Bias in Hiring Applicants With Mental Illness and Criminal Justice Involvement: A Follow-Up Study With Employers

Abstract

Negative employer perceptions of job candidates with psychiatric and criminal backgrounds is one factor contributing to high unemployment rates among these groups. The current study replicated and extended Batastini et al., which evaluated stigmatizing beliefs toward hypothetical job applicants who had a known psychiatric and/or criminal history or neither (i.e., healthy control), as well as the effectiveness of a brief training component to mitigate biased attitudes. However, the current study addressed two major limitations of the original study by (a) including participants (N = 259) who reported current, past, or expected hiring experience and (b) including employer benefits in the training component. Results were generally consistent with prior research suggesting that people with psychiatric and criminal histories experience greater stigma from employers; however, the brief educational training component demonstrated minimal impact on reducing negative attitudes regardless of the applicant’s identified psychiatric or criminal background.

Keywords

stigma mental illness criminal justice employment hiring

Stigma refers to the negative outcomes that occur when a social, economic, or other characteristic label is attributed to an individual or a group of individuals sharing a particular quality (Hayward & Bright, 1997; Link & Phelan, 2001). Stigmatized groups or individuals are more likely to be deprived of certain life opportunities when compared with their nonstigmatized counterparts (e.g., Corrigan, 2004; Corrigan & Shapiro, 2010; Corrigan & Watson, 2002; Overton & Medina, 2008). Persons with mental illnesses (PMI) with and without criminal justice (CJ) involvement are vulnerable populations who often experience barriers to job attainment and maintenance (Graffam, Shinkfield, & Hardcastle, 2008; Varghese, Hardin, Bauer, & Morgan, 2009)—both of which are key components to successful mental health recovery and community reintegration (Bonta & Andrews, 2007; Dunn, Wewiorski, & Rogers, 2008; Graffam, Shinkfield, Lavelle, & Hardcastle, 2004; Stuart, 2006). Therefore, the current study aimed to examine the association between stigmatizing attitudes and employment-related outcomes using hypothetical job applicants with and without mental illness (MI) and/or CJ involvement.

While PMI work to manage their psychiatric symptoms, they must also work to overcome misconceptions about their illness (Corrigan & Watson, 2002; Rüsch, Angermeyer, & Corrigan, 2005). Even if an individual is able to effectively manage his or her psychiatric symptoms enough to maintain stable employment, he or she may still face discrimination from employers when actually attempting to secure viable employment opportunities (Corrigan, 2004; Rüsch et al., 2005). For example, Hand and Tryssenaar (2006) found that employers were less willing to hire an applicant with MI compared with their non-MI counterparts. In addition, employers expressed a greater concern regarding certain characteristics of the applicants with MI (e.g., emotional control, persistence, resolving conflict) that might interfere with their work performance than other applicants (Hand & Tryssenaar, 2006). Furthermore, 36% of employees with psychiatric symptoms reported being the target of on-the-job discrimination within the last 5 years (McAlpine & Warner, 2002). Given these stigmatizing beliefs, PMI experience disproportionately high rates of unemployment and often earn lower salaries than employees without psychiatric illness (Alexander & Link, 2003; Burke-Miller et al., 2006; Cook et al., 2006; Link, 1982; Link & Phelan, 2001; Overton & Medina, 2008).

Individuals with a history of CJ involvement seem to face similar difficulties. According to the U.S. Department of Justice (n.d.), more than 10,000 people are released from state and federal prisons every week. With current shifts in the “tough-on-crime” culture in the United States (e.g., moving away from mandatory sentencing), this number will likely increase. Yet, nearly 60% of CJ-involved individuals in the United States are unemployed 1 year post-release from incarceration (Petersilia, 2001). Other data indicate a 35% unemployment rate for ex-prisoners 8 months after their release (Visher, Debus-Sherrill, & Yahner, 2011), whereas the average 12-month unemployment rate for the general population in the same year was 8.9% (Bureau of Labor Statistics, n.d.). It has been estimated that the number of Americans with a criminal record is equivalent to the number of Americans who possess a 4-year college degree (Friedman, 2015). In sum, high rates of unemployment bear poorly on the reduction of psychiatric and criminal recidivism rates for these populations (Duran, Plotkin, Potter, & Rosen, 2013; Harer, 1994; Rauma & Berk, 1987).

For people with criminal and/or psychiatric histories, challenges in seeking employment opportunities can arise from employer bias. For example, more than 60% of employers indicated that they would be unlikely to hire an individual with a criminal record (Holzer, Raphael, & Stoll, 2002). Similarly, employers evidenced significant negative biases when making hiring decisions about applicants with a criminal history when compared with those without a criminal history (Varghese et al., 2009). This result is further compounded by additional racial biases of employers, such that Whites tend to have more job success than ethnic minorities (Pager, 2003). Taken together, it appears that a criminal record negatively affects the employability of CJ-involved individuals (Decker, Ortiz, Spohn, & Hedberg, 2015; Pager, 2003; Pager, Western, & Sugie, 2009). Moreover, people with intellectual or psychiatric disabilities and those with prior CJ involvement were rated the least likely to secure a job when compared with individuals with a history of chronic physical illness or a physical disability (Graffam et al., 2008).

It should be acknowledged that, although stigma negatively affects the employability of individuals with psychiatric symptoms and/or CJ involvement, it certainly is not the only factor that contributes to difficulties in securing employment. For example, these individuals also tend to have deficits in interpersonal skills, cognitive abilities, training or formal education, and stable work histories (Baron & Salzer, 2002; Holzer, Raphael, & Stoll, 2003; Pager, 2003; Visher, Winterfield, & Coggeshall, 2005). Although, there is educational and vocational programming in both mental health and correctional settings designed to counter these deficits (Guerrero, 2011; Hoffmann, Jäckel, Glauser, Mueser, & Kupper, 2014), employer bias remains a major challenge to the employment of PMI and individuals with a history of CJ involvement. For example, Batastini, Bolaños, and Morgan (2014) found that hypothetical job candidates with psychiatric and criminal histories were perceived less favorably than healthy controls even though participants were instructed to assume candidates possessed the appropriate educational and physical qualifications. These findings suggest that negative attitudes can result from the label itself and not necessarily the applicant’s actual job skills.

Existing research has shown that stigma can be minimized through education and personal contact (Hirschfield & Piquero, 2010; Rüsch et al., 2005). For example, employers who reported positive interactions and beliefs about PMI indicated greater willingness to hire an applicant with MI (Hand & Tryssenaar, 2006). In addition, Kosyluk, Corrigan, & Landis (2014) indicated that employers with experience hiring PMI endorsed less stigmatizing attitudes toward these individuals, which indicates that positive hiring experiences can increase the likelihood of hiring individuals with MI.

Using a sample of college students who were instructed to play the role of an employer, Batastini et al. (2014) showed that hypothetical job applicants with a combined psychiatric and criminal history were rated as the least acceptable candidates for the job and were estimated to be the worst potential workers when compared with applicants with a psychiatric or criminal history, or a control group (i.e., no psychiatric or criminal history). However, the presence of brief training about the benefits of work for employees increased positive perceptions, such that applicants with both characteristics (i.e., psychiatric and criminal histories) no longer differed from the other applicants on ratings of job acceptability. Furthermore, the training component appeared to significantly reduce bias toward applicants with only a criminal history, making them just as likely to be hired as the control applicants. Unfortunately, additional education did not affect participants’ decision to hire applicants with a psychiatric history (despite increases in acceptability ratings). Although the results of this study highlighted the challenges that PMI and/or CJ-involved individuals continue to experience when seeking employment, it was limited by its use of college participants as well as the content of the psychoeducation (PE) component (i.e., only detailing benefits of employment to the employee and community).

Therefore, the purpose of the current study is to expand the findings of Batastini et al. (2014) by (a) examining the effects of PE and personal experience on employer attitudes toward a hypothetical job applicant with MI and/or prior CJ involvement using a sample of respondents with current, prior, or anticipated experience with the hiring process and (b) including potential benefits to employers as a component of the PE conditions. In general, it was hypothesized that the results of this study would be similar to those found in Batastini et al. (2014). However, we further hypothesized that, by adding advantages to employers, negative perceptions associated with employability and general stigma would show greater reductions for all applicant types, not just for those with a criminal history. In addition, we hypothesized that greater familiarity with PMI and/or CJ-involved individuals would be associated with fewer biases. Not only will this study provide a more accurate understanding about the role of stigma in hiring applicants with MI and/or prior CJ involvement, but it may also provide further support for a training model aimed at increasing employers’ acceptance of these individuals as worthy job candidates that will, in turn, improve community reintegration and mental health recovery.

Method

Participants

To be included in this study, participants had to be 18 years of age or older, speak fluent English, and either had/currently have a job in which hiring employees was/is part of their responsibilities or were seeking a job in which hiring employees would be part of their responsibilities. That is, the sample included individuals who either are, have been, or will likely be in a hiring position. Of the 249 participants recruited for this study, 150 (60.2%) were female, 96 (38.6%) were male, one (0.4%) identified as transgender, and two (0.8%) did not provide a response. Participants had a combined mean age of 45.68 (SD = 12.62) with 10.14 (SD = 9.63) years of experience working as a human resource or hiring professional. However, most participants (n = 162, 65.1%) indicated that they had not received formal, specialized training in human resource management or employee relations. Approximately 15% of the sample reported no current or past experience hiring.

Measures

Similar to Batastini et al. (2014), participants were asked to complete the following measures in addition to a basic demographics questionnaire:

Two researcher-generated items were used to evaluate participants’ opinions regarding the employability of a hypothetical job candidate (i.e., Joe) after reading a condition-specific vignette (see examples below). The first of these questions asked participants to rate the acceptability of the applicant for a position as a groundskeeper at an apartment complex using a dichotomous yes/no response format (the vignette noted that Joe had all the necessary educational and physical qualifications for the job). In the second question, participants were asked to rate how likely they would be to seriously consider the applicant for the job on a 7-point Likert-type scale (1 = not at all likely, 7 = very likely).

The Generic Work Behavior Questionnaire (GWBQ)

The GWBQ (Michon, Kroon, van Weeghel, & Schene, 2004) is an 18-item measure used to evaluate the general work performance of employees. It contains four subscales, each representing core dimensions of overall work behavior: (a) task competence, (b) initiative/self-confidence, (c) dependability, and (d) social work behavior. Items are composed of opposing statements with responses ranging on a continuum from 1 to 5 regarding the applicant’s performance at his job. Total scores on the GWBQ are derived by averaging item scores across subtests. For the purposes of this study, the original instructions were altered such that participants were asked to rate the hypothetical job applicant according to how well they would expect him to perform in each of the domains assessed by the GWBQ. In the present study, the internal consistency for the GWBQ total score (which was used in the primary analyses) was strong (Cronbach’s α = .94).

The Attribution Questionnaire (AQ)

The AQ (Brown, 2008; Corrigan, 2008) is a 27-item assessment of nine common stereotypes that individuals tend to hold about people with mental illnesses (i.e., blame, anger, pity, help, dangerousness, fear, avoidance, segregation, and coercion; Brown, 2008; Corrigan, 2008). The AQ provides a brief vignette about a man diagnosed with schizophrenia. The vignette discloses information about the man’s life, including factors related to his MI (Corrigan, 2008; Corrigan, Markowitz, Watson, Rowan, & Kubiak, 2003). Brown (2008) identified a six-factor structure of the AQ consisting of (a) Fear/Dangerousness, (b) Help/Interact, (c) Responsibility, (d) Forcing Treatment, (e) Empathy, and (f) Negative Emotions. In the current study, AQ items were selected from the Fear/Dangerousness (higher score = more fearful), Help/Interact (higher score = less likely to help), and Responsibility (higher score = more responsible for current situation) subscales (Brown, 2008). The 19 items that make up these factors were of particular interest because they were easily adaptable to fit both the mental health and CJ conditions. As in Batastini et al. (2014), item responses were based on the study-specific vignettes (see example below) rather than the original AQ vignette. Participants were asked to answer these items on a scale of 1 (not at all) to 9 (very much). In the present study, the Cronbach’s alpha for the AQ total score was good (α = .83).

The Social Distance Scale (SDS)

The SDS (Link, Cullen, Frank, & Wozniak, 1987; Martin, Pescosolido, & Tuch, 2000) is a six-item assessment of respondents’ willingness to engage with an individual who has a known MI. Again, SDS items were modified for this study to match the vignettes presented within each condition. Items were rated on a 4-point scale from definitely (1) to definitely not (4), in which higher total scores indicate more desired social distance from the hypothetical job applicant. The following is a sample item: “How willing would you be to move next door to Joe?” The Cronbach’s alpha for the SDS total score in the present study was strong (α = .91).

The General Perceived Dangerousness Scale (GPDS)

For this study, the GPDS (Alexander & Link, 2003; Link et al., 1987) assessed the extent to which participants’ perceived the hypothetical job candidate to be a threat or danger. The GPDS contains six-items with responses ranging from definitely true (4) to definitely false (1). Higher total values indicate greater perceived dangerousness toward the applicant. Again, item language was altered whenever necessary to accommodate the background of the applicant. In the current study, the GPDS produces good internal consistency (α= .82). Together, the AQ, SDS, and GPDS were used to measure participants’ general social stigma toward the applicant.

The Level of Familiarity Scale (LoF)

The LoF (also referred to as the Level of Contact Report; Corrigan, Edwards, Green, Diwan, & Penn, 2001; Corrigan, Green, Lundin, Kubiak, & Penn, 2001; Holmes, Corrigan, Williams, Canar, & Kubiak, 1999; Corrigan, 2008) includes 11 yes/no items regarding various levels of personal contact with a person who has a known MI and/or criminal record. Items that are affirmatively endorsed are ranked according to degree of intimacy, with lower values given to more superficial experiences and higher values given to deeper level experiences. In this study, adding up all endorsed item numbers generated a total score for the LoF that ranged from 0 and 66 (e.g., if Items 3 and 7 were endorsed as “yes,” the total score would be 10), where higher scores indicated greater familiarity with the respective mental health and/or criminal background of the job applicant. For the combined conditions, participants were asked to complete separate versions of the LoF for MI and CJ populations (i.e., LoF_MI and LoF_CJ, respectively).

Procedure

This research was approved by the Texas Tech University Human Research Protection Program. Participants were recruited from two sources: (a) randomly selected affiliate chapters of the Society for Human Resource Management (SHRM; n = 98), which is the premier national organization for human resource professionals and (b) Qualtrics Panel Management (QPM; n = 151), which is a web-based survey recruitment service that attracts voluntary participants globally. Selected SHRM chapters were contacted by email or phone to request that the survey be distributed to their membership. Chapters that declined to do so were removed from any further recruitment efforts. QPM was selected as a secondary recruitment source because of its ease and accessibility in targeting specialized respondent populations such as those with a human resource background.

All participants were first provided with information that briefly explained the purpose of the study, the voluntary nature of participation, and the amount of compensation they would receive upon completion of the survey. SHRM participants were given the opportunity to enter a drawing to win one of several iTunes gift cards or an Apple iPad Mini. QPM participants received approximately US$3.00 or the monetary equivalent in redeemable gifts cards or prize points offered through Qualtrics. This payment amount was determined by Qualtrics personnel who were responsible for managing project recruitment. Participant contact information was maintained separately from their survey responses. Consenting participants were provided with a hyperlink that directed them to a 20- to 30-min online survey via Qualtrics. After clicking the link, participants were randomly assigned to 1 of 8 possible conditions.

All participants were then instructed to read 1 of 4 case vignettes describing a hypothetical job candidate who was either an individual with a history of MI, an individual with prior CJ involvement, an individual with both a history of MI and CJ involvement, or an individual with neither a MI nor a CJ history (i.e., the healthy control condition). The vignettes provided in each condition were equivalent (including the applicant’s age, gender, crime type, and interview behaviors) except for the mental health or CJ status of the applicant. See Batastini et al. (2014) for a sample vignette.

Half of the participants in each of these four status conditions also received a PE component that they read prior to the assigned case vignette. Participants who received the PE component were exposed to an excerpt citing academic publications that highlighted the benefits of employment; however, unlike the original Batastini et al. (2014) study, which only focused on advantages to the employee and the community in general, this study also emphasized known advantages to the employer or hiring agency. As an example, below is the excerpt that was provided to participants before being asked to read about the applicant with a CJ background:

There are several advantages to employers for hiring ex-offenders. First, most ex-offenders have completed vocational and occupational training programs that are designed to help them develop the necessary skills for succeeding in the world of work. They may have special qualifications to perform jobs that are difficult to fill. Second, the government offers incentives to companies who hire ex-offenders (e.g., through the Federal Bonding Program from the U.S. Department of Labor; Work Opportunity Tax Credit). Additionally, because ex-offenders often know their options for employment are limited, they may be more likely to work harder to prove themselves (Federal Bureau of Prisons, n.d.; Hawkins, 2011).

An ex-offender’s ability to obtain and maintain employment is also recognized as being very important to successful community reintegration and prevention of reoffending. Benefits associated with employment include increased self-esteem and confidence, less crime, better public safety, reduced government costs related to corrections, and improved attitudes toward community-released offenders. (Graffam et al., 2008)

For comparison, the PE component that was paired with the healthy control condition read as follows:

Individuals who engage in work that is personally enjoyable or fits their interests are more likely to be committed to or invested in their employers. Employee commitment is associated with longer job tenure and increased productivity. Furthermore, matching an employee’s abilities with specific job requirements reduces job turnover and enhances task performance. (Schacter, Gilbert, & Wegner, 2011)

Employment can also help individuals develop a sense of self-identity and self-worth, improve their social skills, increase productivity in everyday life, and foster intellectual growth. Work can also provide pleasure, satisfaction, and emotional stability. (Hoare & Machin, 2006)

Following the presentation of case vignettes (and PE, if applicable), participants were then presented with the outcome measures described above.

Results

Data Screening and Preparation

Data collected from the online Qualtrics program were exported directly into SPSS-20. Missing data were imputed using Expectation Maximization (EM; Tabachnick & Fidell, 2013). First, Little’s Missing Completely at Random Test (MCAR) was conducted for each condition. The results of these analyses indicated that data were missing completely at random within each condition. Furthermore, the percentage of missing values within each condition ranged from 0.04% to 1.10%. Because the data were missing completely at random, EM procedures were used to impute missing data values for each condition separately. In short, EM replaces missing values according to what might be expected based on the patterns of the data for other nonmissing variables in the data file (Tabachnick & Fidell, 2013). The scores in the completed data file (i.e., the data file with no missing data points) were used to calculate the scores for the measures used in the current study.

Preliminary Analyses

Prior to examining the outcomes of interest, we examined differences in demographic variables between conditions using chi-square (to assess the association between two categorical variables) and ANOVA (to compare mean differences between groups with a single continuous outcome variable). Specifically, results suggest that there were no statistically significant between-group differences regarding gender (p = .415), education level (p = .594), current employment status (i.e., full-time, part-time; p = 575), years of experience working as a human resource professional (p = .214), SHRM membership (p = .548), and age (p = .267). Furthermore, an ANOVA was used to assess between-group differences regarding LoF scores. There were no statistically significant differences between groups for participants who read about an applicant with CJ involvement; however, there were significant differences in LoF scores among participants who read about an applicant with MI, F(3, 123) = 4.81, p = .003, such that the MI condition yielded lower LoF scores (M = 16.31, SD = 2.76) than the MI only-PE (M = 18.23, SD = 2.06) and the MI and CJ-involved (MI + CJ; M = 18.09, SD = 2.01) conditions. Taken together, these results suggest that participants who read about an applicant with CJ involvement did not differ in their level of familiarity with these individuals, but participants who read about an applicant with MI did differ with respect to how familiar they were with this population. Despite this, LoF_MI scores were not adjusted in the subsequent analyses because not all participants completed the LoF_MI. That is, some participants completed the LoF_CJ, the LoF_MI, or no LoF depending on their assigned condition (see Table 1). If LoF_MI scores were adjusted, SPSS would have excluded participants without LoF_MI scores in the analyses, thus making analyses of the hypotheses impossible.

Table 1:

Means and Standard Deviations of Likelihood of Hire Ratings, Generic Work Behavior Scale Total Scores by Study Condition

Scale	Condition	M	SD
Likelihood of Hire Ratings	MI + CJ	3.59*	1.61
	MI + CJ/PE	4.53	1.36
	MI	4.75	1.50
	MI/PE	4.94	1.32
	CJ	4.24	1.66
	CJ/PE	4.79	1.29
	Healthy	4.97	1.47
	Healthy/PE	5.21	1.25
GWBQ Total Scores	MI + CJ	3.42*	0.45
	MI + CJ/PE	3.52	0.62
	MI	3.81	0.58
	MI/PE	3.78	0.67
	CJ	3.95	0.59
	CJ/PE	3.58	0.49
	Healthy	3.90	0.70
	Healthy/PE	3.89	0.61

Note. Likelihood of Hire ratings range from 1 (not at all likely) to 7 (very likely). Total scores on the GWBQ range from 1 (poor performance) to 5 (good performance). MI = mental illness; CJ = criminal justice; PE = psychoeducation; GWBQ = Generic Work Behavior Questionnaire.

Statistically significant at p < .05 (for Likelihood of Hire, MI + CJ differed from all conditions except for CJ-only and MI + CJ-PE; for GWBQ, MI + CJ differed from CJ-only and healthy control conditions).

Employability Outcomes

Employability was assessed using the two researcher-generated items (i.e., acceptability of the applicant for the job and likelihood of hire), and the GWBQ (i.e., expected applicant performance). First, chi-square analyses were used to assess acceptability of the applicant across all study conditions. Results indicated a statistically significant relationship between condition and acceptability rating, χ²(3, N = 249) = 11.23, p = .011. Furthermore, applicant acceptability was significantly different between participants who received PE and those who did not, χ²(1, N = 249) = 5.20, p = .023. Additional post hoc chi-square analyses were conducted to examine specific differences in acceptability ratings across the eight conditions. The results indicated that participants in the MI + CJ condition without PE were perceived as significantly less acceptable for the job compared with all other conditions (CJ-PE, p = .005; Healthy, p = .010; Healthy-PE, p = .003; MI, p = .003; MI-PE, p = .026) except the CJ-only condition (p = .236). That is, 31.2% of participants rated the MI + CJ applicant as unacceptable when compared with a range of 0% to 18.9% of participants in other conditions who provided the same response. In addition, it is noteworthy that the MI + CJ applicant was significantly (p = .040) more likely to be considered acceptable for the job if participants were first exposed to the PE material (i.e., only 10% of MI + CJ-PE participants rated the applicant as “unacceptable”), suggesting that PE improved acceptability outcomes when the hypothetical job candidate had a combined history of MI and CJ involvement. Last, the CJ-only condition was the second highest condition to be rated as least acceptable (18.9%) when compared with the Healthy-PE (0%; p = .024) and MI-only (3.1%; p = .041) conditions. There were no other statistically significant acceptability ratings.

Second, a MANOVA (to compare mean differences between groups with multiple continuous outcome variables) was used to examine differences across conditions on the Likelihood of Hire item and GWBQ total score. The overall test was significant, Wilks’s Λ(14, 480) = .83, η²_p = .09, p < .001. It should be noted that PE appeared to increase means across conditions (suggesting more positive ratings); although, differences did not always reach statistical significance. Tukey post hoc analyses revealed that the MI + CJ applicant had the lowest mean (M = 3.59) ratings on Likelihood of Hire, which was significantly lower than CJ-PE (M = 4.79, SD = 1.29, p = .031), MI only (M = 4.75, SD = 1.50, p = .034), MI only-PE (M = 4.94, SD = 1.32, p = .006), Healthy (M = 4.97, SD = 1.47, p = .005), and Healthy-PE (M = 5.21, SD = 1.25, p = .001). Therefore, the hypothesis was partially supported in that the MI + CJ condition produced the lowest Likelihood of Hire ratings than most groups, and the Healthy group produced significantly higher ratings than MI + CJ condition. However, an effect of PE was not observed for this outcome. For the GWBQ, Tukey post hoc analyses revealed that the MI + CJ condition evidenced lower mean scores (M = 3.42, SD = 0.45) on expected applicant performance than the CJ-only (M = 3.95, SD = 0.59, p = .006) and the Healthy (M = 3.90, SD = 0.70, p = .032) conditions. Similarly, there did not appear to be an effect of PE on perceptions of anticipated work performance. See Table 1 for a summary of means and standard deviations.

Stigmatization Outcomes

To examine between-group differences on stigmatizing attitudes and beliefs, as measured by the GPDS, the SDS, and the AQ total score, a MANOVA was conducted. The overall test was significant, Wilks’s Λ(21, 686.83) = .87, η²_p = .05, p = .039. Tukey post hoc analyses failed to reveal significant differences between the conditions on mean GPDS scores. However, mean SDS scores were significantly higher for the MI + CJ condition (M = 15.50, SD = 4.30) than the MI-only condition (M = 11.94, SD = 4.40, p = .020); however, no other comparisons were statistically significant. This result indicates that, regardless of PE, participants preferred more social distance between themselves and applicants with both a MI and prior CJ involvement than applicants with only a MI. Moreover, there were no statistically significant differences across the eight conditions on the AQ total score.

Despite the lack of significant findings regarding the AQ total score, another MANOVA was used to further examine potential differences between conditions on the subscales of the AQ (i.e., Fear, Help, Responsibility, Forced Treatment, and Empathy). The overall test was significant, Wilks’s Λ(35, 999.40) = .58, η²_p = .11, p < .001. Tukey post hoc analyses revealed that there were no statistically significant differences across conditions on the Fear, Help, Forced Treatment, and Empathy subscales. There were, however, significant differences between the conditions regarding mean Responsibility scores. Specifically, the CJ-only condition evidenced significantly higher mean Responsibility scores (M = 21.97, SD = 4.72) than the MI-only (M = 12.59, SD = 4.66, p < .001), MI only-PE (M = 11.58, SD = 4.30, p < .001), MI + CJ (M = 14.78, SD = 4.50, p < .001), MI + CJ-PE (M = 17.07, SD = 4.28, p = .001), Healthy (M = 17.19, SD = 6.18, p = .002), and Healthy-PE (M = 15.17, SD = 4.92, p < .001). Overall, these results suggest that participants found CJ-only applicants more responsible for their current life situation (e.g., struggling due to a history of CJ involvement). In addition, mean trends (though some did not reach statistical significance) suggest that participants found MI applicants (with and without PE) least responsible for their current situation in life. For total score means and standard deviations, see Table 2.

Table 2:

Means and Standard Deviations of Social Distance Scale, General Perceived Dangerousness Scale Total Scores, and Attribution Questionnaire Total Score and Subscale Scores

Scale	Condition	M	SD
Social Distance Scale	MI + CJ	15.50*	4.30
	MI + CJ/PE	13.27	4.70
	MI	11.94	4.40
	MI/PE	12.42	4.52
	CJ	13.43	3.65
	CJ/PE	12.79	3.45
	Healthy	12.81	5.14
	Healthy/PE	12.70	3.12
General Perceived Dangerousness Scale	MI + CJ	15.67	3.30
	MI + CJ/PE	16.60	3.67
	MI	17.28	3.80
	MI/PE	17.33	3.42
	CJ	16.84	3.24
	CJ/PE	17.34	3.36
	Healthy	17.69	4.12
	Healthy/PE	17.58	4.10
Fear/Dangerousness (AQ)	MI + CJ	22.47	11.39
	MI + CJ/PE	23.93	15.20
	MI	18.59	12.43
	MI/PE	17.56	13.03
	CJ	16.77	10.69
	CJ/PE	18.78	11.80
	Healthy	19.50	16.28
	Healthy/PE	16.55	10.71
Responsibility (AQ)	MI + CJ	14.78	4.50
	MI + CJ/PE	17.07	4.28
	MI	12.59	4.66
	MI/PE	11.58	4.30
	CJ	21.97*	4.72
	CJ/PE	19.07	5.22
	Healthy	17.19	6.18
	Healthy/PE	15.17	4.92
Help/Interact (AQ)	MI + CJ	29.50	7.26
	MI + CJ/PE	27.10	7.98
	MI	24.94	6.62
	MI/PE	26.12	7.83
	CJ	26.16	6.53
	CJ/PE	25.19	6.43
	Healthy	24.75	7.79
	Healthy/PE	25.13	7.16
Forced Treatment (AQ)	MI + CJ	10.49	6.11
	MI + CJ/PE	13.60	9.59
	MI	9.22	7.45
	MI/PE	8.70	7.32
	CJ	9.73	6.37
	CJ/PE	9.97	7.59
	Healthy	10.66	9.55
	Healthy/PE	9.67	6.46
Empathy (AQ)	MI + CJ	16.25	5.28
	MI + CJ/PE	15.47	5.72
	MI	15.72	5.39
	MI/PE	17.61	5.67
	CJ	18.06	5.95
	CJ/PE	16.52	4.84
	Healthy	18.16	5.85
	Healthy/PE	17.86	4.07
AQ Total Score	MI + CJ	93.49	21.72
	MI + CJ/PE	97.17	26.38
	MI	81.06	21.73
	MI/PE	81.56	23.77
	CJ	92.69	18.17
	CJ/PE	89.53	21.04
	Healthy	90.37	25.21
	Healthy/PE	84.37	18.92

Note. Total scores on the Social Distance Scale range from 1 (definitely) to 4 (definitely not). Total scores on the General Perceived Dangerousness Scale range from 1 (definitely false) to 4 (definitely true). The Fear/Dangerousness (AQ) and Help/Interact (AQ) scores range from 7 to 63 with higher scores indicating greater fear toward the applicant and less willingness to help the applicant, respectively. The Responsibility and Empathy (AQ) scores range from 3 to 27 with higher scores indicating greater personal responsibility and empathy attributed to the applicant’s current problems, respectively. The Forced Treatment (AQ) score ranges from 4 to 28 with higher scores indicating greater belief that the applicant should be forced to engage in treatment. In general, AQ total score ranges from 26 to 182 with higher scores indicating more stigmatizing attitudes and beliefs toward the applicant. MI = mental illness; CJ = criminal justice; PE = psychoeducation; AQ = Attribution Questionnaire.

Statistically significant at p < .05 (for Responsibility [AQ], CJ-only was significantly higher than all other conditions except CJ-only-PE; for the Social Distance Scale, the MI + CJ was significantly higher than MI-only).

Impact of Familiarity

A series of linear regressions (to examine the association between one or more predictor variables and a continuous outcome variable) were used to test the association between LoF scores and employability and stigmatization outcomes. For all analyses, the LoF score served as the predictor variable with each of the above outcome measures serving as criterion variables (i.e., Likelihood of Hire, GWBQ, PDS, SDS, and AQ total scores). LoF_MI scores were only a statistically significant predictor of social distance (b = .57, p = .001, R² = .08) and Likelihood of Hire (b = −.18, p = .002, R² = .07). These findings suggest that the more exposure participants had to persons with a MI, the more distance they wanted from hypothetical job applicants with an MI history, and the less likely they were to seriously consider these applicants for hire. LoF_CJ scores were not predictive of likelihood of hire, GWBQ total scores, or any of the measures assessing stigmatizing attitudes. In addition, binary logistic regression (to examine the association between one or more predictor variables and a dichotomous outcome variables) was used to examine the association between LoF scores and ratings of candidate acceptability for the job (a dichotomous variable; “yes” was coded as 1 and “no” was coded as 2). Results showed that participants with higher LoF_MI (OR = 1.54, p = .002; Cox & Snell R² = .09; Nagelkerke R² = .16) and higher LoF_CJ (OR = 1.54, p = .002; Cox & Snell R² = .03; Nagelkerke R² = .06) scores were more likely to say that the applicant was an unacceptable candidate. Means and standard deviations are reported in Table 3.

Table 3:

Level of Familiarity Total Scores by Study Condition

Scale	Condition	M	SD
LoF_CJ	MI + CJ	17.41	2.11
	MI + CJ/PE	16.40	2.19
	MI	—	—
	MI/PE	—	—
	CJ	16.52	1.99
	CJ/PE	17.08	2.29
	Healthy	—	—
	Healthy/PE	—	—
LoF_MI	MI + CJ	18.09	2.01
	MI + CJ/PE	17.40	2.16
	MI	16.31*	2.76
	MI/PE	18.23	2.06
	CJ	—	—
	CJ/PE	—	—
	Healthy	—	—
	Healthy/PE	—	—

Note. Level of Familiarity (LoF) scores range from 11 to 22 for both MI and CJ forms. CJ = criminal justice; MI = mental illness; PE = psychoeducation.

p < .05.

Last, we examined whether prior experience hiring an applicant with MI or CJ was specifically associated with perceptions of employability. To do this, a dichotomous (i.e., yes/no) experimenter-generated question (“Have you ever hired someone who you knew had a criminal history/MI?”) from the demographics form was used to predict ratings of acceptability, likelihood of hire, and GWBQ scores. Results showed that experience hiring someone with a known psychiatric history was not significantly associated with how acceptable participants viewed the hypothetical job applicant, χ²(1, N = 176) = 1.88, p = .171. However, participants who reported previously hiring someone with a criminal record were more likely to hire the hypothetical job applicant, χ²(1, N = 150) = 4.32, p = .038. A MANCOVA was conducted to examine the differences between expected work performance (GWBQ total score) and participants’ likelihood of hire rating for the hypothetical job applicant, depending on the participant’s reported experience hiring employees with a history of MI or CJ involvement. Years of experience working in a human resource profession and level of familiarity were adjusted for these analyses to better isolate the impact of past hiring on current employment decisions. The results indicated that ratings of the hypothetical job applicant’s work performance and likelihood of hire did not significantly differ based on previous experience hiring an individual with known mental health problems, Wilks’s Λ(2, 114) = .98, η²_p = .03, p = .238, or a criminal record, Wilks’s Λ(2, 100) = .96, η²_p = .04, p = .153.

Discussion

It is well established that the ability to secure and maintain employment is integral to the successful reentry for PMI and CJ populations (see Andrews & Bonta, 2010; Dunn et al., 2008; Rakis, 2005; Stuart, 2006). Although extant existing programming within the mental health and CJ systems helps to educate and prepare these individuals for the job market (Guerrero, 2011; Hoffmann et al., 2014), the decision to hire ultimately depends on the perceptions of employers. Yet, research suggests that those in hiring positions often view individuals with a MI or prior CJ involvement as less favorable job candidates, even when they possess the required skills and qualifications (e.g., Batastini et al., 2014; Graffam et al., 2008; Holzer et al., 2002). Fortunately, inaccurate perceptions can be mitigated with the use of training efforts and experiential exposure to stigmatized groups (Alexander & Link, 2003; Batastini et al., 2014; Corrigan & Shapiro, 2010).

Regarding ratings of job acceptability, results were consistent with Batastini et al. (2014) in that the applicant with both a MI and CJ background was rated as the least acceptable candidate compared with the other applicant types (notably, this difference did not reach statistical significance when compared with the CJ-only applicant). Interestingly, the presence of PE training was successful in improving acceptability ratings among this condition (i.e., from 68.8% acceptable to 90% acceptable, respectively). CJ-only and MI-only applicants did not appear to benefit in the same way from the addition of employer training. Similarly, the combined MI + CJ applicant was also the least likely to actually be hired; again, it is worth noting that this difference did not reach statistical significance when compared with the CJ-only applicant. However, an effect of the PE training material was not observed for this outcome across any of the study conditions. This result is contrary to the initial Batastini et al. (2014) study, in which training enhanced hiring decisions for the CJ-only applicant such that he was as likely to get the job as the healthy control. Last, expected applicant performance was again lowest for the combined MI + CJ applicant. Interestingly, the CJ-only applicant was estimated to be the best worker, though not statistically superior to the healthy control. In the Batastini et al. (2014) study, the CJ-only and healthy control applicants were also viewed as comparable workers. As with likelihood of hire, the use of training was ineffective in improving perceptions of job performance across all applicant types.

Consistent with the employment outcomes, results of the general stigma outcomes showed that participants tended to prefer more social distance between themselves and the applicant with both a psychiatric and criminal history. This result occurred regardless of training exposure. As in Batastini et al. (2014), participants also perceived the CJ-only applicant as more responsible for his current life difficulties than the other applicant conditions, while the MI applicant (again regardless of training exposure) was generally seen as the least responsible. Regarding experience, greater self-reported exposure to persons with a MI was associated with an increased desire for social distance, as well as lower ratings of job acceptability and likelihood of hire. Greater familiarity with offenders was also associated with lower ratings of job acceptability, but was not associated with any other outcomes of interest. However, it should be noted that even among participants with the highest familiarity scores, the level of intimate exposure with MI or CJ-involved individuals was relatively low (maximum of 22 out of 66 for both groups). As such, it remains possible that judgments about employability and social acceptability were based primarily on superficial exposure (e.g., media portrayals) than on more personalized, real-world experiences with these individuals. Existing literature suggests that the latter is most effective in reducing biased attitudes (Alexander & Link, 2003).

Taken together, it appears that training had a small impact on the most stigmatized applicant; yet, it was not influential enough to increase his chances of ultimately securing the job. That is, although the applicant with both a MI and prior CJ involvement was rated as generally acceptable for the job following brief training, simply being deemed acceptable was insufficient to improve participants’ perceptions about his anticipated work performance and, ultimately, whether or not he would be hired. Findings from this study also suggest that employer perceptions may be largely driven by the labels “mentally ill” and “ex-offender” (or similarly “ex-con,” “felon”) rather than the individual’s actual job-specific skills. For example, even though the CJ-only applicant was seen as the best potential worker, his likelihood of hire rating did not differ from the MI + CJ applicant, who was rated the lowest on both employability domains. Furthermore, all participants were again instructed to assume that the hypothetical applicant was educationally and physically qualified for the position.

Therefore, the fact that an applicant has a known MI and/or a criminal history seems to prime employers to make assumptions about the individual, likely based on extraneous information such as general misperceptions or prior negative experiences. For those with a MI, that assumption appears to be about their inability to successfully perform their required job duties (e.g., psychological distress or poor coping strategies will function as barriers to completing the tasks at hand). This is consistent with prior research suggesting that employers anticipate a lack of social and emotional skills among individuals with MI (Hand & Tryssenaar, 2006). For those with criminal histories, the ability to carry out one’s job responsibilities appears to be less of a concern. Rather, it is likely that these applicants are turned down more so because of anticipated problems stemming from an assumed antisocial personality pattern (e.g., criminals are not trustworthy; once a criminal always a criminal).

Compared with the 2014 study, the PE material was less effective overall in altering biased attitudes. That is, those with actual or anticipated experience hiring employees were not as persuaded by the training component as undergraduate participants, even when it was made more comprehensive by including benefits to both employer and employee, as well as the general public. This finding was particularly true for the CJ-involved applicant, for whom training did not enhance his likelihood of being hired as it did in the initial study. In addition to preexisting assumptions associated with the “offender” label, it is possible that employers have become so accustomed to excluding applicants based on their criminal history that it was unrealistic to expect a relatively brief educational component to effectively reverse such a routine practice. For example, many job applicants are outright eliminated from the applicant pool prior to any in-person contact because they checked “yes” to having a criminal record on the application form. In some cases, employers have a legal obligation to remove ex-offenders from further consideration (Harris & Keller, 2005). However, even large companies have barred applicants with prior felony convictions from applying in the first place (e.g., Domino’s Pizza) or have imposed overly strict background check requirements (e.g., Bank of America, Lowe’s; Rodriquez & Emsellem, 2011).

Although we believe that both studies, as well as the general literature base, continue to support (at least to some degree) the idea that education is a promising method for reducing employer bias, there are several limitations and directions for future research that should be discussed. First, like the original Batastini et al. (2014) study, results are limited to the hypothetical scenario presented to participants. Hiring decisions likely fluctuate depending on the severity of mental health problems and/or the criminal behavior depicted. Similarly, decisions are liable to be context-dependent. The vignette used in these studies implied that the applicant would not only have contact with other apartment residents but may also have access to their housing unit. It makes sense that prospective employers would be apprehensive about hiring a former thief (i.e., Joe was described as stealing a television set from a family member to resell for cash) to work in this environment. In many states, employers can even be held legally responsible through negligent hiring laws for injury (to person or property) as a result of hiring a potentially dangerous or incompetent employee (Kittling, 2010). The fact that participants who reported hiring an employee with a known criminal or psychiatric history did not find the applicant any more employable than participants who never hired an employee with a known criminal or psychiatric history further suggests there may have been something about our particular vignette (either Joe himself or the match between Joe’s characteristics and the job description) that made participants reluctant to hire him.

As detailed in Batastini et al. (2014), other characteristics about the hypothetical applicant could have also influenced participants’ responses. These include, but are not limited to, his youthful age, gender, ethnicity, or noted behaviors during the interview. Even though Joe’s ethnicity was purposefully omitted as an attempt to reduce ethnic/racial bias and better isolate his criminal and/or psychiatric status, it is possible that such biases were not in fact mitigated due to socially reinforced stereotypes concerning who is likely to be involved in the CJ or mental health systems. For example, African American applicants tend to be more affected by their criminal record than their White counterparts (Decker et al., 2015). Results demonstrating different outcomes based on race/ethnicity may suggest the need to include a multicultural training component that brings these biases into awareness. Likewise, female applicants may face different challenges compared with men, including the need for greater flexibility in child care (Shivy et al., 2007), which could deter employers. Therefore, follow-up studies should include vignettes that vary other relevant aspects of applicants’ demographic backgrounds.

Second, although this study used a sample in which the majority (85%) of participants self-reported current or prior experience hiring individuals, few participants (33.7%) reported having specialized training or formal education in human resource management or employee relations. It may be the case that individuals with an explicit educational background in the field would have different perspectives on hiring applicants with a MI or prior CJ involvement. Third, as previously acknowledged, the brief nature of the training material was probably insufficient to reduce negative perceptions about MI and/or CJ applicants, especially among a sample of participants who may routinely reject such applicants early in the hiring process.

Future research should focus on clarifying what specific biases or misgivings employers commonly have about hiring these types of applicants. Training programs can then be developed to distinctly target known concerns by dispelling myths and offering appropriate resources. Related to ex-offenders, training paradigms might also discuss actuarial data on the number of crime-free years it takes for an individual’s risk of reoffending to match that of a nonoffender. For example, Bushway, Nieuwbeerta, and Blokland (2011) found that young, novice offenders are “redeemed” after 10 years; whereas older, first-time offenders had considerably shorter redemption periods (i.e., between 2 and 6 years). Redemption data (including other predictive factors beyond the passage of time, such as age of first offense and number of prior convictions) could help employers make more informed decisions about a potential employee’s level of risk or danger. It is further recommended that employer training programs include the organization as a whole—not just the individual(s) directly in charge of hiring—to address larger societal concerns about hiring and working alongside fellow employees with psychiatric and/or criminal histories (Jones Young & Powell, 2015).

Last, both studies were one-sided in that they focused on employer attitudes in relative isolation to the applicant’s skillset. Barriers to employment for individuals with MI and prior CJ involvement are most likely to be overcome if combined efforts are made to reduce employer (and public) biases while improving the marketability of a population with generally unstable employment histories, less education and training, poorer social and communication skills, and a limited understanding about basic professional behaviors (e.g., appropriate interview conduct, résumé development; Baron & Salzer, 2002; Holzer et al., 2003; Visher et al., 2005). One study (Hoffmann et al., 2014), using a randomized trial to compare traditional vocational programs with a supported employment program, lends some support to this idea. Participants in the supported employment program, which provided enhanced assistance to individuals with severe MI (e.g., seeking appropriate job placements, follow-along support to maintain the job) as well as to their employers (e.g., monthly progress meetings with work supervisors, funding incentives), obtained positions at a higher rate, worked more hours per week, earned higher wages, had longer job tenures, and experienced fewer psychiatric hospitalizations over a 5-year outcome period than their counterparts. These results suggest that personalized interactions with employees and their employers yields more positive experiences and, in turn, more sustainable outcomes.

Perhaps the ability to connect applicants with employers prior to their release from hospitalization or incarceration could also be beneficial in building rapport and humanizing these individuals in the eyes of future employers, thereby reducing false expectations about the type of worker they will be based solely on their status as an individual with a MI or CJ involvement. Other recent civil rights initiatives calling for fair hiring policies (e.g., “Ban the Box” campaign; see Rodriquez & Emsellem, 2011) are likely to further reduce the social stigma associated with psychiatric and criminal records in the hiring process. In addition to increasing the number of employed ex-offenders (e.g., Atkinson & Lockwood, n.d.), Ban the Box laws are also associated with significant reductions in continued offending. D’Alessio, Stolzenberg, and Flexon (2015), for example, showed that ex-offenders released in Hawaii (a state with one of the earliest initiated Ban the Box laws) were 57% less likely to have a repeat felony arrest after the law was implemented. Other state-level reforms include expunging low-level offenses, reducing sealing wait periods (i.e., the amount of time that must pass before a prior crime can be sealed from public record), providing released offenders with certificates of rehabilitation, and curbing negligent hiring laws (i.e., limiting employer liability for hiring people with criminal records; see National Employment Law Project, 2011). Combating the issue of unemployment among discriminated populations by modernizing legal policies and enhancing educational practices (for employees and employers) is likely to have a significant impact at both the individual and community levels.

Footnotes

Batastini was located at Texas Tech University during the time this research was conducted.

Ashley B. Batastini, PhD, is an assistant professor in the Department of Psychology at the University of Southern Mississippi. Her research interests include the implementation and efficacy of correctional mental health treatment and issues related to forensic mental health assessment.

Angelea D. Bolaños, MA, is a doctoral student in the Department of Psychological Sciences at Texas Tech University. Her research interests involve treatment of persons with severe mental illness involved in the criminal justice system, and her professional interests involve issues related to forensic assessment.

Robert D. Morgan, PhD, is the John G. Skelton Jr. regents endowed professor and chair in the Department of Psychological Sciences at Texas Tech University. His research interests are in correctional mental health treatment, effects of incarceration, and forensic mental health assessment.

Sean M. Mitchell, MA, is a clinical psychology doctoral candidate in the Department of Psychological Sciences at Texas Tech University. His primary research interests include interpersonal risk factors for suicide with a focus on offender populations.

References

Alexander

L. A.

Link

B. G.

(2003). The impact of contact on stigmatizing attitudes toward people with mental illness. Journal of Mental Health, 12, 271-289. doi:10.1080/0963823031000118267

Andrews

D. A.

Bonta

(2010). The psychology of criminal conduct (5th ed.). New Providence, NJ: Matthew Bender.

Atkinson

D. V.

Lockwood

(n.d.). The benefits of Ban the Box. The Southern Coalition for Social Justice. Retrieved from http://www.southerncoalition.org/wp-content/uploads/2014/10/BantheBox_WhitePaper-2.pdf

Baron

R. C.

Salzer

M. S.

(2002). Accounting for unemployment among people with mental illness. Behavioral Sciences & the Law, 20, 585-599. doi:10.1002/bsl.513

Batastini

A. B.

Bolaños

A. D.

Morgan

R. D.

(2014). Attitudes toward hiring applicants with mental illness and criminal justice involvement: The impact of education and experience. International Journal of Law and Psychiatry, 37, 524-533. doi:10.1016/j.ijlp.2014.02.025

Bonta

Andrews

(2007). Risk-need-responsivity model for offender assessment and rehabilitation. Ottawa, Ontario: Public Safety Canada.

Brown

S. A.

(2008). Factors and measurement of mental illness stigma: A psychometric examination of the Attribution Questionnaire. Psychiatric Rehabilitation Journal, 32, 89-94. doi:10.2975/32.2.2008.89.94

Bureau of Labor Statistics. (n.d.). Databases, tables, & calculators by subject. Retrieved from http://data.bls.gov/timeseries/LNS14000000

Burke-Miller

J. K.

Cook

J. A.

Grey

D. G.

Razzano

L. A.

Blyler

C. R.

Leff

H. S.

. . . Carey

M. A.

(2006). Demographic characteristics and employment among people with severe mental illness in a multisite study. Community Mental Health Journal, 42, 143-159. doi:10.1007/s10597-005-9017-4

10.

Bushway

Nieuwbeerta

Blokland

(2011). The predictive value of criminal background checks: Do age and criminal history affect time to redemption? Criminology, 49, 27-60. doi:10.1111/j.1745-9125.2010.00217.x

11.

Cook

J. A.

Mulkern

Grey

D. D.

Burke-Miller

Blyler

C. R.

Razzano

L. A.

. . . Steigman

P. A.

(2006). Effects of local unemployment rate on vocational outcomes in a randomized trial of supported employment for individuals with psychiatric disabilities. Journal of Vocational Rehabilitation, 25, 71-84. Retrieved from http://www.psych.uic.edu/eidp/EIDP.JVR2006.pdf

12.

Corrigan

P. W.

(2004). How stigma interferes with mental health care. American Psychologist, 59, 614-625. doi:10.1037/0003-066X.59.7.614

13.

Corrigan

P. W.

(2008). A toolkit for evaluating programs meant to erase the stigma of mental illness (Illinois Institute of Technology). Retrieved from https://www.yumpu.com/en/document/view/8786847/a-toolkit-for-evaluating-programs-meant-to-erase-the-stigma-of-

14.

Corrigan

P. W.

Edwards

A. B.

Green

Diwan

S. L.

Penn

D. L.

(2001). Prejudice, social distance, and familiarity with mental illness. Schizophrenia Bulletin, 27, 219-226.

15.

Corrigan

P. W.

Green

Lundin

Kubiak

M. A.

Penn

D. L.

(2001). Familiarity with and social distance from people who have serious mental illness. Psychiatric Services, 52, 953-958. doi:10.1176/appi.ps.52.7.953

16.

Corrigan

P. W.

Markowitz

F. E.

Watson

Rowan

Kubiak

M. A.

(2003). An attribution model of public discrimination towards persons with mental illness. Journal of Health and Social Behavior, 44, 162-179. Retrieved from https://www-jstor-org.web.bisu.edu.cn/discover/10.2307/1519806

17.

Corrigan

P. W.

Shapiro

J. R.

(2010). Measuring the impact of programs that challenge the public stigma of mental illness. Clinical Psychology Review, 30, 907-922. doi:10.1016/j.cpr.2010.06.004

18.

Corrigan

P. W.

Watson

A. C.

(2002). Understanding the impact of stigma on people with mental illness. World Psychiatry, 1, 16-20.

19.

D’Alessio

S. J.

Stolzenberg

Flexon

J. L.

(2015). The effect of Hawaii’s Ban the Box law on repeat offending. American Journal of Criminal Justice, 40, 336-352. doi:10.1007/s12103-014-9251-9

20.

Decker

Ortiz

Spohn

Hedberg

(2015). Criminal stigma, race, and ethnicity: The consequences of imprisonment for employment. Journal of Criminal Justice, 43, 108-121. doi:10.1016/j.jcrimjus.2015.02.002

21.

Dunn

E. C.

Wewiorski

N. J.

Rogers

(2008). The meaning and importance of employment to people in recovery from serious mental illness: Results of a qualitative study. Psychiatric Rehabilitation Journal, 32, 59-62. doi:10.2975/32.1.2008.59.62

22.

Duran

Plotkin

Potter

Rosen

(2013). Integrated reentry and employment strategies: Reducing recidivism and promoting job readiness. New York, NY: The Council of State Governments Justice Center.

23.

Federal Bureau of Prisons. (n.d.). Employing former inmates. Washington, DC: Author. Retrieved from https://www.bop.gov/business/employing_former_inmates.jsp

24.

Friedman

(2015, November 17). Just facts: As many Americans have criminal records as college diplomas. Brennan Center for Justice at New York University School of Law. Retrieved from https://www.brennancenter.org/blog/just-facts-many-americans-have-criminal-records-college-diplomas

25.

Graffam

Shinkfield

A. J.

Hardcastle

(2008). The perceived employability of ex-prisoners and offenders. International Journal of Offender Therapy and Comparative Criminology, 52, 673-685. doi:10.1177/0306624X07307783

26.

Graffam

Shinkfield

Lavelle

Hardcastle

(2004). Attitudes of employers, corrective services workers, employment support workers, and prisoners and offenders towards employing ex-prisoners and ex-offenders. Criminology Research Council, 2, 1-65. Retrieved from http://hdl.handle.net/10536/DRO/DU:30010282

27.

Guerrero

(2011). Prison-based educational and vocational training programs. In Gideon

Sung

(Eds.), Rethinking corrections: Rehabilitation, reentry, and reintegration (pp. 193-217). Thousand Oaks, CA: SAGE.

28.

Hand

Tryssenaar

(2006). Small business employers’ views on hiring individuals with mental illness. Psychiatric Rehabilitation Journal, 29, 166-173.

29.

Harer

M. D.

(1994). Recidivism among federal prisoners released in 1987. Journal of Correctional Education, 46, 98-127.

30.

Harris

P. M.

Keller

K. S.

(2005). Ex-offenders need not apply. Journal of Contemporary Criminal Justice, 21, 6-30. doi:10.1177/1043986204271678

31.

Hawkins

(2011, May). Why you should consider hiring ex-convicts. Intuit, Small Business. Retrieved from http://blog.intuit.com/employees/why-you-should-consider-hiring-ex-convicts/

32.

Hayward

Bright

J. A.

(1997). Stigma and mental illness: A review and critique. Journal of Mental Health, 6, 345-354. doi:10.1080/09638239718671

33.

Hirschfield

P. J.

Piquero

A. R.

(2010). Normalization and legitimation: Modeling stigmatizing attitudes toward ex-offenders. Criminology, 48, 27-55. doi:10.1111/j.1745-9125.2010.00179.x

34.

Hoare

N. P

Machin

M. M.

(2006). Maintaining wellbeing during unemployment. Australian Journal of Career Development, 15, 19-27.

35.

Hoffmann

Jäckel

Glauser

Mueser

K. T.

Kupper

(2014). Long-term effectiveness of supported employment: 5-year follow-up of a randomized controlled trial. American Journal of Psychiatry, 171, 1183-1190.

36.

Holmes

E. P.

Corrigan

P. W.

Williams

Canar

Kubiak

(1999). Changing attitudes about schizophrenia. Schizophrenia Bulletin, 25, 447-456.

37.

Holzer

H. J.

Raphael

Stoll

(2002). Will employers hire ex-offenders? Employer preferences, background checks, and their determinants (Report No. 1243-02). Retrieved from https://pdfs.semanticscholar.org/c425/da764a2613c0a14d16c758fcf74bbf89d76d.pdf

38.

Holzer

H. J.

Raphael

Stoll

M. A.

(2003, May, 19-20). Employment barriers facing ex-offenders. Paper presented at the New York University Law School Urban Institute Reentry Roundtable, New York City. Retrieved from http://www.urban.org/sites/default/files/alfresco/publication-pdfs/410855-Employment-Barriers-Facing-Ex-Offenders.PDF

39.

Jones Young

N. C.

Powell

G. N

. (2015). Hiring ex-offenders: A theoretical model. Human Resource Management Review, 25, 298-312. doi:10.1016/j.hrmr.2014.11.001

40.

Kittling

N. M.

(2010, November). Negligent hiring and negligent retention: A state by state analysis. Paper presented at the Annual American Bar Association Section of Labor and Employment Law Conference, Chicago, IL.

41.

Kosyluk

K. A.

Corrigan

P. W.

Landis

R. S.

(2014). Employer stigma as a mediator between past and future hiring behavior. Rehabilitation Counseling Bulletin, 57, 102-108. doi:10.1177/0034355213496284

42.

Link

B. G.

(1982). Mental patient status, work, and income: An examination of the effects of a psychiatric label. American Sociological Review, 47, 202-215.

43.

Link

B. G.

Cullen

F. T.

Frank

Wozniak

J. F.

(1987). The social rejection of former mental patients: Understanding why labels matter. American Journal of Sociology, 92, 1461-1500.

44.

Link

B. G.

Phelan

J. C.

(2001). Conceptualizing stigma. Annual Review of Sociology, 72, 363-385. doi:10.1146/annurev.soc.27.1.363

45.

Martin

J. K.

Pescosolido

B. A.

Tuch

S. A.

(2000). Of fear and loathing: The role of “disturbing behavior,” labels, and causal attributions in shaping public attitudes toward people with mental illness. Journal of Health and Social Behavior, 41, 208-223.

46.

McAlpine

D. D.

Warner

(2002). Barriers to employment among persons with mental illness: A review of the literature. Newark, NJ: Center for Research on the Organization and Financing of Care for the Severely Mentally Ill, Institute for Health, Health Care Policy and Aging Research, Rutgers University.

47.

Michon

H. W. C.

Kroon

van Weeghel

Schene

A. H.

(2004). The Generic Work Behavior Questionnaire (GWBQ): Assessment of core dimensions of generic work behavior of people with severe mental illnesses in vocational rehabilitation. Psychiatric Rehabilitation Journal, 28, 40-47.

48.

National Employment Law Project. (2011). State reforms promoting employment of people with criminal records: 2010-11 legislative round-up. New York, NY: Author. Retrieved from http://www.nelp.org/content/uploads/2015/03/PromotingEmploymentofPeoplewithCriminalRecords.pdf

49.

Overton

S. L.

Medina

S. L.

(2008). The stigma of mental illness. Journal of Counseling & Development, 86, 143-151. doi:10.1002/j.1556-6678.2008.tb00491.x

50.

Pager

(2003). The mark of a criminal record. American Journal of Sociology, 108, 937-975.

51.

Pager

Western

Sugie

(2009). Sequencing disadvantage: Barriers to employment facing young black and white men with criminal records. The Annals of the American Academy of Political and Social Science, 623, 195-199.

52.

Petersilia

(2001). When prisoners return to the communities: Political, economic, and social consequences. Federal Probation, 65, 3-8. Retrieved from http://www.ncjrs.gov/pdffiles1/nij/184253.pdf

53.

Rakis

(2005). Improving the employment rates of ex-prisoners under parole. Federal Probation, 69, 7-12.

54.

Rauma

Berk

R. A.

(1987). Remuneration and recidivism: The long-term impact of unemployment compensation on ex-offenders. Journal of Quantitative Criminology, 3, 3-27.

55.

Rodriquez

Emsellem

(2011, March). 65 million “need not apply”: The case for reforming criminal background checks for employment. New York, NY: The National Employment Law Project. Retrieved from http://www.nelp.org/content/uploads/2015/03/65_Million_Need_Not_Apply.pdf

56.

Rüsch

Angermeyer

M. C.

Corrigan

P. W.

(2005). Mental illness stigma: Concepts, consequences, and initiatives to reduce stigma. European Psychiatry, 20, 529-539. doi:10.1016/j.eurpsy.2005.04.004

57.

Schacter

D. L.

Gilbert

D. T.

Wegner

D. M.

(2011). Psychology (2nd ed.). New York, NY: Worth.

58.

Shivy

Moon

Mann

Holland

Eacho

(2007). Ex-offenders reentering the workforce. Journal of Counseling Psychology, 54, 466-473. doi:10.1037/0022-0167.54.4.466

59.

Stuart

(2006). Mental illness and employment discrimination. Current Opinion in Psychiatry, 19, 522-526. doi:10.1097/01.yco.0000238482.27270.5d

60.

Tabachnick

B. G.

Fidell

L. S.

(2013). Using multivariate statistics (6th ed.). Boston, MA: Pearson Education.

61.

U.S. Department of Justice. (n.d.). Prisoners and prison re-entry. Washington, DC: Author. Retrieved from https://www.justice.gov/archive/fbci/progmenu_reentry.html

62.

Varghese

F. P.

Hardin

E. E.

Bauer

R. L.

Morgan

R. D.

(2009). Attitudes toward hiring offenders: The roles of criminal history, job qualifications, and race. International Journal of Offender Therapy and Comparative Criminology, 54, 1-14. doi:10.1177/0306624X09344960

63.

Visher

C. A.

Debus-Sherrill

S. A.

Yahner

(2011). Employment after prison: A longitudinal study of former prisoners. Justice Quarterly, 28, 698-718. doi:10.1080/07418825.2010.535553

64.

Visher

C. A.

Winterfield

Coggeshall

M. B.

(2005). Ex-offender employment programs and recidivism: A meta-analysis. Journal of Experimental Criminology, 1, 295-315. doi:10.1007/s11292-005-8127-x