Abstract
The explanatory power of the moral community hypothesis is evaluated through an analysis of survey data collected from a sample of 11,789 state prison inmates housed in 275 prisons located throughout the United States. The results of the multilevel regression analysis are not consistent with what would be predicted by the moral community hypothesis. The theoretical implications of this line of research are also discussed.
I would like to thank John Galliher, Wayne Brekhus, David Brunsma, Ed Brent, and Greg Casey for helpful suggestions on earlier versions of this article. Access to the data analyzed in this article was provided by membership in the Inter-University Consortium for Political and Social Research. Earlier versions of this article were presented at the 2008 and 2010 annual meetings of the Midwest Sociological Society.
Notes
Note. a Significant correlation with total violations, two-tailed significance test.
Note. a Random effect.
*p < .05. **p < .01. ***p < .001.
Note. a The low value of the alpha coefficient is partially explained by the small number of items that make up the index.
Note. a Random effect.
*p < .05. **p < .01. ***p < .001.
Sturgis (2008) conducted an extensive literature review on studies that had examined the individual level relationship between religiosity and crime-deviance. Of the 74 studies that Sturgis identified, 69 found an inverse relationship between the two variables.
See Sturgis (Citation2008) for details on the literature that has examined the moral community hypothesis.
Johnson et al. (1997) found mixed support for the proposition that participation in Prison Fellowship activities such as religious seminars and bible studies is inversely related to institutional misconduct. In addition, Nelson-Green (Citation1994) determined that there was a positive relationship between aggregate religiosity and the mean level of institutional misconduct in federal prisons.
As noted, data is available on both state and federal inmates. An analysis was also conducted on the federal data, however, due to space limitations and the fact that the same pattern is present in the state and the federal data, the analysis of the federal data is not discussed in this article.
Specifically, the following sampling weights were utilized in this analysis: the weighting control factor, the duplication control factor, the noninterview factor, the offense category ratio adjustment factor, and the control count ratio adjustment factor. For a description of these weights, see the codebook that accompanies the data file (U.S. Department of Justice, 2000, pp. 8–14).
It should be noted that when religiosity was operationalized as belief in an afterlife, four out of five studies that were examined demonstrated the presence of an inverse relationship between religiosity and deviance.
Similarly, Tittle and Welch (Citation1983, p. 662) argue that, “the theoretical significance, empirical substitutability, and comparability provided by attendance make it a reasonable and convenient general indicator of religiosity.”