“But Everybody Does It…”: The Effects Of Perceptions, Moral Pressures, And Informal Sanctions On Tax Cheating

Abstract

Data from the Notre Dame Study of Catholic Parish Life are used to examine how perceptions of tax evasion within a community affect community members' judgments about the moral wrongfulness of tax evasion, their fear of informal sanctions directed against such behavior, and, ultimately, their inclination to evade taxes in the future. We used multilevel linear structural equation modeling to estimate the direct and indirect effects of these variables. We find that the more prevalent an individual perceives tax evasion to be within the community, the less likely the individual will be to judge the act harshly, the less likely he or she will be to fear informal sanctions directed against it, and the more inclined the individual will be to commit tax evasion in the future. The direct effects of individuals' perceptions, their judgments about the moral wrongfulness of tax evasion, and the threat of informal actions are all statistically significant and in the predicted direction. The indirect effects of such perceptions of offending and the direct effects of moral judgments illustrate how informal social control loses force within a moral community and deviant acts may become increasingly acceptable through process of defining deviancy down.

Notes

Data used in this study were collected with funding provided by the Lilly Endowment, Inc. We would like to thank the anonymous reviewers for their valuable comments.

¹These characteristics may not operate uniformly across all denominations or congregations within a denomination. For example, many mainline Protestant congregations may not have the level of social cohesiveness evident in fundamentalist Protestant congregations and Catholic parishes. Thus, they might fail to reproduce the deterrence process described here.

²The vintage of this dataset is a secondary concern, as the goal of this study is not to estimate marginal distributions of social perceptions and moral judgments about criminal and deviant behavior, but rather to disentangle connections among individuals' perceptions, judgments, and estimates of informal sanction threat. In the absence of a more contemporary set of data providing such community-based measures we consider the CPL best suited for the purposes of the current study.

³Hispanic parishes and parishioners were excluded for several practical reasons: language and translation problems that confront any mailed survey of Hispanic parishioners; divergent cultural patterns among Hispanic groups that complicate the construction of survey items; and finally, budget constraints that limited the scope of the research.

⁴Concern about potential problems attributable to non-response bias led the researchers who originally gathered the data to conduct tests for signs of such bias (see Leege and Welch 1989). Three principal techniques were used: (1) direct calls to an initial wave of non-responders, (2) examination of letters or notes that accompanied any questionnaires that were returned uncompleted, and (3) examination of the last waves of questionnaires that were returned before the survey closed. This last technique was based on the assumption that nonrespondents may resemble more closely those who return their questionnaires later in the process rather than individuals who return their questionnaires earlier. These comparisons examined correlations between the date of return for the questionnaire and a variety of relevant variables (e.g., demographic, ideological, religious, etc.), including the measures of criminal/deviant behaviors. In general, some modest patterns of non-response bias appeared in only 4 of the 36 parish communities that were sampled. In three of these parishes, which were all predominantly white, nonrespondents appeared to he slightly older and somewhat more conservative than other respondents. It is particularly interesting that nonrespondents did not appear to differ significantly from respondents in their willingness to report past misbehavior or their intention to commit such acts in the foreseeable future. In the lone black parish that displayed any evidence of potential problems related to nonresponse bias, respondents were more likely to be women and committed Catholics. Given that black respondents are omitted from our analysis, this differential is rendered irrelevant.

⁵We included race in the model in preliminary analyses, but the model failed to converge largely because of the severe skewness associated with the race variable. Because only 127 black respondents were available for analysis after cases with missing values were deleted, we were unable to perform separate parallel tests of the model on the remaining subsamples of black and white respondents. As a result, we conducted the analysis on the white subsample only. This approach clearly limits the generalizability of our results.

⁶Although the majority of Americans tend to pay their federal income tax once a year, our item captures patterns of other types of income tax payments as well, including annual payments made to states, counties, and municipalities. Moreover, individuals who are employed in certain types of occupations (e.g., regional sales representatives, independent contractors, etc.) that are characterized by monthly or seasonal fluctuations in earnings often opt, or are required, to make their tax payments on a quarterly basis. Thus, some individuals have considerably more than simply five times (one time for each of the five years addressed in the item) in which to cheat on their tax payments.

⁷Although the literature on perceived fear of formal legal sanctions emphasizes certainty of apprehension, studies of informal sanctions have tended to emphasize this dimension somewhat less. Our measure focuses solely on the magnitude of unspecified interactional constraints (i.e., “how difficult would it be for you to get along…”) individuals perceive they would experience when a deviant act is discovered. These constraints, therefore, may represent actual external sanctions (e.g., ridicule, ostracism, other punishment, etc.) imposed by others within the community or the individual's internal reactions (e.g., embarrassment or shame) that derive from such external informal sanction. In this respect the measure resembles some measures of shame and/or embarrassment used in more recent studies (Blackwell 2000).