Belief in Redeemability and Punitive Public Opinion: “Once a Criminal,Always a Criminal” Revisited

Abstract

In 2009, Maruna and King presented results from a British survey showing that the public’s belief in the redeemability of people who committed offenses curbed their level of punitiveness. Based on a 2017 national survey in the United States (n = 1,000), the current study confirms that redeemability is negatively related to punitive attitudes. In addition, the analyses reveal that this belief predicts support for rehabilitation and specific inclusionary policies (i.e., ban-the-box in employment, expungement of criminal records, and voting rights for people with a felony conviction). Findings regarding measures for punishment and rehabilitation were confirmed by a 2019 Mechanical Turk (MTurk) survey. These results suggest that beliefs about capacity for change among people who committed offenses are key to understanding crime-control public policy.

Keywords

attitudes corrections public opinion rehabilitation punitiveness redeemability

Tracking public opinion since 1971, Enns (2016) has persuasively documented how the growth in punitive attitudes contributed to the United States becoming the “incarceration nation.” The good news, he notes, is that for more than a decade, “public punitiveness has been on the decline and politicians of both parties have noticed” (Enns, 2016, p. 164; see also Pickett, 2019). In this context, policy shifts largely correspond to how people who committed offenses have been socially constructed as either redeemable or the intractable dangerous “other.”

Thus, from the Progressive era into the 1970s, people who committed offenses were portrayed as malleable and worthy of reform (Allen, 1981; Cullen & Gilbert, 1982; Rothman, 1980). The invention of the correctional system—including probation, parole, indeterminate sentencing, and the juvenile court—was predicated on the belief that the criminogenic needs of these individuals could be treated successfully (Rothman, 1980). The goal was “normalization,” with corrections aiming to enrich people who committed offenses in the hope that they would be transformed and “and fit into the main organizing institutions of social life” (Simon, 1993, p. 101). As Garland (2001, pp. 38–39) observes, “penal-welfarism” prevailed in which the state was responsible “for the care of offenders as well as their punishment and control.”

In the 1970s, however, the image of people who broke the law changed dramatically. “The recurrent image of the offender,” observes Garland (2001), “ceased to be that of the needy delinquent or the feckless misfit and became much more threatening—a matter of career criminals, crackheads, thugs, and predators” (p. 102). In Thinking About Crime, James Q. Wilson (1975) offered the stunning conclusion that “Wicked people exist. Nothing avails except to set them apart from innocent people” (p. 235). Two decades later, DiIulio (1995, p. 23) termed juveniles “super-predators,” describing them as “hardened” and “remorseless” and as having “absolutely no respect for human life and no sense of the future.” Garland (2001, p. 136) captured these views under the umbrella of the “criminology of the other,” in which “offenders are treated as a different species of threatening, violent individuals for whom we can have no sympathy and for whom there is no effective help.” More recently, Simon (2014, p. 23) has documented the emergence in this area of the “prevailing wisdom” that “most criminals have a high and unchanging potential for criminal activity.” This conception of such individuals was inextricably linked to the view that keeping society risk-free required a policy of “total incapacitation.” “The politics of fear that produced mass incarceration,” notes Simon (2014, p. 131), “relied on the image of the prisoner as an unchanging lethal threat.”

These portrayals of individuals’ criminality as a stable or fixed entity sound increasingly debatable—even if they were prevalent merely a decade ago. First, the “criminology of the other” in which all people who break the law are seen as inherently and permanently risky has been challenged by the “criminology of the diverse.” Essentially, politicians are now disaggregating the criminal class. A favorite category “non-violent drug offenders” are depicted as posing no danger and as candidates for prison release (see Obama, 2017). The recent federal reform, The First Step Act of 2018, includes as a core component a “risk and needs assessment system” to determine individuals’ risk of recidivating and to develop individualized treatment plans (Office of the Attorney General, 2019, p. 1). Second, the language of redemption is now common (see Cullen et al., 2020). Recall the “REDEEM Act” proposed by Senators Cory Booker and Rand Paul in 2015. Similarly, in the U.S. House of Representatives, Doug Collins, a conservative Georgia Republican, offered the Prison Reform and Redemption Act (Cohen, 2019). Most notably, at the White House signing ceremony for the First Step Act, President Trump commented that the “bipartisan bill will make our communities safer and give individuals released from prison a second chance at life after they have served their time, so important” (Schwartz, 2018). He then added the following: “Redemption is at the heart of the American Idea, and that’s what this is about” (Schwartz, 2018).

The point is that public beliefs about people who committed offenses, especially about redeemability, may be consequential when writ large—if they do, in fact, lead to reduced punitiveness and greater support for rehabilitation and other inclusionary policies. Building on the work of Maruna and King (2009), the current project has implications regarding this possibility. A decade ago, Maruna and King (2009) advanced inquiry into the sources of public punitiveness by introducing, measuring, and showing the consequences of belief in individuals’ “redeemability”—that is, belief in “the impermanence of criminality” (p. 15). Their work falls under the umbrella of attribution theory, which examines how people explain behavior and, in turn, how these “attributions” make certain interventions seem logical. Prior to their contribution, one aspect of attribution theory had shaped research: whether the public perceived that the cause of crime was “dispositional” or “situational” (see Heider, 1958). The dispositional perspective attributes crime (and other behaviors) to individual traits or choices and tended to justify the use of punitive correctional policies, whereas the situational perspective attributes crime to more malleable external or social factors that legitimize rehabilitative responses (see, e.g., Cullen et al., 1985)—a connection largely supported in the literature (Cochran et al., 2003; Falco & Turner, 2014; Maruna and King, 2009; Pickett & Baker, 2014; Unnever et al., 2010).

Maruna and King (2009, p. 9) understood, however, that attribution theory in social psychology had evolved to incorporate additional cognitive dimensions, including “stability/instability” (Weary, 1984; Weiner, 2010). In a related development, other social psychologists—led most notably by Carol Dweck—framed the same issue in terms of the “implicit” theories people held about human attributes, distinguishing between “entity” theory that saw traits as fixed and “incremental theory” that saw traits as malleable. As Levy et al. (1998) note, “a fixed versus a malleable view of traits may set up a framework for understanding social information and social judgments” (p. 1422). In an important innovation, Maruna and King (2009, p. 9) called for the application of this reasoning to the study of public opinion about people who committed offenses:

This measure asks how permanent or transient members of the public feel the causes of criminality to be. That is, regardless of the origins of criminal behavior (situational or dispositional), do they believe that “once a criminal, always a criminal” or do they believe that even the most persistent offenders can redeem themselves and turn their lives around.

Maruna and King (2009, p. 9) observed that the public likely has “divergent views on this key issue,” and that such thinking about such individuals has potential consequences (see also Plaks et al., 2009). They hypothesized that those who believed that criminality was a stable, fixed trait—“once a criminal, always a criminal”—would endorse more punitive correctional policies. By contrast, those who believed that these individuals were redeemable—who could change and be saved from a life in crime—would be less inclined to embrace uniformly harsh sanctioning of people considered to be wayward.

To test this thesis, they collected data (n = 941) as part of the Cambridge University Public Opinion Project in which “surveys were sent to a random sample of British household in six wards chosen for their diversity in income and population demographics” (Maruna & King, 2009, p. 13). Their analysis included a four-item measure of redeemability (e.g., “Most offenders can go on to lead productive lives with help and hard work”) and an eight-item measure of punitiveness (e.g., “My general view towards offenders is that they should be treated harshly”). The results proved striking. When added to their model, redeemability beliefs had a “strong negative” relationship with punitiveness (β = −.398) and explained 14% of the variation (Maruna & King, 2009, p. 18).

Although limited, subsequent research on redeemability or similar constructs (e.g., belief in a “growth mindset”) has generally supported this finding (see, e.g., Moss et al., 2019; Rade et al., 2018; Reich, 2017; Sloas & Atkin-Plunk, 2019; Tam et al., 2013 see also Plaks et al., 2009). Notably, several studies have shown that redeemability not only is negatively related to punitiveness but also is positively associated with support for specific progressive policies, such as restorative justice (Moss et al., 2019), the use of parole (Dodd, 2018), expanding housing assistance programs for people returning from prison and increasing employment opportunities (Ouellette et al., 2017), and increasing employers’ willingness to hire persons with a felony conviction (Reich, 2017).

In this context, the current study seeks to build on Maruna and King’s (2009) research in three important ways. First, whereas their study was conducted in Great Britain more than a decade ago, we explore their redeemability thesis in a distinct social context with a 2017 national-level study of American respondents. We use an opt-in internet survey conducted for us by YouGov to field these data. The findings are then confirmed with a second study employing a 2019 Mechanical Turk (MTurk) sample. Second, we control for factors typically included in U.S. public opinion surveys but omitted from Maruna and King’s (2009) research, including political variables (party and ideology), race, and religiosity. As we report below, our analyses replicate their finding that belief in redeemability is negatively related to measures of punitiveness. Third, building on the emerging policy-specific research in this area, we then proceed to show that redeemability is associated not only with lower punitive sentiments but also with greater support for rehabilitation and specific inclusionary correctional policies (e.g., ban-the-box on employment applications, expunging criminal records, and voting rights for people with a felony conviction). As will be noted, these findings suggest that redeemability is a potentially significant source of public policy preferences and warrants continued investigation.

Before presenting the analysis, two important considerations merit discussion. First, a key assumption of the now-voluminous research on attribution theory and studies of redeemability/immutability/growth mindset is that beliefs about people who committed offenses and their capacity to change are “antecedents of punitiveness” and other correctional policies (Tam et al., 2013, p. 603; see also Plaks et al., 2009). The underlying theory is that views of such individuals—in this case, as capable of being reformed—provide the logic or rational basis for policies (such as providing them with opportunities to change). Maruna and King (2009, p. 19) recognized, however, that this premise was theoretical and, with cross-sectional data, they could not rule out the possibility that punitiveness was the source, not the cause, of people embracing an implicit entity perspective and a view of people who committed offenses as irredeemable. But they rejected this view, noting that

although strong, the relationship between belief in redeemability and punitiveness does not suggest circularity between the two concepts. That is, they are not two sides of the same coin . . . After all, support for harsh punishment is not logically incompatible with a belief in redemption. (Maruna and King, 2009, p. 19)

For example, belief in the mutability of people who committed offenses could logically encourage a belief in punitive policies such as deterrence through prison terms, boot camps, and scared straight programs. Furthermore, experimental research shows that when implicit theories are manipulated (e.g., by giving subjects information favoring the entity attribution), social inferences change in a consistent way, thus suggesting a causal influence (Chiu et al., 1997). Field research also demonstrates that those holding incremental theories are more likely to behave accordingly across diverse contexts—to coach their underperforming employees, to respond to transgressions by romantic partners in a constructive way, and to volunteer to help the homeless (Plaks et al., 2009).

In the current study, we offer an empirical check on whether punitiveness renders spurious the significant relationship between redeemability and support for rehabilitation (two studies) and the policies of voting rights for persons with a felony conviction, ban-the-box, and expungement (one study). These measures are described in more detail below. We created a punitiveness scale consisting of three items (support for capital punishment, for harsher courts, and for punishment/protecting society as the goal of prisons), and then we included this with the redeemability measure in the multivariate analyses of rehabilitation and the three inclusionary policies. In all analyses, redeemability remained statistically significant at the p < .001 level, with the β for the rehabilitation scale .644 or higher and the odds ratio for the other policies 2.305 or higher (full results available upon request from the first author). These findings suggest that redeemability is associated with a range of outcomes independent of the respondents’ level of punitiveness.

Second, it is beyond the scope of this article to explore the origins of belief in redeemability, especially individual differences in this orientation. Furthermore, little research on this topic exists. Still, it is possible to identify at least four potential sources for future inquiry to consider. First, using the Big Five, scholars have shown that personality traits—especially the dimension of agreeableness—are related to more positive and hopeful attitudes toward people who committed offenses, including being less punitive (e.g., Olver & Barlow, 2010; Roberts et al., 2013), and to a “redemptive self” that fosters sensitivity to the suffering of others and the desire to pursue goals that benefit others (Guo et al., 2016). Second, those who have a personal growth mindset might be more likely to see the possibility that troubled individuals are malleable and can be reformed. Possessing a growth mindset has been linked to neuroscience factors (Ng, 2018) and parenting practices that avoid communicating to children that failure is debilitating (Haimovitz & Dweck, 2016).

Third, Haidt’s (2012) moral foundation theory argues that the care/harm intuition makes some individuals—usually identified as political liberals—seek to ameliorate, rather than impose pain on others. Accordingly, their morality would likely favor conceptions of people who committed offenses as redeemable and thus as candidates to be saved from a life in crime. Fourth, correctional research shows that those who embrace religious forgiveness (Applegate et al., 2000), are high on empathy (Unnever et al., 2005), and display racial sympathy (Chudy, 2017; Hannan et al., 2019) are likely to be less punitive and more supportive of correctional reform, suggesting they view people who committed offenses as malleable.

Method

Sample

To test the effects of belief in redeemability, we commissioned YouGov to interview a national sample of American adults (18 years and older). YouGov is considered a reliable source of survey data for academic research in social science disciplines, including criminal justice (e.g., Thielo et al., 2019). YouGov administered our survey to a national sample of 1,000 U.S. respondents from March 3–7, 2017.

In fielding the sample, YouGov uses a two-stage, sample-matching design. First, YouGov selects a matched (on the joint distribution of a large number of covariates, e.g., political ideology, voter registration status) sample of respondents from its online panel (more than two million adult U.S. panelists) using distance matching with a synthetic sampling frame (constructed from probability samples, including the Current Population Survey and the American Community Survey [ACS]). It then uses propensity score matching to weight the sample to resemble the U.S. population on the matched covariates (Ansolabehere & Rivers, 2013). Given the matching and weighting of variables, the method of sample selection is assumed disregardable; clear evidence exists showing that findings from YouGov surveys generalize to the U.S. population (Ansolabehere & Schaffner, 2014; Simmons & Bobo, 2015). Moreover, several studies support the notion that YouGov’s sampling design may outperform probability sampling strategies (Kennedy et al., 2016).

When compared with estimates from the U.S. Census and the ACS (in parentheses), our weighted sample looks much like the U.S. population: non-Hispanic White, 66.8% (64.5%); male, 48.5% (48.7%); bachelor’s degree, 26.5% (28.4%); married, 44.1% (48.2%); Northeast, 18.7% (17.2%); Midwest, 20.1% (20.9%); South, 36.0% (38.1%); West, 25.3% (23.8%). When compared with the Pew Research Center’s estimates of party identification among registered voters (in parentheses), our weighted sample also looks like the U.S. population: lean Republican or Republican, 34.5% (42%); lean Democrat or Democrat, 43.9% (50%). Given these similarities in major population demographics, we have confidence that the sample generalizes to all American adults.

Dependent Variables

The dependent variables in our analyses gauge the respondents’ support for a variety of punitive and inclusionary criminal justice policies. To assess the respondents’ attitudes toward punitive crime policies, we focus on three widely used measures of punitiveness (see Enns, 2016): support for the death penalty, support for harsher courts, and belief that the main goal of prisons should be punitive, rather than rehabilitative. The wording for these measures was taken from questions employed for decades by the General Social Survey (death penalty and harsher courts questions) and by the Harris Poll (main goal of prisons question; see Cullen et al., 2000; Enns, 2016). Thus, we measured the respondents’ support (0 = oppose/no opinion, 1 = favor) for the death penalty by asking “are you in favor of the death penalty for a person convicted of murder?” Harsher courts was measured by asking the respondents “in general, do you think the courts in this area deal too harshly or not harshly enough with criminals?” Responses were coded such that 0 = don’t know/about right/too harsh and 1 = not harsh enough. Support for a punitive goal of prisons (0 = rehabilitation/not sure, 1 = punishing the individual/protecting society) was measured by asking the respondents

what do you think should be the main emphasis in most prisons—punishing the individual convicted of a crime, trying to rehabilitate the individual so that he or she might return to society as a productive citizen, or, protecting society from future crimes he or she might commit?

As noted, the purpose of this research is to assess the impact of redeemability not only on punitiveness but also on policies that seek to “include,” rather than “exclude” people who committed offenses from the community. Thus, survey items were incorporated to gauge the respondents’ support for four inclusionary policies: correctional rehabilitation, banning the box on employment applications, expunging criminal records, and allowing persons with a felony conviction the right to vote. These specific policies were selected because of the attention they have received as barriers to the full social and legal participation of people who committed offenses in society (e.g., Alexander, 2010; Burton et al., 1987, 2014; Chin, 2017; Jacobs, 2015; Manza & Uggen, 2006; Pager, 2007).

To measure support for rehabilitation, we adapted questions from the prior work of Applegate et al. (1997) and Cullen et al. (1983). Accordingly, support for rehabilitation is a mean index (α = .841, factor loadings = .739–.788) measured with responses to five items that asked how much (1 = strongly disagree, 6 = strongly agree) the respondents supported the following statements: (1) “it is important to try to rehabilitate adults who have committed crimes and are now in the correctional system,” (2) “it is a good idea to provide treatment for offenders who are supervised by the courts and live in the community,” (3) “rehabilitation programs should be available even for offenders who have been involved in a lot of crime in their lives,” (4) “all rehabilitation programs have done is to allow criminals who deserve to be punished to get off easily,” and (5) “I would not support expanding the rehabilitation programs that are now being undertaken in our prisons.” Items 4 and 5 were reverse coded so that higher values on the mean index correspond to greater support for rehabilitation. The use of these reverse-coded items to measure this construct is important in that it reduces the potential for acquiescence bias (Pickett & Baker, 2014).

To measure support for ban-the-box (0 = keep the box, 1 = ban-the-box) policies, the respondents were first asked to read the following introductory paragraph:

As you may know, many job applications contain a “box” that a person applying for the job must check if they have a criminal record from their past. Recently, however, many elected officials have passed ban-the-box laws. These laws say that employers must remove this “box” on job applications that people must check if they have been arrested and/or convicted of a crime. With ban-the-box laws, employers can still conduct criminal background checks and choose to not hire someone who has a criminal record. However, they can only do this AFTER they have looked at the person’s job application and decided to interview them or give them a job offer.

Then, the respondents were asked to choose one of the following options: (1) ban-the-box laws are a good idea because ex-offenders’ skills and qualifications for jobs will be considered. This could help them get jobs because they won’t just be rejected right away for having criminal records”; or (2) ban-the-box laws are a bad idea because they make employers waste time considering hiring people that they may end up rejecting later when they find out about their criminal records.”

Similarly, support for expungement (0 = oppose, 1 = support) was measured by asking the respondents to first read the following introductory paragraph:

There has been some debate recently about expunging criminal records for offenders who have completed their sentences and thus paid for their crime. When a criminal record is expunged, this means that the criminal record is removed or sealed and thus is no longer something that the public, including employers, can see. It’s like starting over from scratch. We would like to know your views on this matter. Some people argue that expunging criminal records is a good policy because it gives criminal offenders the opportunity to wipe their slate clean and get their lives back on track. Other people believe that expunging criminal records is a bad policy because public access to criminal records helps keep communities safe.

The respondents then were asked to choose one of the following options: (1) “expunging criminal records is a good policy because it gives criminal offenders a chance to get their lives back on track,” or (2) “expunging criminal records is a bad policy because public access to criminal records helps keep communities safe.”

We should note that the ban-the-box and expungement questions contain arguments for and against each policy. This approach was chosen because some respondents might not be familiar with these policies or reasons given for supporting and opposing them. As a reviewer pointed out, however, it is possible that for people to “support either policy for reasons not included in the response options.” It also is possible that including the phrase “get their lives back on track” in the expungement question might have made the redemptive nature of this reform more salient. As noted below, the consistency of the results across all dependent variables suggests that question-wording did not produce any counterintuitive results. Still, future research should explore whether the content of items affects subjects’ responses.

Finally, to measure whether the respondents believe individuals convicted of felonies should retain the right to vote, the respondents chose from one of the three options: (1) they should lose permanently their right to vote, (2) they should lose their right to vote only until they have completed their sentence, and (3) they should not lose their right to vote at all. Thus, Felon voting rights was coded such that 0 = they should lose permanently their right to vote/they should lose their right to vote only until they have completed their sentence and 1 = they should not lose their right to vote at all.

Note that we made our original coding decisions on the dependent variables to try to capture respondents who clearly were punitive (favored capital punishment or harsher courts) or were clearly in favor of progressive policy proposals. To check that these coding choices did not affect the results, we recoded all these variables and reran the analyses. None of the findings with regard to redeemability were affected (redeemability remained significant at the p < .001 level). Specifically, we conducted the analyses excluding the respondents who answered “not sure” and “don’t know,” and we also ran multinomial logistic regressions where applicable (e.g., the models predicting Harsher Courts, Punish and Protect, and Felon Voting Rights). Ordered logit models were not estimated because these data violated the proportional odds assumption. Nonetheless, we found no substantive differences.

Independent Variable

To measure the respondents’ belief in redeemability, we adapted questions from Maruna and King (2009). Three of the four questions were the same or quite similar to those used by Maruna and King in their four-item scale (Items 1, 3, and 4 below). Given the focus on adults, we chose not to use their item on “young offenders” and replaced it with Item 2. Similar measures of this construct have been used in prior research (e.g., Dodd, 2018; Leverentz, 2011; Ouellette et al., 2017). Thus, redeemability is a mean index (α = .718, factor loadings = .490–.781); note that the Cronbach’s alpha for Maruna and King’s (2009) measure was .64. Our measure of redeemability was created from the respondents’ opinions (1 = strongly disagree, 6 = strongly agree) to the following four statements: (1) “most offenders can go on to lead productive lives with help and hard work,” (2) “given the right conditions, a great many offenders can turn their lives around and become law-abiding citizens,” (3) “most criminal offenders are unlikely to change for the better,” and (4) “some offenders are so damaged that they can never lead productive lives.” Items 3 and 4 were reverse coded such that higher values on the mean index represent a greater belief in redeemability. Again, because items in this scale were asked in both directions (i.e., belief in and non-belief in redeemability), the potential for acquiescence bias is lessened (Pickett & Baker, 2014).

Control Variables

We control for a number of factors that theory and prior research suggest may be confounders of the relationships between our independent and dependent variables. These include the respondents’ race (1 = White), gender (1 = Male), age (in years), marital status (1 = Married), household composition (1 = Child in Residence), education (1 = no high school, 6 = graduate degree), income (1 = < US$10K, 16 = US$500K+), political party affiliation (1 = Republican), and ideology (1 = Conservative). As is common in YouGov surveys, there was a comparatively large amount of item missing data on the income variable (13.5%). Using linear regression imputation, we imputed the missing income values based on the scores on the other variables in the analysis. We also control for religious beliefs. Religiosity is a standardized mean index (α = .741, factor loadings = .511–.787) based on three questions assessing the importance of religion in the respondents’ lives, their frequency of church attendance, and their frequency of praying. Finally, we control for the respondents’ region of residence (1 = Southerner). Table 1 provides the descriptive statistics for all the variables included in the analyses and the bivariate correlations between each independent and dependent variable included in the analyses.

Table 1:

Descriptive Statistics and Correlation Matrices of Independent and Control Variables With the Outcomes

Variables	M or %	SD	Bivariate correlation matrices with outcomes
Variables	M or %	SD	Death penalty	Harsher courts	Punitive prison goal	Rehabilitation	Ban-the-box	Expungement	Felon voting rights
Outcome variables
Death penalty (% favor)	54.53	—	—	—	—	—	—	—	—
Harsher courts (% favor)	40.12	—	—	—	—	—	—	—	—
Punitive prison goal (% favor)	53.50	—	—	—	—	—	—	—	—
Felon voting rights (% favor)	76.64
Rehabilitation	4.19	1.02	—	—	—	—	—	—	—
Ban-the-box (% favor)	64.72	—	—	—	—	—	—	—	—
Expungement (% favor)	45.06	—	—	—	—	—	—	—	—
Independent variable
Redeemability	3.56	0.87	−.41*	−.44*	−.45*	.69*	.32*	.37*	.38*
Control variables
Republican (%)	23.46	—	.19*	.17*	.14*	−.22*	−.11*	−.14*	−.20*
Conservatism (%)	34.66	—	.29*	.22*	.23*	−.31*	−.21*	−.22*	−.28*
Religiosity	0.01	0.81	.10	.13*	.08	−.16*	−.07	−.06	−.11*
Married (%)	44.10	—	.08	.08	.05	−.12*	−.02	−.10	−.07
Child in residence (%)	24.64	—	.07	.01	−.01	−.05	−.01	.03	−.04
Southerner (%)	36.00	—	−.01	−.02	.03	−.01	.05	.01	−.03
Education	3.17	1.46	−.06	−.10	−.03	.18*	.05	.02	.06
Income	5.28	3.38	.01	−.04	.06	.05	−.02	−.09	.02
Age	48.08	16.92	.10	.18*	.12*	−.19*	−.01	−.17*	−.14*
Male (%)	48.48	—	.03	−.06	.04	.01	−.06	.04	.01
White (%)	66.76	—	.13*	.05	.09	−.01	.02	−.07	−.06

Note. Bonferroni corrected p values reported. “*” denotes correlation is significant at p < .001, which is the new adjusted α = .05 level.

Although the control variables cover key demographics, political party and ideology, marital status, region, and religiosity, the data set did not include measures of other factors that could affect policy attitudes, such as racial resentment, crime salience (e.g., fear or perceived risk of crime victimization), and causal attributions of crime. As will be seen, the findings suggest that the effects of redeemability are robust. In this context, future research should examine the stability of this finding with a range of other control variables included in the statistical model. Note that in the Maruna and King (2009) study, redeemability retained statistical significance with crime salience (fear) and causal attribution in the analysis. Furthermore, as discussed later, we present a second analysis from an Amazon MTurk survey that adds confidence to the conclusion that redeemability is a robust predictor of policy preferences.

Analytic Plan

In Tables 1 and 4, we present both descriptive statistics and bivariate correlations between each independent variable and outcome variable in the study. Because of the number of tests run on these data in Tables 1 and 4 (84 correlations computed in Table 1 and 48 in Table 4), we present our results using Bonferroni adjusted p values to account for the possibility of inflated Type 1 error. After adjusting the significance threshold using the Bonferroni correction, we found that correlations with p values less than .0005 (rounded to < .001) would attain significance at the new α = .05 level. We then proceed with regression analyses to explore the relationship between redeemability and the outcome variables, independent of the effects of the control variables. To do this, we estimate seven regression analyses in Table 3 (six binary logistic and one ordinary least squares [OLS]) and four regression analyses in Table 5 (four binary logistic and one OLS). We report odds ratios for models with binary outcomes and beta values for models with continuous outcomes.

Results

Analysis Based on the YouGov Survey

This project’s central focus is on the effect of belief in the redeemability of people who committed offenses on a range of punitive and inclusionary correctional policies. Table 2 reports the percentage agreeing and disagreeing with the four items measuring belief in redeemability. Notably (see Items 1 and 2), about 8 in 10 respondents stated that “most” or “many” of these individuals can “lead productive lives” and “become law-abiding citizens.” Achieving this outcome will require “help and hard work” and “the right conditions.” Redemption thus is seen as achievable. On the other hand (see Items 3 and 4), more than half (57.3%) of the sample agreed that “most criminals are unlikely to change,” and about 8 in 10 (82.5%) expressed the view that “some” are so “damaged that they can never lead productive lives.” In a way, the American public might be seen as having a nuanced view of those in the correctional system—viewing redemption as a possibility for many people who committed offenses, but not as a realistic goal for all them.

Table 2:

Public Belief in Redeemability

Items	% total agree	% total disagree	% strongly agree	% agree	% somewhat agree	% somewhat disagree	% disagree	% strongly disagree
1. Most offenders can go on to lead productive lives with help and hard work. (n = 989)	78.7	21.3	11.8	25.3	41.6	14.1	5.6	1.6
2. Given the right conditions, a great many offenders can turn their lives around and become law-abiding citizens. (n = 993)	80.8	19.2	13.3	27.8	39.6	12.9	4.3	2.1
3. Most criminals are unlikely to change for the better. (n = 990)	57.3	42.7	9.1	16.3	31.9	30.1	9.5	3.1
4. Some offenders are so damaged that they can never lead productive lives. (n = 989)	82.6	17.4	21.8	30.5	30.3	9.8	5.0	2.6

Note. Statistics are weighted using the person weights from the public use file for the full 2010 ACS sample. The number of respondents answering each question differs and is listed. ACS = American Community Survey.

Table 3 reports the multivariate analyses for the effect of redeemability on three punitive policies, support for rehabilitation, and three policy proposals that have gained increasing notoriety—ban-the-box on employment applications, the expungement of criminal records, and the extension of voting rights to persons with a felony conviction. We estimated logistic regression models for the three punitive measures as well as the models predicting support for felon voting rights, ban-the-box, and expungement. We used OLS regression for estimating the model predicting support for rehabilitation. Beyond specific significant findings (e.g., Whites less supportive of the death penalty), two variables have consistent relationships in the expected direction across the dependent variables. First, conservative political ideology is significantly associated with increases in support for the death penalty and for a punitive prison goal (i.e., punishment or protect society as the main goal of prisons) and with significant decreases in support for rehabilitation, ban-the-box, expungement, and voting rights. Conservatism is not significantly related to harsher courts, though the relationship is in the expected direction.

Table 3:

Impact of Redeemability on Support for Correctional Policies

Variables	Punitive policies			Inclusionary policies
	Death penalty	Harsher courts	Punitive prison goal	Felon voting rights	Rehabilitation		Ban-the-box	Expungement
	OR (SE)	OR (SE)	OR (SE)	OR (SE)	b (SE)	β	OR (SE)	OR (SE)
Redeemability	0.388*** (.05)	0.312*** (.05)	0.248*** (.04)	3.399*** (.61)	0.765*** (.04)	.657	2.523*** (.33)	2.514*** (.30)
Republican	1.035 (.28)	1.153 (.30)	0.820 (.21)	0.586* (.15)	−0.055 (.08)	−.024	1.212 (.29)	1.033 (.27)
Conservatism	2.383** (.64)	1.414 (.35)	1.920** (.47)	0.605* (.15)	−0.190* (.09)	−.093	0.590* (.15)	0.565* (.14)
Religiosity	1.070 (.14)	1.206 (.17)	1.134 (.15)	0.873 (.13)	−0.063 (.04)	−.051	0.937 (.12)	1.090 (.14)
Married	1.484 (.32)	1.469 (.31)	0.950 (.20)	1.065 (.25)	−0.153* (.06)	−.078	0.955 (.19)	0.965 (.20)
Child in residence	1.173 (.21)	0.831 (.14)	0.653** (.12)	0.745 (.12)	0.010 (.07)	.006	1.398 (.26)	1.256 (.19)
Southerner	0.905 (.07)	0.892 (.06)	1.112 (.08)	1.049 (.08)	0.030 (.02)	.044	1.145* (.08)	1.018 (.07)
Education	1.034 (.07)	1.005 (.07)	1.080 (.07)	0.944 (.07)	0.047* (.02)	.073	1.007 (.07)	0.933 (.07)
Income	0.947* (.02)	0.954 (.03)	0.981 (.03)	1.053 (.04)	0.025*** (.01)	.093	0.961 (.03)	0.984 (.03)
Age	0.996 (.01)	1.010 (.01)	0.991 (.01)	0.985 (.01)	0.001 (.01)	.019	1.015* (.01)	0.992 (.01)
Male	1.188 (.24)	0.635* (.13)	0.937 (.19)	0.859 (.19)	−0.001 (.06)	−.001	1.083 (.22)	1.352 (.26)
White	1.704* (.39)	0.991 (.24)	1.285 (.28)	1.040 (.28)	0.117 (.07)	.057	1.405 (.32)	1.144 (.25)
N	998	998	999	1,000	1,000		978	981
R ²	.155	.165	.180	.188	.502		.111	.118

p < .05. **p < .01. ***p < .001 (two-tailed).

Most notable, given the thrust of this project, belief in redeemability is the variable with the strongest and most consistent relationship with punitive and inclusionary policies. As seen in Table 1, where bivariate correlations are presented, redeemability is significantly related to the seven outcomes in the predicted direction. These results remain in the multivariate analyses reported in Table 3, irrespective of the control variables. For the three punitive correctional policies, embedded in national surveys for decades, redeemability is associated with decreases in support for the death penalty, making courts harsher, and prisons as serving punitive goals. With regard to inclusionary policies, this belief is strongly related to the scale measuring support for rehabilitation (β = .657). Equally instructive, redeemability is significantly and positively related to three very specific policy proposals—support for ban-the-box, expungement, and voting rights for persons with a felony conviction. Thus, across a diversity of correctional policies, measured in different ways, redeemability has robust relationships.

Analysis Based on the Amazon MTurk Survey

To assess further the potential effects of redeemability on public policy preferences, we were able to use a March 2019 MTurk survey—designed for other purposes (see Graham, 2019)—that included the identical measures used in the current study for redeemability (α = .792, factor loadings = .755–.827) and four policy measures (support for the death penalty, harsher courts, punitive prison goal, and rehabilitation). Beyond attempting to replicate the results reported above, we wished to assess the robustness of the effects for redeemability by including three variables omitted in the YouGov survey. First, based on the work of Kinder and Sanders (1996), we incorporated a standard four-item scale measuring racial resentment (α = .878, factor loadings = .844–.872)—again, a variable found to be a strong predictor in public opinion research on crime policies and social policies (see, e.g., Bobo & Johnson, 2004; Chudy, 2017; Filindra & Kaplan, 2016; Hutchings & Valentino, 2004; Pickett & Baker, 2014; Unnever et al., 2008). Second, to measure crime salience, a one-item measure was added that measured anger about crime (“When you think about crime in your community, do you feel . . .” “Not angry at all,” “A little angry,” “Moderately angry,” “Angry,” “Very angry.”). Third, to measure crime attribution, we used an item available in the survey called blank slate, tapping whether criminal behavior was attributed either to a crime gene or crime gene X environment interaction (= 0) or to purely nongenetic factors (= 1). This item was used to capture whether the attribution was more dispositional or situational in orientation. Note that the results for redeemability remained the same when this attribution was measured with the crime gene X environment response included in the blank slate category with nongenetic factors.

Note that MTurk surveys use respondents who opt-in to participate in return for a small payment—in this survey, US$0.15 per minute totaling US$3.75 for their participation (see Thielo et al., 2019). Compared with the YouGov sample, the MTurk sample is similar in party membership and percent married, but less conservative, more educated, and more supportive of redeemability (M = 3.80 versus 3.56; see Tables 1 and 4). Because the analysis included racial resentment, the MTurk sample was restricted to Whites (n = 396). More details on the measures and survey are available from the senior author.

Table 4:

Descriptive Statistics and Correlation Matrices of Independent and Control Variables With the Outcomes, MTurk Study

Variables (n = 396)	M or %	SD	Bivariate correlation matrices with outcomes
Variables (n = 396)	M or %	SD	Death penalty	Harsher courts	Punitive prison goal	Rehabilitation
Outcome variables
Death penalty (% favor)	40.7	—	—	—	—	—
Harsher courts (% favor)	21.5	—	—	—	—	—
Punitive prison goal (% favor)	40.9	—	—	—	—	—
Rehabilitation	4.0	0.9	—	—	—	—
Independent variable
Redeemability	3.8	0.9	−.210*	−.258*	−.323*	.766*
Control variables
Racial resentment	0.6	0.2	.349*	.218*	.311*	−.429*
Blank slate (%)	52.3	—	−.002	−.055	.034	.063
Anger about crime	2.5	1.1	.129	.111	.168	−.194*
Republican (%)	24.2	—	.240*	.092	.296*	−.311*
Conservatism (%)	27.5	—	.261*	.091	.258*	−.254*
Married (%)	45.5	—	.039	.066	.014	−.066
Education	4.26	1.3	−.079	−.065	.007	−.130
Full-time employment (%)	71.2		.049	.061	.109	−.202*
Income	4.2	1.5	.043	.005	.053	−.058
Age	39.5	11.6	.017	.055	.013	.138*
Male (%)	58.3	—	.011	−.132*	−.026	−.055

Note. Bonferroni corrected p values reported. “*” denotes correlation is significant at p < .001, which is the new adjusted α = .05 level. MTurk = Mechanical Turk.

As seen in Table 4, the zero-order bivariate correlations between redeemability and the four dependent variables were significant and in the expected direction. More instructive, Table 5 reveals that redeemability remains significant in the multivariate models, even with a range of control variables in the analysis. Consistent with previous research, racial resentment was also significant across all models, associated with increases in punitive sentiments and decreases in support for rehabilitation. The R² for the four models was similar to the results in the YouGov survey (see Tables 3 and 5). These results lend credence to the conclusion that belief in redeemability is an important source of crime policy views.

Table 5:

Impact of Redeemability on Support for Correctional Policies, MTurk Study

	Death penalty		Harsher courts		Punitive prison goal		Rehabilitation
Variables	OR	(SE)	OR	(SE)	OR	(SE)	b	(SE)	β
Redeemability	0.744*	(.140)	0.508***	(.164)	0.539***	(.148)	.684***	(.034)	.663
Racial resentment	10.449***	(.559)	6.012**	(.679)	5.486**	(.572)	−.802***	(.134)	−.213
Blank slate	1.083	(.235)	0.752	(.281)	1.467	(.383)	−.051	(.056)	−.028
Anger about crime	1.146	(.105)	1.121	(.121)	1.197	(.107)	−.021	(.025)	−.027
Republican	1.427	(.305)	0.844	(.238)	2.188*	(.313)	−.180*	(.078)	−.087
Conservatism	0.691	(.301)	1.071	(.347)	0.804	(.309)	.076	(.077)	.038
Married	0.982	(.239)	1.405	(.282)	0.798	(.246)	−.081	(.058)	−.045
Education	0.860	(.095)	0.863	(.112)	0.983	(.097)	−.028	(.023)	−.041
Income	1.088	(.083)	0.973	(.100)	1.073	(.086)	.011	(.020)	.018
Full-time employment	1.111	(.266)	1.512	(.322)	1.520	(.276)	−.174**	(.062)	−.089
Age	1.003	(.011)	1.011	(.013)	1.006	(.011)	.005*	(.003)	.067
Male	1.097	(.242)	0.482**	(.280)	0.811	(.250)	.024	(.057)	.013
R ²	.156		.124		.197		.654

Note. Models contain only White respondents (n = 396). MTurk = Mechanical Turk.

p < .05. **p < .01. ***p < .001 (two-tailed).

Discussion

Not that long ago, research confirming the results of important prior studies was devalued based on the rationale that “we already know this” or “this relationship has already been shown.” In the past decade, however, there has been a growing recognition in both the sciences generally and in criminology in particular that a replication crisis may well exist (see, e.g., Ioannidis, 2005; Kulig et al., 2017). For a host of reasons—both statistical and contextual—reported results in any single study might prove idiosyncratic and not reflect empirical reality. Organized skepticism, a core norm of the scientific enterprise, advises that provisional findings be tested for their stability. The current study is an effort that takes on significance in this regard by reexamining Maruna and King’s (2009) finding that belief in redeemability is a strong predictor of public punitiveness.

Thus, using a near-identical measure of redeemability, we assessed whether this belief in the impermanence of criminality—that is, the conviction that persons who have broken the law have the capacity to become law-abiding—reduces support for punishment. Importantly, the current study was conducted a decade later and, in another country (the United States rather than the United Kingdom). Maruna and King’s (2009) key finding that redeemability decreases punitiveness was replicated in a national-level sample (weighted to be representative of the American public) and in an MTurk national sample. In retrospect, this result makes “common sense.” If respondents hold the belief that people who committed offenses are permanently criminal, if not dangerous, then punishment serves the purpose of retribution (exacting just deserts), protection (incapacitation), and coercion (compelling conformity through nasty means). By contrast, if the hope exists that such individuals can change for the better, then the perceived utility of punishment evaporates. No longer viewed as “the other,” the challenge now becomes how the correctional system can help those with good still within them to live a good life.

In this vein, it is predictable that those holding redeemability beliefs would possess a rehabilitation-oriented correctional ideology. Since the founding of the American penitentiary in the 1820s, the conviction that those considered to be wayward could be saved from a life in crime has justified reformative correctional practices (Cullen & Gilbert, 1982; Rothman, 1980). Still, the salience of belief in redeemability as a source of public opinion is seen in the finding that this belief was significantly related to support not only for the global goal of rehabilitation but also for specific policy proposals—including such diverse measures as ban-the-box on employment applications, the expungement of criminal records, and the extension of voting rights of people convicted of a felony. Taken together, these results are empirically remarkable. It appears that belief in redeemability predicts opinions on virtually every correctional policy—whether that is punitive or rehabilitative or whether the survey probes general ideology or specific policy proposals (see also Dodd, 2018; Moss et al., 2019; Ouellette et al., 2017; Reich, 2017).

This fact has implications for the future study of public opinion on punitiveness and correctional policy preferences. When primary data are collected, it now seems imperative that a measure of redeemability be included in the analysis, both to avoid a misspecified model and with the intent of further replicating the Maruna and King (2009) now-classic article. As in any area, however, advances in concept and measurement will need to be forthcoming (see Graham, 2019). At least two improvements in this line of inquiry merit consideration.

First, we recommend returning to Maruna’s (2001) Making Good to capitalize on his analysis not only of redemption scripts (belief in the possibility of change and a good life) but also condemnation scripts (belief in the inevitability of a life in crime). The inclusion of condemnation scripts is relevant because these involve the belief that people who committed offenses are ensnared in life-course-persistent criminal trajectory. Furthermore, if Maruna’s discussion of these scripts is unpacked, it is possible to identify components of narratives about these individuals. For example, with regard to condemnation scripts, three components can be extracted: people who committed offenses are (1) doomed to deviance, (2) constrained by circumstances beyond their control, and (3) are unable to live a boring conventional life when the temptation of immediate gratification awaits. Similarly, with regard to redemption scripts, the discussion alerts us to notions that (1) past acts may have been bad but the person who violated the law—the “real me”—is good, (2) the person now has what it takes to be a good citizen and lead a good life, and (3) having experienced the negative consequences of crime while younger, the person now can have a “generative” influence and help others to avoid making the same mistaken choices. Although applied to the subjects in Maruna’s (2001) work, each of these components can be operationalized and included in a multifaceted measure of public views regarding redeemability.

Second, also at issue is how openness to redeemability is shaped by who respondents “have in mind” when answering questions that ask about “offenders” in general. Through experiments and factorial design methods, it would be possible to explore how a range of social and criminal characteristics of people who committed offenses affect belief redeemability. Given that race and racial bias are integral to public policy and public opinion (Alexander, 2010; Pager, 2007; Tonry, 2011; Unnever et al., 2008), any design must incorporate race and ethnicity into the evaluations respondents might make.

A key issue is whether the public will embrace the possibility that people who commit violent offenses, in particular, are capable of reform and returning to society absent of criminogenic risk. “Violent crime is the great divide between punitiveness and nonpunitiveness,” observe Cullen et al. (2000), “The American public is risk averse. It sees no reason to ‘take chances’ with offenders who have shown they will physically hurt others” (p. 59). A 2016 national poll (n = 2,001) by Morning Consult lends credence to this conclusion. The respondents were asked if they would support “reducing prison time” for people who have committed a crime and had “a low risk of committing another crime.” For “non-violent offenders,” 78% supported cutting back sentences, but for “violent offenders,” the support was 29%—a 49 percentage point difference (Morning Consult, 2016, p. 3). The challenge is that more than half of all people in state prisons are locked up for a violent offense (Bronson & Carson, 2019). Reversing mass incarceration, thus, will entail the early release of some of these people. The key to doing so might well be showing evidence that these individuals, often years older than when first incarcerated, have shown clear evidence of their redeemability (e.g., completed treatment programs, no violence during their imprisonment). In the popular media, the individuals serving time in prison portrayed in Shawshank Redemption—most notably the characters Andy Dufresne (Tim Robbins) and Boyd “Red” Redding (Morgan Freeman)—exist as plausible images. It is instructive that Shawshank Redemption is now the internet’s favorite movie and is ranked as the number one film of all time (ahead of The Godfather) in polls taken in both the United States and United Kingdom (O’Callaghan, 2017).

As this discussion suggests, research on redeemability has clear policy implications. As noted previously, it is difficult for anyone entering criminology at this time to comprehend fully the role of the “criminology of the other” in shaping past discourse about crime and crime-control policy (Garland, 2001). These images justified a host of get-tough policies aimed at the “total incarceration” of the wicked and incurable (Simon, 2014). Especially when attributed to members of the so-called Black underclass, the denial of redeemability led to what Simon (1993, p. 259) disquietingly identified as “the waste management model,” which deployed surveillance technologies and incapacitation policies that would “allow this population to be maintained securely at the lowest possible cost.” “Expensive techniques of discipline, training, and normalization are not warranted,” observed Simon (1993), “if the basic assumption is that there is no realistic potential to alter the offender’s status as toxic waste” (p. 259).

If we are indeed at a correctional turning point as some have argued (see, e.g., Clear & Frost, 2014; Petersilia & Cullen, 2015), then it is likely that the prevailing rhetoric about people who committed offenses has indeed changed—and can be documented through content analyses of political speeches and news reports and editorials. We should find as well that redeemability beliefs underlie important policy developments such as the now-widespread reentry movement that was initiated around 2004 (Jonson & Cullen, 2015; Mears & Cochran, 2015; see also Rade et al., 2018). Consider, for example, the comments of then-governor of Ohio, Republican John Kasich (2015) in which he was explaining why he had endorsed progressive policies (e.g., sentencing reform, addiction treatment, reduction in collateral consequences):

Look, redemption is real, second chances are real. We need to not only practice that individually, but we need to practice that collectively . . . We’re all in this together restoring a human being’s hope, opportunity, and purpose. It changes the world.

Finally, belief in redemption is not only related to punitiveness and support for rehabilitation, the correctional ideological orientations often probed in public opinion studies (Cullen et al., 2000; Maruna & King, 2009). As shown in our analyses, this faith that people who committed offenses can change also leads to support for specific policies that would foster the inclusion, rather than the exclusion, of these individuals from society. We focused on ban-the-box, criminal record expungement, and the voting rights of people with a felony conviction. Maruna (2001, 2011a, 2011b), however, has suggested a next step to this reform—“redemption rituals” or “rehabilitation ceremonies” (see also Cullen et al., 2020). If public belief in redeemability can become more widespread and intense, then the ideological space would exist to create official ceremonies that would restore to people who committed offenses the legal rights of all American citizens. Thus, the goal would be to translate belief in redeemability into the correctional policy of redemption (Cullen et al., 2020).

Footnotes

ORCID iD

Alexander L. Burton

Alexander L. Burton is a doctoral student in the School of Criminal Justice at the University of Cincinnati. His research interests include corrections and public opinion on corrections and criminal justice policy. He is a coauthor of “Punishing rampage: Public opinion on sanctions for school shooters,” Justice Quarterly (2020).

Francis T. Cullen is a Distinguished Research Professor Emeritus and Senior Research Associate in the School of Criminal Justice at the University of Cincinnati. He is a past president of the American Society of Criminology and the Academy of Criminal Justice Sciences. His current research interests include redemption as a correctional policy, the impact of racial beliefs on public punitiveness, and the criminology of Donald Trump.

Velmer S. Burton, Jr. is a Senior Vice Chancellor for University Strategy and Performance and Professor of Criminal Justice at the University of Arkansas at Little Rock. He holds a Ph.D. from University of Cincinnati and an EdD from the University of Pennsylvania. He is a coauthor of “Confronting the collateral consequences of a criminal conviction: A special challenge for social work with offenders,” Journal of Forensic Social Work (2014).

Amanda Graham is an assistant professor in the Department of Criminal Justice and Criminology at Georgia Southern University. She received her PhD from the University of Cincinnati in the School of Criminal Justice. Her research interests focus on policing, police legitimacy, public opinion, scale development, and measurement.

Leah C. Butler is an assistant professor, University of Nebraska, Omaha. Ph.D. Candidate, School of Criminal Justice, University of Cincinnati. Coauthor of “Public opinion about the death penalty,” Routledge Handbook on Capital Punishment (2018).

Angela J. Thielo is an assistant professor of criminal justice at the University of Louisville. She received her PhD from the University of Cincinnati in the School of Criminal Justice. Her research interests center on the sustainable implementation of effective correctional practices and public attitudes toward correctional policy. She was coeditor of a special issue on “Downsizing prisons” that appeared in Victims & Offenders.

References

Alexander

(2010). The new Jim Crow: Mass incarceration in the age of colorblindness. The New Press.

Allen

F. A.

(1981). The decline of the rehabilitative ideal: Penal policy and social purpose. Yale University Press.

Ansolabehere

Rivers

(2013). Cooperative survey research. Annual Review of Political Science, 16, 307–329.

Ansolabehere

Schaffner

B. F.

(2014). Does survey model still matter? Findings from a 2010 multi-mode comparison. Political Analysis, 22, 285–303.

Applegate

B. K.

Cullen

F. T.

Fisher

B. S.

(1997). Public support for correctional treatment: The continuing appeal of the rehabilitative ideal. Prison Journal, 77, 237–258.

Applegate

B. K.

Cullen

F. T.

Fisher

B. S.

Vander Ven

(2000). Forgiveness and fundamentalism: Reconsidering the relationship between correctional attitudes and religion. Criminology, 38, 719–754.

Bobo

L. D.

Johnson

(2004). A taste for punishment: Black and White Americans’ views on the death penalty and the war on drugs. Du Bois Review, 1, 151–181.

Bronson

Carson

E. A.

(2019). Prisoners in 2017. Bureau of Justice Statistics, U.S. Department of Justice.

Burton

V. S.

Jr. Cullen

F. T.

Travis

L. F.

(1987). The collateral consequences of a felony conviction: A national study of state statutes. Federal Probation, 5(3), 52–60.

10.

Burton

V. S.

Jr. Fisher

C. M.

Jonson

C. L.

Cullen

F. T.

(2014). Confronting the collateral consequences of a criminal conviction: A special challenge for social work with offenders. Journal of Forensic Social Work, 4, 80–103.

11.

Chin

(2017). Collateral consequences. In Luna

(Ed.), Reforming criminal justice: Punishment, incarceration, and release (Vol. 4, pp. 371–395). Arizona State University.

12.

Chiu

C.-Y

Hong

Y.-Y.

Dweck

C. S.

(1997). Lay dispositionism and implicit theories of personality. Journal of Personality and Social Psychology, 73, 19–30.

13.

Chudy

(2017). Racial sympathy in American politics [Unpublished doctoral dissertation, University of Michigan].

14.

Clear

T. R.

Frost

N. A.

(2014). The punishment imperative: The rise and failure of mass incarceration in America. New York University Press.

15.

Cochran

J. K.

Boots

D. P.

Heide

K. M.

(2003). Attribution styles and attitudes toward capital punishment for juveniles, the mentally incompetent, and the mentally retarded. Justice Quarterly, 20, 65–93.

16.

Cohen

D. M.

(2019). Justice, not jailbreak: The context and consequences of the First Step Act. Victims & Offenders, 14, 1084–1098.

17.

Cullen

F. T.

Clark

G. A.

Cullen

J. B.

Mathers

R. A.

(1985). Attribution, salience, and attitudes toward criminal sanctioning. Criminal Justice and Behavior, 12, 305–331.

18.

Cullen

F. T.

Fisher

B. S.

Applegate

B. K.

(2000). Public opinion about punishment and corrections. Crime and Justice, 14, 1–79.

19.

Cullen

F. T.

Gilbert

K. E.

(1982). Reaffirming rehabilitation. Anderson.

20.

Cullen

F. T.

Gilbert

K. E.

Cullen

J. B.

(1983). Implementing determinate sentencing in Illinois: Conscience and convenience. Criminal Justice Review, 8, 1–16.

21.

Cullen

F. T.

Lee

Butler

L. C.

Thielo

A. J.

(2020). Rehabilitation and redemption: Building a new corrections. In Chouhy

Cochran

J. C.

Jonson

C. L.

(Eds.), Advances in criminological theory. Criminal justice theory: Explanations and effects (Vol. 26, pp. 309–335). Routledge.

22.

DiIulio

Jr. (1995, November 27). The coming of the super-predators. Weekly Standard, 23–28.

23.

Dodd

(2018). The punitive woman? Gender differences in public attitudes toward parole among an Australian sample. International Journal of Offender Therapy and Comparative Criminology, 62, 3006–3022.

24.

Enns

(2016). Incarceration nation: How the United States became the most punitive democracy in the world. Cambridge University Press.

25.

Falco

D. L.

Turner

N. C.

(2014). Examining causal attributions towards crime on support for offender rehabilitation. American Journal of Criminal Justice, 39, 630–641.

26.

Filindra

Kaplan

N. J.

(2016). Racial resentment and Whites’ gun policy preferences in contemporary America. Political Behavior, 38, 255–275.

27.

Garland

(2001). The culture of control: Crime and social order in contemporary society. University of Chicago Press.

28.

Graham

(2019). Measuring procedural justice: A case study in criminometrics [Unpublished doctoral dissertation, University of Cincinnati].

29.

Guo

Klevan

McAdams

D. P.

(2016). Personality traits, ego development, and the redemptive self. Personality and Social Psychology Bulletin, 42, 1551–1563.

30.

Haidt

(2012). The righteous mind: Why good people are divided by politics and religion. Vintage.

31.

Haimovitz

Dweck

C. S.

(2016). What predicts children’s fixed and growth intelligence mind-sets? Not their parents’ views of intelligence but their parents’ views of failure. Psychological Science, 27, 859–869.

32.

Hannan

K. R.

Cullen

F. T.

Butler

L. C.

Burton

A. L.

Graham

Burton

V. S.

Jr. (2019, November). Racial sympathy and support for capital punishment: Revisiting White Americans’ views toward African Americans [Paper presentation]. Paper presented at the annual meeting of the American Society of Criminology, San Francisco, CA, United States.

33.

Heider

(1958). The psychology of personal relations. Wiley.

34.

Hutchings

V. L.

Valentino

N. A.

(2004). The centrality of race in American politics. Annual Review of Political Science, 7, 383–408.

35.

Ioannidis

J. P. A.

(2005). Why most published research findings are false. PLOS Medicine, 2, 696–701.

36.

Jacobs

J. B.

(2015). The eternal criminal record. Harvard University Press.

37.

Jonson

C. L.

Cullen

F. T.

(2015). Prisoner reentry programs. Crime and Justice, 44, 517–575.

38.

Kasich

(2015). Bipartisan summit: A video message from John Kasich, Governor of Ohio. https://www.youtube.com/watch?v=CrXe6D8yMl0

39.

Kennedy

Mercer

Keeter

Hatley

McGeeney

Gimenez

(2016). Evaluating online nonprobability surveys. Pew Research Center.

40.

Kinder

D. R.

Sanders

L. M.

(1996). Divided by color: Racial politics and democratic ideals. University of Chicago Press.

41.

Kulig

T. C.

Pratt

T. C.

Cullen

F. T.

(2017). Revisiting the Stanford Prison Experiment: A case study in organized skepticism. Journal of Criminal Justice Education, 28, 74–111.

42.

Leverentz

(2011). Neighborhood context of attitudes toward crime and reentry. Punishment & Society, 13, 64–92.

43.

Levy

S. R.

Stroessner

S. J.

Dweck

C. S.

(1998). Stereotype formation and endorsement: The role of implicit theories. Journal of Personality and Social Psychology, 74, 1421–1436.

44.

Manza

Uggen

(2006). Locked out: Felon disenfranchisement and American democracy. Oxford University Press.

45.

Maruna

(2001). Making good: How ex-convicts reform and rebuild their lives. American Psychological Association.

46.

Maruna

(2011a). Judicial rehabilitation and the “clean bill of health” in criminal justice. European Journal of Probation, 3, 97–117.

47.

Maruna

(2011b). Reentry as a rite of passage. Punishment & Society, 13, 3–28.

48.

Maruna

King

(2009). Once a criminal, always a criminal? “Redeemability” and the psychology of punitive public attitudes. European Journal of Criminal Policy and Research, 15(1–2), 7–24.

49.

Mears

D. P.

Cochran

J. C.

(2015). Prisoner reentry in the era of mass incarceration. SAGE.

50.

Morning Consult. (2016). National tracking poll, Project 160812, September 01-02, 2016. Washington, DC: Author.

51.

Moss

S. A.

Lee

Berman

Rung

(2019). When do people values rehabilitation and restorative justice over the punishment of offenders? Victims & Offenders, 14, 32–51.

52.

(2018). The neuroscience of growth mindset and intrinsic motivation. Brain Sciences, 8(2), 20.

53.

Obama

(2017). The president’s role in advancing criminal justice reform. Harvard Law Review, 130, 811–866.

54.

O’Callaghan

(2017, September 22). How the Shawshank Redemption became the internet’s favourite film. BFI Film Forever. https://www.bfi.org.uk/news-opinion/news-bfi/features/shawshank-redemption-greatest-film-frank-darabont

55.

Office of the Attorney General. (2019). The First Step Act of 2018: Risk and needs assessment system. U.S. Department of Justice.

56.

Olver

M. E.

Barlow

A. A.

(2010). Public attitudes toward sex offenders and their relationship to personality traits and demographic characteristics. Behavioral Sciences and the Law, 28, 832–849.

57.

Ouellette

H. M.

Applegate

B. K.

Vuk

(2017). The public’s stance on prisoner reentry: Policy support and personal acceptance. American Journal of Criminal Justice, 4, 768–789.

58.

Pager

(2007). Marked: Race, crime, and finding work in an era of mass incarceration. University of Chicago Press.

59.

Petersilia

Cullen

F. T.

(2015). Liberal but not stupid: Meeting the promise of downsizing prisons. Stanford Journal of Criminal Law and Policy, 2, 1–43.

60.

Pickett

J. T.

(2019). Public opinion and criminal justice policy: Theory and research. Annual Review of Criminology, 2, 405–428.

61.

Pickett

J. T.

Baker

(2014). The pragmatic American: Empirical reality or methodological artifact? Criminology, 52, 195–222.

62.

Plaks

J. E.

Levy

S. R.

Dweck

C. S.

(2009). Lay theories of personality: Cornerstones of meaning in social cognition. Social and Personality Psychology Compass, 3, 1069–1081.

63.

Rade

C. B.

Desmarais

S. L.

Burnette

J. L.

(2018). An integrative theoretical model of public support for ex-offender reentry. International Journal of Offender Therapy and Comparative Criminology, 62, 2131–2152.

64.

Reich

S. E.

(2017). An exception to the rule: Belief in redeemability desistance signals, and the employer’s decision to hire a job applicant with a criminal record. Journal of Offender Rehabilitation, 56, 110–136.

65.

Roberts

S. C.

Vakirtzis

Kristjánsdóttir

Havlíček

(2013). Who punishes? Personality traits predict individual variation in punitive sentiment. Evolutionary Psychology, 11, 186–200.

66.

Rothman

D. J.

(1980). Conscience and convenience: The asylum and its alternatives in Progressive America. Little, Brown.

67.

Schwartz

(2018, November 14). Trump announces criminal justice reform: “Redemption is at the heart of the American Idea.” https://www.realclearpolitics.com/video/2018/11/14/trump_announces_criminal_justice_reform_redemption_is_at_the_heart_of_the_american_idea.html

68.

Simmons

A. D.

Bobo

L. D.

(2015). Can non-full-probability internet surveys yield useful data? A comparison with full-probability face-to-face surveys in the domain of race and social inequality attitudes. Sociological Methodology, 45, 357–387.

69.

Simon

(1993). Poor discipline: Parole and the social control of the underclass, 1890–1990. University of Chicago Press.

70.

Simon

(2014). Mass incarceration on trial: A remarkable court decision and the future of prisons in America. The New Press.

71.

Sloas

L. B.

Atkin-Plunk

C. A.

(2019). Perceptions of balanced justice and rehabilitation for drug offenders. Criminal Justice Policy Review, 30, 990–1009.

72.

Tam

K.-P.

Shu

T.-M.

K.-S.

Tong

Y.-Y.

(2013). Belief about immutability of moral character and punitiveness toward criminal offenders. Journal of Applied Psychology, 43, 603–611.

73.

Thielo

A. J.

Cullen

F. T.

Burton

A. L.

Moon

M. M.

Burton

V. S.

Jr. (2019). Prisons or problem-solving: Does the public support specialty courts? Victims & Offenders, 14, 267–282.

74.

Tonry

(2011). Punishing race: A continuing American dilemma. Oxford University Press.

75.

Unnever

Cullen

Jonson

(2008). Race, racism, and support for capital punishment. Crime and Justice, 37, 45–96.

76.

Unnever

J. D.

Cochran

J. K.

Cullen

F. T.

Applegate

B. K.

(2010). The pragmatic American: Attributions of crime and the hydraulic relation hypothesis. Justice Quarterly, 27, 431–457.

77.

Unnever

J. D.

Cullen

F. T.

Fisher

B. S.

(2005). Empathy and support for capital punishment. Journal of Crime and Justice, 28, 1–34.

78.

Weary

(1984). Current issues in attribution theory and research. Annual Review of Psychology, 35, 427–459.

79.

Weiner

(2010). Attribution theory. In Peterson

Baker

McGraw

(Eds.), International encyclopedia of education (Vol. 6, pp. 558–563). Elsevier.

80.

Wilson

J. Q.

(1975). Thinking about crime. Vintage Books.