Search for a command to run...
Abstract Male prisoners (n = 450) estimated public reaction to the label "ex-con" and endorsed likely stigma management strategies. Although most anticipate significant rejection, they prefer preventative telling to withdrawal and secrecy as adjustments. OLS regression finds factors promoting exclusion (devaluation/discrimination beliefs and reduced job opportunities) related to withdrawal and secrecy, whereas factors favoring inclusion ("wise" family and friends and plans to attend church) increase preventative telling and reduce withdrawal and secrecy. To enhance public safety by increasing ex-convicts' openness and visibility, policies should promote social support and economic opportunities. Successful reentry may improve rule-following in general, whereas exclusionary practices may work against post-release control. The authors are grateful to Jerry Gray, Eliza Pavalko, and several anonymous reviewers for their helpful comments and suggestions. Notes iIntriguingly, Frank Tannenbaum was himself an ex-convict, having served a year in prison for social activism prior to his academic career (Goode Citation1994). iiBecause we were not able to assist with the data collection, we have no way of knowing whether or not packets were distributed in every housing unit It is possible that some staff may have neglected to do so. iiiQuestionnaires that were returned by mail did not differ significantly from those collected by the staff, with the exception of one measure: There was a significantly larger proportion of questionnaires completed by prisoners with indeterminate sentences among those collected by the staff (30.8% versus 20.4% of those mailed in). A control for indeterminate sentences was included in the analyses to correct for this potential source of bias. a Offenses in the sample sum to more than 100 percent because they are not mutually exclusive. In the ODRC data, only the "most serious" crime was counted. b Miscellaneous property crimes include arson, auto theft, and larceny. Source: Bureau of Research/Office of Policy and Offender Reentry: July 1, 2006 Census of ODRC Institutional Population, Demographic and Offense Summary. ivPrison populations are typically characterized by low literacy and educational achievement (Haigler et al Citation1994; Harlow Citation2003). The bias in our sample may merely reflect the willingness or ability of better-educated individuals to fill out a survey. a Offenses are not mutually exclusive. Sex offense is omitted category. Six point "Strongly Disagree" = 1, "Strongly Agree" = 6;® Indicates reversed scoring. vAs Table A2 shows, for our purposes the referent was changed and the language level was reduced, per request of the Ohio Department of Corrections and Rehabilitation Review Board. Three of the statements were created by us, and are not found in prior research. Only one item (Secrecy #1: "When I meet people for the first time, I will not tell them that I was in prison") differed conceptually between this and prior research (Link et al Citation1989), loading on secrecy rather than withdrawal. viPrior income was not included in the analyses for several reasons. On this measure, there were an unacceptably high number of missing values, reducing statistical power. Inflationary influences could not be controlled, given that the respondents could have stopped working either very recently or years previously. Nor was there any way of ascertaining if income reported was entirely attained through legal means. Finally, in trials (not shown) income has no significant relationship with stigma management strategies. viiOther specifications for offenses were tried, including single primary offense, rank ordered offense, and total number of offenses. These variables, when substituted for the total listing of all offenses, did not alter the analyses. Because the total listing (with the three smallest categories combined) is the most comprehensive, it was used in the analyses. The constancy of these results suggests that the lack of mutual exclusivity among offense categories is substantively unimportant. viiiTrials without any controls for offenses were conducted, as were trials with other categories of offenses omitted In each case, results were similar to those shown. Because sex offenders do differ from other prisoners in bivariate analyses of perceived stigmatization and stigma management strategies, and because of the theoretical importance of community notification statutes, sex offender as the reference category was retained. ixBecause a sizeable proportion of our sample (n= 160) had no firm release date, continuous data on actual release dates was substantially incomplete. In order to include information from all respondents, we constructed the variable "release date groups," which includes those with indeterminate as well as determinate sentences and those with life sentences. In models tried (but not shown) each of the individual release date categories (omitting "no parole" and "life") were included as separate measures. None were significant, nor did they change the model estimation. a Sex offense is the reference category; ∗p < .05; ∗∗∗p < .001 (two tailed). a Items significantly above the mean on one-sample t-tests Six point "Strongly Disagree" = 1, "Strongly Agree" = 6; R indicates reverse scoring. xIf item #1 on the withdrawal scale ("I will switch jobs if I think someone knows I am an ex-con"), which is also related to employment, is included in scales with the other three job-related secrecy items, the summary scale mean drops to 260 (SD=118). xiTrials separately employing only scaled job-related items and general concealment items as dependent variables in the regression analyses were also conducted (but not shown). Results indicate that analyzing these factors separately does not increase the precision of the estimates, nor were results appreciably different from trials employing the full secrecy scale. a Sex offense is the reference category; ∗p < .05; ∗∗p < .01; ∗∗∗p < .001 (two-tailed). xiiThe positive relationship between past job experience and withdrawal may reflect past struggles in the job market, or it may reflect a refusal to even try to find a job, with perceived difficulties providing a ready excuse. a Item #1 on Link et al. (Citation1989) withdrawal scale: "It is easier for me to be friendly with people who have been psychiatric patients" (with ex-con referent) did not load on any of these factors. It was retained, but used as a single-item independent variable in the analyses. b This item was item #7 on Link et al. (Citation1989) withdrawal scale. It did not load on withdrawal in these analyses.