidate calls for flag- salute legislation that appears to be inconsistent with prior Court decisions, is this a repudiation of the legitimacy of the institution? Perhaps all can agree that the failure to make this assertion or to introduce this bill or to call for such legislation might be indicative of institutional legitimacy, but challenges to public policy- whether that policy is made by courts or legislatures-are not so clearly evidence of threats to institutional legitimacy. In a polity that is ever changing, it seems unreasonable to use willingness to attempt to change public policy as an index of the illegitimacy of the institution.

From the point of view of civil liberties and political tolerance, these findings suggest that the intervention of courts and legisla- tures may have some limited effect on the willingness of citizens to tolerate their most hated political enemies. Yet it is hard to be sanguine on the basis of these findings. Despite a decision by the United States Supreme Court, a significant proportion of citizens is still willing to challenge the rights of political minorities. To the extent that this intolerance is neutralized, it is probably through a mixture of processes. Some accept the rulings of political institu- tions as authoritative; some view it as futile to pursue dissent any further, and some would attempt to change the decision were there a low-cost means of doing so. Whatever special legitimizing powers the courts have probably play only a minor role in the overall political equation.29

Perhaps the most important contribution of the judiciary under conditions of intense political conflict is that courts vastly slow the process of reaching a decision, thus allowing passions to subside. Courts also employ a language and process of decision- making to which ordinary citizens are not privy. The legitimacy of these institutions then is not marked by citizen willingness to com- ply with unpopular court decisions but is instead registered by citi- zen willingness to submit the issue to the courts in the first place.

29 However, since courts (especially the federal courts) are typically insu- lated from direct majority pressures for tyrannical policy, this finding of a slight difference may be of greater practical consequence. An institution that can influence the majority to accept unpopular decisions but that is relatively insulated from direct accountability to the majority can serve a very important function on the minority rights side of the democratic equation.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

494 UNDERSTANDINGS OF JUSTICE

The manner in which political systems allocate freedom and justice, for the majority and especially for the minority, is of cen- tral importance in the study of politics. This analysis represents a small step toward understanding the complex processes whereby legal institutions can depend on fairness and legitimacy to control political intolerance, and ultimately to contribute to the mainte- nance of democracy.

APPENDIX: THE SURVEY DESIGN

The 1987 General Social Survey (GSS) was conducted in the spring with a response rate of approximately 75 percent. Its sam- ple is a full probability sample. In June and July, the respondents in the 1987 GSS were resurveyed. Of the 1,466 subjects in the spring GSS, 1,106 were eligible to be interviewed. (NORC reserved approximately 350 respondents for possible use in future panel studies. These subjects were not eligible to be re-interviewed.) This subsample was selected randomly, within gender strata. Be- cause the 1987 GSS cross-section had a relatively large differential non-response rate by gender, there is some gender imbalance in the pool of subjects eligible for the second interview. Conse- quently, stratified random sampling was thought desirable. For the re-interviews, males and females were selected with equal probability. Approximately 87 percent of these subjects were re- interviewed. Most of these were in-person interviews, although because some subjects had moved, a small percentage were con- ducted by telephone.

The 1987 GSS also included a special oversample of 353 black respondents, which was also a full probability sample. Extraordi- nary efforts had to be mounted to draw this supplementary sam- ple: roughly five thousand households nationwide were sampled and contacted to locate the black subjects. Their response rate was 79 percent. All of these subjects were eligible for the re-interview, and re-interviews were successfully completed with nearly 90 per- cent of the black subjects in the original sample. The black over- sample also overrepresents females. Because the universe of the black oversample was selected for the re-interview project, no sampling techniques could ameliorate this problem. Thus, inter- views were completed with a total of 1,267 respondents. Since the sample is stratified by race, most analyses are conducted on weighted data.

Special thanks are due Tom Smith and Jim Davis, co-principal investigators on the GSS, and the GSS Board of Overseers for their assistance on this project. Dick Rubin at the NORC was in- strumental in bringing the re-interview survey to its highly suc- cessful conclusion.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

GIBSON 495

REFERENCES

ADAMANY, David (1973) "Legitimacy, Realigning Elections, and the Supreme Court," 1973 Wisconsin Law Review 790.

ADAMANY, David, and Joel B. GROSSMAN (1983) "Support for the Supreme Court as a National Policymaker," 5 Law and Policy Quarterly 405.

ADLER, Jane W., Deborah R. HENSLER, and Charles E. NELSON (1983) Simple Justice: How Litigants Fare in the Pittsburgh Court Arbitration Program. Santa Monica, CA: Rand.

BAAS, Larry R., and Dan THOMAS (1984) "The Supreme Court and Policy Legitimation: Experimental Tests," 12 American Politics Quarterly 335.

BARNUM, David G. (1982) "Decision Making in a Constitutional Democracy: Policy Formation in the Skokie Free Speech Controversy," 44 Journal of Politics 480.

BICKEL, Alexander (1962) The Least Dangerous Branch. Indianapolis: Bobbs-Merrill.

BLACK, Charles, Jr. (1960) The People and the Court. New York: MacMillan. CALDEIRA, Gregory A. (1986) "Neither the Purse nor the Sword: Dynamics

of Public Confidence in the Supreme Court," 80 American Political Sci- ence Review 1209.

CALDEIRA, Gregory A., and James L. GIBSON (1989) "The Etiology of Sup- port for the Supreme Court," Paper presented at the 1989 Annual Meeting of the Midwest Political Science Association, Chicago, Illinois, April 13-15, 1989.

CASPER, Jonathan D. (1976) "The Supreme Court and National Policy-Mak- ing," 70 American Political Science Review 50.

COOPER, Joel (1976) "Deception and Role-Playing: On Telling the Good Guys from the Bad Guys," 31 American Psychologist 605.

COOPER, Phillip J. (1988) Hard Judicial Choices: Federal District Court Judges and State and Local Officials. New York: Oxford University Press.

CROWNE, Douglas P., and David MARLOWE (1960) "A New Scale of Social Desirability Independent of Psychopathology," 24 Journal of Consulting Psychology 349.

DAHL, Robert A. (1971) Polyarchy. New Haven: Yale University Press. (1957) "Decision-Making in a Democracy: The Supreme Court seems a

National Policy-Maker," 6 Journal of Public Law 279. EASTON, David (1975) "A Re-Assessment of the Concept of Political Sup-

port," 5 British Journal of Political Science 435. (1965) A Systems Analysis of Political Life. New York: Wiley.

ENGSTROM, Richard L., and Micheal W. GILES (1972) "Expectations and Images: A Note on Diffuse Support for Legal Institutions," 6 Law & Soci- ety Review 631.

FORWARD, John, Rachelle CANTER, and Ned KIRSCH (1976) "Role-Enact- ment and Deception Methodologies: Alternative Paradigms?" 31 Ameri- can Psychologist 595.

FUNSTON, Richard (1975) "The Supreme Court and Critical Elections," 69 American Political Science Review 795.

GIBSON, James L. (in press) "The Structure of Attitudinal Tolerance in the United States," British Journal of Political Science.

(1986) "Pluralistic Intolerance in America: A Reconsideration," 14 American Politics Quarterly 267.

(1987) "Homosexuals and the Ku Klux Klan: A Contextual Analysis of Political Intolerance," 40 Western Political Quarterly 427.

GIBSON, James L., and Richard D. BINGHAM (1985) Civil Liberties and Na- zis: The Skokie Free-Speech Controversy. New York: Praeger.

GIBSON, James L., and Kent L. TEDIN (1988) "The Etiology of Intolerance of Homosexual Politics," 69 Social Science Quarterly 587.

HAMLIN, David (1980) The Nazi/Skokie Conflict: A Civil Liberties Battle. Boston: Beacon Press.

HYDE, Alan (1983) "The Concept of Legitimation in the Sociology of Law," 1983 Wisconsin Law Review 379.

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions

496 UNDERSTANDINGS OF JUSTICE

JAROS, Dean, and Robert I. MENDELSOHN (1967) "The Judicial Role and Sentencing Behavior," 11 Midwest Journal of Political Science 471.

JAROS, Dean and Robert ROPER (1980) "The U.S. Supreme Court: Myth, Diffuse Support, Specific Support, and Legitimacy," 8 American Politics Quarterly. 85.

LEHNE, Richard, and John REYNOLDS (1978) "The Impact of Judicial Ac- tivism on Public Opinion," 22 American Journal of Political Science 896.

LIND, E. Allan, and Tom R. TYLER (1988) The Social Psychology of Proce- dural Justice. New York: Plenum Press.

McCLOSKY, Herbert, and Alida BRILL (1983) Dimensions of Tolerance: What Americans Believe About Civil Liberties. New York: Russell Sage.

McEWEN, Craig A., and Richard J. MAIMAN (1986) "In Search of Legiti- macy: Toward an Empirical Analysis," 8 Law and Policy 257.

(1984) "Mediation in Small Claims Court: Achieving Compliance Through Consent," 18 Law & Society Review II.

(1981) "Small Claims Mediation in Maine: An Empirical Assessment," 33 Maine Law Review 237.

MUET.TLER, John (1988) "Trends in Political Tolerance," 52 Public Opinion Quarterly 1.

MULLER, Edward N. -(1979) Aggressive Political Participation. Princeton: Princeton University Press.

MURPHY, Walter F., and Joseph TANENHAUS (1969) "Public Opinion and the United States Supreme Court: A Preliminary Mapping of Some Pre- requisites for Court Legitimation of Regime Changes," in J. B. Grossman and J. Tanenhaus (eds.), Frontiers of Judicial Research. New York: John Wiley.

(1968) "Public Opinion and the United States Supreme Court: Mapping of Some Prerequisites for Court Legitimation of Regime Changes," 2 Law & Society Review 357.

RASINSKI, Kenneth, Tom R. TYLER, and Kim FRIDKIN (1985) "Exploring the Function of Legitimacy: Mediating Effects of Personal and Institu- tional Legitimacy on Leadership Endorsement and System Support," 49 Journal of Personality and Social Psychology 386.

STOUFFER, Samuel C. (1955) Communism, Conformity, and Civil Liberties. New York: Doubleday.

SULLIVAN, John L., and George E. MARCUS (1988) "A Note on 'Trends in Political Tolerance,'" 52 Public Opinion Quarterly 26.

SULLIVAN, John L., James E. PIERESON, and George E. MARCUS (1982) Political Tolerance and American Democracy. Chicago: University of Chicago Press.

SULLIVAN, John L., Michal SHAMIR, Patrick WALSH, and Nigel S. ROB- ERTS (1985) Political Tolerance in Context: Support for Unpopular Mi- norities in Israel, New Zealand, and the United States. Boulder: West- view Press.

TYLER, Tom R. (forthcoming) Why People Follow the Law: Procedural Jus- tice, Legitimacy, and Compliance. New Haven: Yale University Press.

(1988) "What Is Procedural Justice?: Criteria Used by Citizens to As- sess the Fairness of Legal Procedures," 22 Law & Society Review 103.

(1984) "The Role of Perceived Injustice in Defendants' Evaluations of Their Courtroom Experience," 18 Law & Society Review 51.

TYLER, Tom R., Kenneth A. RASINSKI, and Eugene GRIFFIN (1986) "Al- ternative Images of the Citizen: Implications for Public Policy," 41 Ameri- can Psychologist 970.

TYLER, Tom R., Kenneth A. RASINSKI, and Kathleen M. McGRAW (1985) "The Influence of Perceived Injustice on the Endorsement of Political Leaders," 15 Journal of Applied Social Psychology 700.

WEBER, Max (1947) The Theory of Social and Economic Organization. New York: Free Press.

CASES CITED

Collin v. Smith, 447 F. Supp. 676 (N.D. Ill. 1978). Collin v. Smith, 578 F.2d 1197 (7th Cir. 1978).

This content downloaded from 66.119.96.100 on Sun, 21 Feb 2016 15:38:42 UTC All use subject to JSTOR Terms and Conditions