Criminal History,Race,and Housing Type: An Experimental Audit of Housing Outcomes

Abstract

Although there are a number of experimental studies showing that private housing can be difficult to secure for those with criminal history, many issues in this area remain unexplored or underexplored. The goal of the current study was to address the following unexplored or underexplored issues: (a) the impact of various types of multiple conviction records on private housing outcomes, including one that possessed a certificate of relief; (b) racial differences in private housing outcomes; (c) the impact of housing type on private housing outcomes; and (d) the impact of a criminal record, race, and housing type interaction on private housing outcomes. This goal was achieved with the use of a field experiment (correspondence audit). Results showed several statistically and substantively significant differences among the criminal record, race, and housing type conditions. These results can be used to better inform individuals with criminal history who are seeking private housing options.

Keywords

certificates of relief collateral consequences of conviction housing experiment

Housing is a critical factor for a successful reentry (Garland et al., 2010; O’Brien, 2001). Although there are several public housing options (Roman & Travis, 2004), several laws and policies limit public housing access to those with criminal history (Carey, 2005; Legal Action Center, 2016; Roman & Travis, 2004; Schneider, 2018). Because of this fact, many of those with criminal history are forced to seek private housing. Although there are a growing number of experimental studies showing that private housing can be difficult to secure for those with criminal history in the United States (Evans et al., 2019; Evans & Porter, 2015),¹ many issues in this area remain unexplored or underexplored with experimental designs. The goal of the current study was to address the following unexplored or underexplored issues with the use of an experimental design: (a) the impact of various types of multiple conviction records on private housing outcomes, including one that possessed a certificate of relief²; (b) racial differences in private housing outcomes; (c) the impact of housing type on private housing outcomes; and (d) the impact of a criminal record, race, and housing type interaction on private housing outcomes.

Obtaining Housing With a Criminal Record³

Attaining safe and affordable housing is often one of the most significant challenges for those with criminal history, and an important factor in successful reentry (Lutze et al., 2014; Poulos, 2020). In fact, research has demonstrated that housing instability increases the risk of recidivism (Makarios et al., 2010; Metraux & Culhane, 2004). Despite its importance for a successful reentry, housing is often difficult to obtain for those with criminal history (Helfgott, 1997; Legal Action Center, 2016; O’Brien, 2001; Petersilia, 2003).

For example, while many reentering individuals return to family members’ or friends’ households (Nelson et al., 1999; Western et al., 2015), parole conditions can sometimes prevent a reentering individual from living with others who also possess criminal history (Petersilia, 2003). Community-based correctional or noncorrectional transitional housing options generally have limited availability often due to poor funding and/or public support (Legal Action Center, 2016; Roman & Travis, 2004). Similarly, homeless shelters generally provide only short-term solutions and are often limited by funding issues, and some research suggests that they may be criminogenic (Metraux & Culhane, 2004).

Government-subsidized housing is another option and this form of assistance is generally provided in the form of vouchers and government-subsidized residences (42 U.S.C.A. § 1437f). These benefits are distributed by local public housing authorities (PHAs). However, a number of policies (e.g., the Housing and Urban Development’s One Strike guidelines and the Quality Housing and Work Responsibility Act of 1998) have limited access to subsidized housing for those with criminal records and their families, especially those convicted of drug and violent crimes (Carey, 2005; Van Olphen et al., 2009). Policy examples include lifetime bans for sex offenses and methamphetamine manufacturing; periodic bans for certain drug offenses; evictions for tenants, household members, or guests who engage in criminal activity on or near the premises; and refusal of individuals believed to be abusing drugs (see Cammett, 2015; Carey, 2005; Crowell, 2016; Legal Action Center, 2016; Lundgren et al., 2010; Silva, 2015; Van Olphen et al., 2009, for a full discussion of the history, specifics, and impacts of such policies). Given these policies, it is unsurprising that Keene and colleagues (2018) found that reentering individuals convicted of nonviolent drug offenses faced consistent denials of public housing options.

The final option is private housing. However, numerous factors also limit access to private housing. First, many individuals with criminal history lack the funds, credit history, or work stability necessary to secure private housing (Nagin & Waldfogel, 1998; Petersilia, 2003; Travis, 2005). Even if an individual can afford private housing, some suggest that they will be charged higher rental rates or deposits because of their perceived risk (Clark, 2007). Second, although federal guidelines exist to regulate the blanket exclusion of those possessing a criminal record in the employment context (Green v. Missouri Pacific Railroad Company, 523 F 2d 1290, 1975), such guidelines do not yet exist for private housing (Oyama, 2009). This creates an opportunity for a larger amount of disparate treatment in the private housing market for those possessing a criminal record (Pager & Shepherd, 2008). Finally, due to liability fears, property owners and managers commonly use criminal background checks to screen out tenants with perceived risk (Oyama, 2009; Reosti, 2020).

Given liability and safety concerns, criminal convictions are a common reason for denial (Clark, 2007; Helfgott, 1997). For example, in a survey of 196 property owners and managers in Washington state, Helfgott (1997) found that 67% would ask about an applicant’s criminal record and that 43% would automatically deny a person possessing a criminal record (violent crimes were the least likely to be considered). The main reasons for denial included protection and safety of the community. In a similar survey of 611 property owners and managers in Akron, Ohio, Clark (2007) also found that a majority of landlords were unwilling to rent to those with criminal history (see Walter et al., 2017, for a review of similar studies). Interestingly, Clark’s (2007) study found that a landlord’s decision to lease to a person possessing a criminal record was based upon many factors such as credit, income, employment history, rental history, type of criminal background (felony vs. misdemeanor), and whether or not there was evidence of rehabilitation (see also Furst & Evans, 2017; Leasure, 2019; Weiss, 2016).

Although these prior studies provide context into the struggles of those with criminal records in obtaining housing and the attitudes of landlords in denying housing to those with criminal records, there are several methodological issues with nonexperimental research designs that limit their ability to identify causal effects and may lead to the underestimation of discrimination (Pager & Quillian, 2005). Therefore, experimental research is critical to accurately identify and gauge the impact of various types of criminal history on housing outcomes, and a growing number of studies have used experimental methods to explore this area (Evans et al., 2019; Evans & Porter, 2015; Leasure & Martin, 2017).

For example, some researchers have utilized a phone audit design, which involves data collection through real-world interactions where “testers” pose as individuals seeking housing. In one grouping of this line of research, testers contacted landlords and real estate agents to inquire about securing a residence (Evans, 2016; Evans & Porter, 2015). Testers were randomly assigned a script containing no conviction or a script where they would disclose one of three types of offenses: child molestation, statutory rape, or drug trafficking. Evans and Porter (2015) found that landlords are significantly less willing to consider prospective tenants with a criminal conviction, particularly when the conviction is for child molestation and when the tester is male. For real estate agents, Evans (2016) found that the financial status of the individual and market conditions were also important variables when considering potential clients.

In a recent adaptation of Evans and Porter (2015), Evans and colleagues (2019) again conducted a phone audit with the same randomly assigned criminal record categories, but included an examination of racial (e.g., African American, White, and Latino) and gender differences.⁴ The results indicated statistically significant differences between the no criminal record category and criminal record categories (lower positive response rates for the record categories).⁵ No statistically significant racial or gender differences were found across the testers.

In another phone audit that was conducted in Ohio, Leasure and Martin (2017) found the following positive response rates: 1-year-old felony = 38%, 1-year-old misdemeanor = 84%, 10-year-old felony = 56%, 1-year-old felony with certificate of relief = 70%. This study was important as it indicated that Ohio’s certificate of relief (i.e., the certificate of qualification for employment or CQE), which was created for employment purposes, may be useful for improving housing outcomes for those with criminal history.⁶ Those authors did not explore racial differences but did find that property managers of single-family houses were more likely to rent to those with criminal history than owners or managers of apartments. Importantly, this study was designed to only include property managers who provided a response (in contrast to audit studies that generally note nonresponses as negative).

Criminal History, Private Housing, and Rational Choice/Expected Utility Theories

Rational choice/expected utility theories propose that individual’s behavior is affected by their rational decision-making. Rationality is entrenched in an individual weighing the costs and benefits of a decision, whereby individuals will most likely make decisions that increase benefits while reducing the potential costs (Beccaria, 1963/1764; Becker, 1968; Cornish & Clarke, 2014, and see Pogarsky et al., 2017, for details on the variations of the theories). As it relates to private housing, the tenants of this theoretical paradigm would suggest that property managers/landlords weigh the costs and benefits before approving any applicant (Clark, 2007). Although this theoretical framework has been used in a number of different contexts (Pratt, 2008), no study has utilized this paradigm to explain private housing outcomes for applicants with criminal history.

Given the safety and liability concerns related to renting to individuals with criminal history, rational choice/expected utility theories would predict that individuals with criminal history would have poorer private housing outcomes than equally situated individuals with no record. However, rational choice/expected utility theories would also predict that individuals with less serious and older criminal history should have better private housing outcomes than those with more serious and recent criminal history. Relatedly, rational choice/expected utility theories would also predict that individuals with criminal history and evidence of rehabilitation (i.e., a certificate of relief) would have better private housing outcomes than those with criminal history and no evidence of rehabilitation (see Leasure & Kaminski, 2021).

Current Study

While the above review shows that there are a growing number of studies using experimental methods to explore the impact of criminal history on housing outcomes, it also shows that several areas of this topic are underexplored or unexplored. First, no experimental study has explored the impact of a multiple conviction record on private housing outcomes. This is an important limitation given research showing that a large majority of individuals with criminal history possess previous misdemeanor and felony convictions (e.g., Intake Study, 2016). Relatedly, only one experimental study has explored the impacts of different criminal histories with varying seriousness levels (i.e., felony vs. misdemeanor) and offense ages (Leasure & Martin, 2017). Second, only one experimental study has explored racial differences in private housing outcomes for those with criminal history (Evans et al., 2019). Again, this is a large gap in the literature given previous research in the employment context demonstrating that racial minorities with criminal history face increased amounts of discrimination (Pager, 2003). Third, only one experimental study has explored the impact of housing types on private housing outcomes for those with criminal history (Leasure & Martin, 2017) and no study has examined roommate listings as a housing type. Exploring differences in private housing outcomes across housing types is important as results could be utilized to direct reentering individuals to private housing opportunities that have a higher likelihood of renting to one with criminal history. Finally, no experimental study has utilized a three-way interaction between criminal history, race, and housing type variables to explore impacts on private housing outcomes. The current study seeks to address these gaps. The specific research questions, guided by theory and previous research, were as follows:

Research Question 1 (RQ1): Are there differences in private housing outcomes between those with various types of multiple conviction records (including one with a certificate of relief) and those with no criminal record?

Research Question 2 (RQ2): Are there differences in private housing outcomes between African American and White applicants?

Research Question 3 (RQ3): Are there differences in private housing outcomes between single-family rentals, multiple-family rentals, and shared housing (roommate) rentals?

Research Question 4 (RQ4): Does the use of an interaction between criminal record, race, and housing type affect any of the results from the previous research questions?

Method

To answer the above research questions, a between-subjects correspondence audit was used (see Berk, 2005; Pager & Quillian, 2005; Pager & Western, 2012; Shadish et al., 2002, for literature discussing the benefits of experiments and issues with other designs, especially surveys).⁷ A correspondence audit is a design that sends pieces of correspondence (e.g., emails) to a population of interest (e.g., property managers), which vary only by their treatments (e.g., criminal record statements; Lahey & Beasley, 2018; Vuolo et al., 2018). A between-subjects correspondence audit sends only one piece of correspondence to each individual within the population of interest. We sent each property manager a single email that was randomly assigned both a criminal record condition and a racially distinct name.⁸ Randomization of race and criminal record was done simultaneously and was achieved using random.org, a true randomization mechanism (Haahr, 2019).

Study Context

Ohio had an estimated population of 11,689,100 in 2019. About 82% of those residents were White and approximately 13% were African American (these were the two largest racial groups in Ohio; U.S. Census Bureau, 2019). More than 50% of Ohio residents reside in Butler, Cuyahoga, Franklin, Hamilton, Lorain, Lucas, Mahoning, Montgomery, Summit, and Stark counties. Akron, Cincinnati, Columbus, Cleveland, Dayton, and Toledo all have populations over 100,000 (Urban Extension, 2021). Furthermore, approximately 66% of Ohio households were owner occupied and the median rent amount was US$808 (U.S. Census Bureau, 2019). Approximately one quarter of Ohio’s population resides in rural areas (Center for Family and Demographic Research, 2004) and rural areas in Ohio generally have higher levels of poverty and lower educational levels (Center for Family and Demographic Research, 2004).

In 2015, Ohio had more than 70,000 people incarcerated in Ohio jails and prisons. In addition, roughly 262,000 persons were under some form of community supervision (Kaeble & Glaze, 2016). Furthermore, 80,000 of these individuals were female (Kaeble & Glaze, 2016). Of Ohio’s incarcerated individuals, 60.5% were White, 36.6% were Black or African American, and about 2% were Hispanic (Bennie, 2017). Drug offenses (drug possession being the most common offense) make up the largest single group of Ohio commitments at approximately 28%. Other common crime categories include crimes against persons at 24%, property offenses at 22%, and fraud offenses at 2%. Furthermore, data from the Ohio Department of Rehabilitation and Correction show that a large majority of individuals under current supervision in Ohio possess previous misdemeanor and felony convictions (Intake Study, 2016).

Sample

The sample for the current study was comprised of craigslist.org advertisements in Ohio. Craigslist.org was chosen because the website allows email contact and lists advertisements for both roommate and house/apartment rentals. Furthermore, Craigslist.org and other popular rental websites (apartments.com) contained a significant amount of overlap in listings, reducing the possibility of website differences driving results. Data were collected from May 2020 to August 2020. All regions of Ohio were included and significant quality control steps were taken to ensure that non-Ohio advertisements were excluded.⁹ Housing types included single-family homes, multiple-family apartments, and shared housing (roommate situations). Advertisements older than 14 days were excluded from the sample to limit the possibility that the property was rented before we sent our correspondence. Roommate advertisements seeking characteristics that our hypothetical applicants did not possess (e.g., women only) were also excluded.¹⁰ Research assistants focused on one Ohio region at a time and utilized all eligible advertisements until the region’s advertisements were exhausted or until a predetermined threshold was met.¹¹ Therefore, the data collection procedure closely mirrored a census approach.

Sample size calculations for correspondence studies are largely dependent on the number of responses received for a particular treatment (Vuolo et al., 2016, 2018). In line with the recommendations of Vuolo et al. (2016), several estimates of positive response rates were used for sample size calculations. The estimates were guided by estimates from previous research (primarily Leasure & Martin, 2017, as their study was conducted in Ohio, explored similar research questions, and found several statistically significant differences), by new amendments to the certificate statute, and by the formulation of the criminal record and base resume information used in this study. After several different estimates of positive responses, a conservative sample size was determined to be approximately 1,000 property managers. These calculations were based on a power of .8 to detect statistically significant effects (p ≤ .05). To reach this sample size, emails were submitted multiple days each week for the duration of the study. Early on in the study, quality control procedures found that many advertisements were duplicates, contained spam, or were located in a bordering state. Research assistant error also accounted for a small portion of flawed entries. Therefore, to ensure that we reached our desired sample size after several waves of further quality control, we increased our initial sample size to 2,000. After deleting the entries containing the above flaws, our final sample size was 1,248.

Key Independent Variables

The first key independent variable contained the randomly assigned criminal record conditions that were conveyed in the emails. The five criminal record conditions largely follow Leasure and Martin (2017). We focused on drug crimes to increase generalizability and because many of the above public housing restriction policies apply to those with drug convictions, meaning some may be forced to seek private housing. The current criminal record conditions also note multiple convictions to increase generalizability (see the abovementioned data from the Ohio Department of Rehabilitation and Correction). The criminal record categories were as follows: (a) a condition noting no criminal record, (b) a condition noting 3-year-old felony drug convictions, (c) a condition noting 3-year-old misdemeanor drug convictions, (d) a condition noting 10-year-old felony drug convictions, and (e) a condition noting 3-year-old felony drug convictions and a CQE. No sanctions were discussed in any condition.¹²

The practice of self-disclosure is crucial to the design of the current study and has been utilized by previous correspondence audits (Evans et al., 2019; Zannella et al., 2020). Furthermore, research shows that self-disclosure can help combat the negative effects of criminal record stigma (EmployeeScreenIQ, 2013) and that disclosure can serve as a form of stigma management (Ricciardelli & Mooney, 2018).

The second key independent variable was the randomly assigned racially distinct name. While Evans and colleagues (2019) failed to find statistically significant racial differences in private housing outcomes for those with criminal history, Gaddis and Ghoshal (2020) did find statistically significant racial differences in their email audit study that focused on a sample of individuals seeking roommates (also Gaddis et al., 2020). Therefore, we explore racial differences in this study. White and African American names were used in this study. First names were selected from Gaddis (2017) who examined racial distinctiveness of first names and Bertrand and Mullainathan (2004) who confirmed racial distinctiveness from birth certificate data and an independent field survey. Last name selection was guided by Crabtree and Chykina (2018) who examined geographically robust racially distinct last names and frequency data from the U.S. Census Bureau (2012). Using these sources of racially distinct names, the name selected for the African American applicant was as DaQuan Jefferson and the name selected for the White applicant was Jake Walsh.

The third key independent variable noted the housing type for the rental listing. This variable was observational and not subject to randomization. The first category for this variable was apartment, defined as a multiple-family building. The second category for this variable was house, defined as a single-family building. The third category for this variable was roommate, defined as a shared house or apartment (i.e., renting a room in a house or apartment). This variable was important to include as previous research indicated differences in positive responses for those with criminal history across various housing types (Leasure & Martin, 2017).

Dependent Variable

The dependent variable examined whether a hypothetical applicant received a positive response (dichotomous yes or no). Positive responses included those that stated the property was available, those that invited the applicant to continue in the application process, those that asked for additional information, and those that directed the applicant to other properties because the current property had been rented. Negative responses included those that noted the applicant would not be approved, those that did not respond, and those that stated the property was rented. Responses were measured by monitoring email accounts for 30 days after submission of the resume. While voicemails and emails were monitored for 30 days after resume submission, virtually all responses occurred within 1 week of submission.

Robustness Checks¹³

Control Variables

The first robustness check procedure involved including control variables into each regression model. Control variables retain value in experiments because random assignment only theoretically creates equal groups and can fail to do so in practice (Berk, 2005; Saint-Mont, 2015). In addition, identifying influential control variables could be valuable to future nonexperimental designs or experiments where there are problems with random assignment. The control variables noted the specific region in Ohio where the property was located and the rental amount.¹⁴ When a property listed multiple rent amounts (due to several offerings), we used an average of all of the rent amounts. The specific region in Ohio was important to note as previous research indicated that residents in areas that have more exposure to individuals with criminal history may be more willing to offer opportunities (Hirschfield & Piquero, 2010).

Response Subcategories

The second robustness check procedure included an examination of the subcategories for the dependent variable. The subcategories are noted above in the dependent variable section. Only cross-tabulations of these subcategories are presented given the small sample sizes. Specifically, this section displays the distribution of the response subcategories across criminal record categories, race categories, and housing type categories.

Analytic Strategy

Quantitative Analyses

All quantitative analyses were conducted with the software Stata 16 (StataCorp., 2019). The first stage of data analysis presented descriptive statistics. The second stage utilized bivariate analysis in the form of chi-square tests. The second stage of analysis focused on the first three research questions. The third stage of analysis utilized multiple logistic regression to examine positive responses from a model specified with a three-way interaction between the criminal record, race, and housing type variables (Brambor et al., 2006, noting to include interactions when there is a theoretical basis).¹⁵ The third stage of analysis focused on the fourth research question. The fourth stage of analysis included the robustness checks noted above. The final stage of analysis included the qualitative examination of the additional information contained in manager responses that went beyond the answer of whether the property was available.

Results of regression analyses were presented using Stata’s margins commands.¹⁶ Reported first was predicted probabilities (predicted probability of a positive response). The prime advantages of using predicted probabilities include ease of interpretation and practical significance (see Williams, 2012).¹⁷ Statistically significant differences between these predicted probabilities were then reported using Stata’s average marginal effects (discrete change from base level; see Williams, 2012, discussing advantages of average marginal effects) and contrasts of predictive margins (Mitchel, 2012). While statistical significance (p < .05) was noted throughout the “Results” section, marginally (p < .10) and substantively significant findings were also discussed as recommended by Wasserstein et al. (2019).

Qualitative Analysis

In many cases, property managers included additional information in their responses beyond the answer to our question of whether the property was available. Several of those responses, collectively or individually, were worthy of note. This additional information was analyzed for common themes and interesting individual responses and is presented in the “Results” section (see Furst & Evans, 2017; Leasure, 2019, for a similar approach).

Quantitative Results

Descriptive Statistics

Overall, 255 applicants (20.43%) received a positive response. Section IV of the Supplemental Technical Appendix (available in the online version of this article) presents a table displaying the distribution of the key independent variables and balance tables that show the distribution of key independent variables across each other and across the control variables. Section IV of the Supplemental Technical Appendix also presents descriptive statistics for the control variables.

Bivariate Analyses¹⁸

RQ1: Criminal Record

Positive response percentages for the criminal record categories were as follows: no record = 28.29%, recent felony convictions = 16.87%, recent misdemeanor convictions = 14.64%, old felony convictions = 23.05%, and recent felony convictions plus a CQE = 18.65%. The difference between these groups was statistically significant: χ² = 18.197, df = 4, p = .001. The difference between the no record and recent felony groups was statistically significant: χ² = 9.289, df = 1, p = .002. The difference between the no record and recent misdemeanor groups was statistically significant: χ² = 13.592, df = 1, p < .001. The difference between the no record and old felony groups was not statistically significant: χ² = 1.854, df = 1, p = .173. The difference between the no record and recent felony plus a CQE groups was statistically significant: χ² = 6.589, df = 1, p = .010. The difference between the recent felony and recent misdemeanor groups was not statistically significant: χ² = 0.450, df = 1, p = .502. The difference between the recent felony and old felony groups was marginally significant: χ² = 2.966, df = 1, p = .085. The difference between the recent felony and recent felony plus CQE groups was not statistically significant: χ² = 0.268, df = 1, p = .605. The difference between the recent misdemeanor and old felony groups was statistically significant: χ² = 5.673, df = 1, p = .017. The difference between the recent misdemeanor and recent felony plus CQE groups was not statistically significant: χ² = 1.415, df = 1, p = .234. The difference between the old felony and recent felony plus CQE groups was not statistically significant: χ² = 1.486, df = 1, p = .223.

The above results indicate that property managers may view age of criminal history as a more important factor than offense level (e.g., felony vs. misdemeanor) when determining the suitability of a particular applicant. Similarly, the results indicate that recent misdemeanor convictions can be nearly as impactful as recent felony convictions. Finally, while the difference was not statistically significant, it should be noted that the mean positive response percentage for the recent felony plus a CQE group was approximately two percentage points higher than the recent felony group positive response percentage.

RQ2: Race

Positive response percentages for the racial categories were as follows: African American = 21.11% and White = 19.74%. The difference between these groups was not statistically significant: χ² = 0.360, df = 1, p = .548. These results do not show statistically or substantively significant differences between the mean percentages of a positive response for African American people and White people. In fact, the mean percentage of a positive response was higher for African American individuals.

RQ3: Housing Type

Positive response percentages for housing type were as follows: apartment = 24.07%, house = 17.60%, and roommate = 18.44%. The difference between these groups was statistically significant: χ² = 6.452, df = 2, p = .040. The difference between apartment and house was statistically significant: χ² = 4.216, df = 1, p = .040. The difference between apartment and roommate was statistically significant: χ² = 4.651, df = 1, p = .031. The difference between house and roommate was not statistically significant: χ² = 0.081, df = 1, p = .775. These results indicate that there are statistically significant differences between the apartment and house groups and the apartment and roommate groups.

RQ4: Multiple Regression With the Three-Way Interaction¹⁹

Regarding criminal record type differences, the results showed as follows: a statistically significant difference between the recent felony and no record groups when the applicant was White and applied to an apartment listing; a statistically significant difference between the recent misdemeanor and no record groups when the applicant was African American and applied to an apartment listing; a marginally significant difference between the recent misdemeanor and no record groups when the applicant was African American and applied to a roommate listing; a marginally significant difference between the recent misdemeanor and no record groups when the applicant was White and applied to an apartment listing; a statistically significant difference between the recent felony plus CQE and no record groups when the applicant was White and applied to an apartment listing; a marginally significant difference between the recent felony plus CQE and no record groups when the applicant was White and applied to a roommate listing; a marginally significant difference between the recent misdemeanor and recent felony groups when the applicant was African American and applied to a roommate listing; a statistically significant difference between the old felony and recent felony groups when the applicant was White and applied to an apartment listing; a marginally significant difference between the old felony and recent misdemeanor groups when the applicant was African American and applied to a roommate listing; a marginally significant difference between the recent felony plus CQE and recent misdemeanor groups when the applicant was African American and applied to a roommate listing; and a statistically significant difference between the recent felony plus CQE and old felony groups when the applicant was White and applied to an apartment listing.

Although not marginally or statistically significant, it is worthy of note that African Americans with a recent felony plus CQE who applied to house listings had a positive response rate that was 13 percentage points higher (17.2% vs. 4.2%) than African Americans with a recent felony who also applied to house listings.

Regarding racial differences, the results showed a statistically significant difference between Whites and African Americans with a recent felony plus a CQE who applied to apartment listings. The results also showed a marginally significant difference between Whites and African Americans with a recent felony who applied to apartment listings.

Regarding the housing type differences, the results showed as follows: a marginally significant difference between house and apartment listings when the applicant was an African American with no record; a marginally significant difference between house and apartment listings when the applicant was an African American with a recent felony; a statistically significant difference between roommate and apartment listings when the applicant was an African American with no record; and a marginally significant difference between roommate and apartment listings when the applicant was an African American with a recent misdemeanor.

Regarding criminal record type differences, the results indicate that differences in positive responses by criminal record type are also affected by housing type and race. Regarding racial differences, the results indicate that differences in positive responses by race are also affected by housing type and criminal record type. Regarding housing type differences, the results indicate that differences in positive responses by housing type are also affected by race and criminal record type.

Qualitative Results

First, several property managers were willing to respond that a particular criminal record would result in a denial due to their rental requirements. For example, one manager said, “Unfortunately, those convictions would deny your application once screened.” In other words, several of the managers seemed willing to answer qualification questions without first asking the applicant to apply. This is an important finding because reentering individuals, who generally do not have a surplus of funds (Garland et al., 2010; O’Brien, 2001), may be able to avoid the time and expense associated with a rental application (that could be ultimately denied due to criminal history) by sending the manager an initial email.

Second, several property managers inquired about the applicant’s current drug use, indicating that they would be likely to consider the applicant if the drug use had ceased. For example, one manager stated, “As long as your clean now because I don’t allow drugs of any kind here.” Another manager’s statement produced the same theme, but seemed to differentiate depending upon the type of drug by saying, “No that don’t matter as long as you don’t do them now I don’t want to be around that [but] weed is ok.” Finally, one manager indicated that renting to an individual with substance abuse/criminal history was riskier because of the presence of the potential presence of the applicant’s associates, by stating, “I will have to admit, that your history concerns me. More so a relapse or your friends.”

Third, several property managers asked for additional details about the applicant’s conviction. For example, one manager asked, “What kind of drug charges they were and have you had any other kinds of trouble with the law before or since?” Another manager stated, “We may require a letter of explanation once the background check is completed.” Other managers asked for website links to the conviction records and for the jurisdiction of the convictions.

Fourth, several property managers noted that they appreciated the applicant’s up-front disclosure of the criminal history. For example, manager stated, “Yes the property is still available and thanks for the Being sincere with us.” Another manager stated,

Oh I got your email and thought about it, it’s not going to be a problem giving out my property to someone who’s admitted and being truthful that he’s got a felony rather [than] being exposed during the tenant background check.

Relatedly, one manager stated, “If you were to get denied based off of your criminal history but were honest on the application you would get your initial $99 deposit you would put down back.” The finding of positive effects for disclosure is an important result, given that this approach could easily be implemented by those with criminal history.

Fifth, several property managers noted crime-free time requirements. For example, one manager stated, “According to our criteria it has to be past 84 months for your situation. I am so sorry I cannot help you with your housing needs at this time.” Another manager stated, “I’m sorry but we do not accept applications from anyone with a misdemeanor within the past 5 years.” Several other managers noted various crime-free requirements (e.g., 10 years and 3 years) and many had different requirements depending on the type of crime (e.g., felony vs. misdemeanor). Relatedly, some managers noted that specific crime types would result in denial. For example, one manager stated, “Once you have completed an application for the property you are interested in, we will also do a background check and as long as you have no felonies or misdemeanors involving sex, drugs or violence, you should be fine.” Another manager stated, “Our qualifications state no felonies.”

Sixth, some property managers noted that applicants with criminal history would face continued scrutiny if approved. For example, one manager stated,

I am not against second chances at all, but with that [being] said if you were to rent from us we would watch you closely. Rent would have to be on time. No disturbances to the neighbors. No heavy traffic to the apartment. No “funny” business at all. There would not be any room for mistake.

Another manager stated, “If you are interested in the place, I will ask you more info than others. Mostly to know your support group and your plan to avoid relapsing.” Finally, one manager stated, “We would consider renting to you if you can pass all the other requirements and provide employment, references.”

Seventh, several property managers asked for additional details about the applicant’s financial status. For example, one manager asked, “Where do you work and how long have you been there? What is your monthly net income? Do you have garnishments, or are you making restitution payments? Finally, are you renting now and how much are you paying?” Many of these managers may be implying that criminal history could be overcome if the applicant could present positive financial factors. Relatedly, several managers asked for additional details about the applicant’s age, familial status, and living situation. For example, one manager asked, “Age? Pets? Kids?” Another manager asked, “Would you please tell me a little bit more about yourself? For example, number of people renting together, reason to look for a new place to rent.” Here again, these managers may be implying that they would consider an applicant with criminal history if that applicant could demonstrate stability.

Finally, there were several noteworthy individual responses. For example, one manager noted that their decision to rent to one with criminal history would be dependent on what type of property the applicant was seeking, by stating, “It depends on the property you’re interested in.” Furthermore, one manager specifically asked for copies of the certificate of relief, indicating that the certificate could play a role in improving the likelihood of approval for one with criminal history. The manager stated, “While we do check criminal backgrounds, we do accept felonies. Please include your certificate as an attachment on your online application.”

Discussion

The above results present several areas worthy of further discussion. First, consistent with previous research and rational choice/expected utility theories, the above results show that those with criminal history are less likely to receive a positive response to a rental inquiry. Given these results, it is important to identify mechanisms that can improve the likelihood of a positive response for one with criminal history. Previous research showed that Ohio’s CQE, which was designed to reduce collateral consequences in employment, may be an effective tool to improve private housing outcomes (Leasure & Martin, 2017). However, the multiple regression results show that the effectiveness of Ohio’s CQE for private housing may vary depending upon the race of the individual as well as the type of housing sought. Therefore, it is important to identify mechanisms that are effective across race and housing type.

Second, while the results generally showed that those with criminal history fared poorer in private housing outcomes than those without such history, there was a great deal of variability in outcomes depending on the type of criminal history. One important finding on this issue, which was consistent with the predictions of rational choice/expected utility theories, was that those with the old felony condition were generally more likely to receive a positive response than the other more recent criminal record conditions. Such results may indicate that recency may be a more prominent factor for managers than seriousness, at least for drug crimes. This is an interesting finding given previous research showing that an individual’s risk of reoffending significantly decreases over time (Blumstein & Nakamura, 2009; Kurlychek et al., 2006). Perhaps property managers have begun to recognize such findings. It is also possible that property managers view those with older criminal history as more likely to possess better credit or more stable income sources. Future qualitative research on the relationship among age, financial standing, and criminal history would be beneficial to the housing and reentry fields. Nonetheless, as the qualitative results showed, some managers do not consider individuals with criminal history, regardless of the age of the criminal history. Therefore, efforts should be made to educate property managers on the risk level of those with older criminal history.

Third, the current study did not find pervasive racial differences in private housing outcomes for those with criminal history. In fact, African Americans had a higher probability of a positive response in most of the comparisons. While the lack of statistically significant differences is consistent with Evans and colleagues (2019), other studies in the private housing area have found statistically significant racial differences (Gaddis et al., 2020; Gaddis & Ghoshal, 2020). Nonetheless, very few studies have examined racial differences in private housing outcomes for those with criminal history, meaning that future research on this issue is necessary.

Fourth, the results showed several statistically and substantively significant differences in private housing outcomes depending on the type of housing sought. Although this overall finding is consistent with Leasure and Martin (2017), that study found that individuals with criminal history had a higher likelihood of a positive response from single-family houses. We found that individuals with criminal history had a higher likelihood of a positive response with apartments. Such conflicting results show the need for future research on this issue, as reentry personnel could direct reentering individuals to housing types that may be more likely to rent to one with criminal history.

Finally, the qualitative results showed that simple up-front disclosure of one’s criminal history might be an effective method to improve private housing outcomes for those with criminal history. This is an interesting finding as previous research in the employment context has shown that self-disclosure can also help combat the negative effects of criminal record stigma (EmployeeScreenIQ, 2013) and that disclosure can serve as a form of stigma management (Ricciardelli & Mooney, 2018). Furthermore, the above results showed that many property managers are willing to provide a response regarding their willingness to rent to an individual with criminal history. Individuals with criminal history may be able to save money and time spent on private housing searches by contacting property managers to determine whether their particular criminal history would be a reason for a denial. Service providers should consider these potential benefits of disclosure when advising those with criminal history on private housing searches.

Several limitations contextualize the above findings. First, this study only examined hosing outcomes for men. Second, only drug crimes were used as the criminal record condition. Third, the study was only conducted in Ohio. Fourth, this study did not examine manager characteristics beyond rent amount and city location (i.e., race and gender). One may reasonably expect differences in results if any of the above approaches were altered. Finally, any design that relies upon voluntary disclosure of criminal history may overestimate discrimination as many managers may not have conducted background checks. Future research should address the above limitations and attempt to replicate our findings so that stronger policy can be developed.

Supplemental Material

sj-docx-1-cjb-10.1177_00938548221082086 – Supplemental material for Criminal History, Race, and Housing Type: An Experimental Audit of Housing Outcomes

Supplemental material, sj-docx-1-cjb-10.1177_00938548221082086 for Criminal History, Race, and Housing Type: An Experimental Audit of Housing Outcomes by Peter Leasure, R. Caleb Doyle, Hunter M. Boehme and Gary Zhang in Criminal Justice and Behavior

Footnotes

Authors’ Note:

The authors would like to thank Samantha Weil-Kaspar and Alexander Shannon for their assistance with data collection.

ORCID iD

Peter Leasure

Supplemental Material

Supplemental Technical Appendix is available in the online version of this article at .

Notes

Peter Leasure, JD, PhD, is an assistant professor in the Department of Criminology and Criminal Justice and Community Impact Research Fellow for the Arthur J. Glatfelter Institute for Public Policy at York College of Pennsylvania.

R. Caleb Doyle, BA, is a recent graduate of York College of Pennsylvania and will begin his first year of law school at Boston College in the Fall of 2021.

Hunter M. Boehme, PhD, is an assistant professor and Juvenile Justice Faculty Fellow in the Department of Criminal Justice at North Carolina Central University.

Gary Zhang, PhD, is a technical writer in the Research and Development Division of the SAS Institute.

References

Beccaria

(1963). On crimes and punishments. Bobbs-Merrill. (Original work published 1764)

Becker

G. S.

(1968). Crime and punishment: An economic approach. In Fielding

N. G.

Clarke

Witt

(Eds.), The economic dimensions of crime (pp. 13–68). Palgrave Macmillan. https://doi.org/10.1007/978-1-349-62853-7_2

Bennie

R. C.

(2017). January 2017 census of ODRC institutional population, demographic and offense summary. Bureau of Research and Evaluation, Ohio Department of Rehabilitation and Correction.

Berk

R. A.

(2005). Randomized experiments as the bronze standard. Journal of Experimental Criminology, 1(4), 417–433. https://doi.org/10.1007/s11292-005-3538-2

Bertrand

Mullainathan

(2004). Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination. American Economic Review, 94(4), 991–1013. https://doi.org/10.1257/0002828042002561

Blumstein

Nakamura

(2009). Redemption in the presence of widespread criminal background checks. Criminology, 47(2), 327–359. https://doi.org/10.1111/j.1745-9125.2009.00155.x

Brambor

Clark

W. R.

Golder

(2006). Understanding interaction models: Improving empirical analyses. Political Analysis, 14(1), 63–82. https://doi.org/10.1093/pan/mpi014

Cammett

(2015). Confronting race and collateral consequences in public housing. Seattle University Law Review, 39, 1123–1155.

Carey

(2005). No second chance: People with criminal records denied access to public housing. University of Toledo Law Review, 36(3), 545–594.

10.

Center for Family and Demographic Research. (2004). Ohio population news. Bowling Green State University. https://www.bgsu.edu/content/dam/BGSU/college-of-arts-and-sciences/center-for-family-and-demographic-research/documents/OPN/Ohio-Population-News-Ohios-Rural-Population-August-2004.pdf

11.

Clark

L. M.

(2007). Landlord attitudes toward renting to released offenders. Federal Probation, 71(3), 20–30.

12.

Cornish

D. B.

Clarke

R. V.

(Eds.). (2014). The reasoning criminal: Rational choice perspectives on offending. Transaction Publishers.

13.

Crabtree

Chykina

(2018). Last name selection in audit studies. Sociological Science, 5, 21–28. https://doi.org/10.15195/v5.a2

14.

Crowell

(2016). A home of one’s own: The fight against illegal housing discrimination based on criminal convictions, and those who are still left behind. Texas Law Review, 95, 1103–1143.

15.

EmployeeScreenIQ. (2013). Employment screening practices and trends: The era of heightened care and diligence. http://www.employeescreen.com/Survey_Report_2013.pdf

16.

Evans

D. N.

(2016). The effect of criminal convictions on real estate agent decisions in New York City. Journal of Crime and Justice, 39(3), 363–379. https://doi.org/10.1080/0735648X.2016.1166068

17.

Evans

D. N.

Blount-Hill

K. L.

Cubellis

M. A.

(2019). Examining housing discrimination across race, gender and felony history. Housing Studies, 34(5), 761–778. https://doi.org/10.1080/02673037.2018.1478069

18.

Evans

D. N.

Blount-Hill

K. L.

Hoyos-Torres

(2020). The frontline of housing access: Comparing criminal stigma among landlords and real estate agents in New York. Critical Criminology, 29, 853–871. https://doi.org/10.1007/s10612-020-09534-0

19.

Evans

D. N.

Porter

J. R.

(2015). Criminal history and landlord rental decisions: A New York quasi-experimental study. Journal of Experimental Criminology, 11(1), 21–42. https://doi.org/10.1007/s11292-014-9217-4

20.

Furst

R. T.

Evans

D. N.

(2017). Renting apartments to felons: Variations in real estate agent decisions due to stigma. Deviant Behavior, 38(6), 698–708. https://doi.org/10.1080/01639625.2016.1197635

21.

Gaddis

S. M.

(2017). How Black are Lakisha and Jamal? Racial perceptions from names used in correspondence audit studies. Sociological Science, 4, 469–489. https://doi.org/10.15195/v4.a19

22.

Gaddis

S. M.

DiRago

Ghoshal

(2020). Moving to the city: An audit study examining the consequences of racial/ethnic discrimination on residential segregation. https://ssrn.com/abstract=2605853

23.

Gaddis

S. M.

Ghoshal

(2020). Searching for a roommate: A correspondence audit examining racial/ethnic and immigrant discrimination among millennials. Socius, 6, 1–16. https://doi.org/10.1177/2378023120972287

24.

Garland

Wodahl

E. J.

Mayfield

(2010). Prisoner reentry in a small metropolitan community: Obstacles and policy recommendations. Criminal Justice Policy Review, 22, 90–10. https://doi.org/10.1177/0887403409359804

25.

Haahr

(2019). True random number service. https://www.random.org

26.

Helfgott

(1997). Ex-offender needs versus community opportunity in Seattle, Washington. Federal Probation, 61, 12–24.

27.

Hirschfield

P. J.

Piquero

A. R.

(2010). Normalization and legitimation: Modeling stigmatizing attitudes toward ex-offenders. Criminology, 48(1), 27–55. https://doi.org/10.1111/j.1745-9125.2010.00179.x

28.

Intake Study. (2016). Ohio Department of Rehabilitation and Correction. https://drc.ohio.gov/Portals/0/Reentry/Reports/Intake/2015%20Intake%20Study_final.pdf?ver=2017-09-08-092854-057

29.

Kaeble

Glaze

(2016). Correctional populations in the United States, 2015. Bureau of Justice Statistics.

30.

Keene

D. E.

Smoyer

A. B.

Blankenship

K. M.

(2018). Stigma, housing and identity after prison. The Sociological Review, 66(4), 799–815. https://doi.org/10.1177/0038026118777447

31.

Kurlychek

M. C.

Brame

Bushway

S. D.

(2006). Scarlet letters and recidivism: Does an old criminal record predict future offending? Criminology & Public Policy, 5(3), 483–504. https://doi.org/10.1111/j.1745-9133.2006.00397.x

32.

Lahey

Beasley

(2018). Technical aspects of correspondence studies. In Gaddis

M. S.

(Ed.), Audit studies: Behind the scenes with theory, method, and nuance (pp. 81–101). Springer.

33.

Leasure

(2019). Securing private housing with a criminal record. Journal of Offender Rehabilitation, 58(1), 30–49.

34.

Leasure

Kaminski

(2021). The effectiveness of certificates of relief for those with multiple conviction records. Journal of Empirical Legal Studies, 18(4), 849–875.

35.

Leasure

Martin

(2017). Criminal records and housing: An experimental study. Journal of Experimental Criminology, 13(4), 527–535.

36.

Legal Action Center. (2016). Helping moms, dads, and kids to come home: Eliminating barriers to housing for people with criminal records. https://indd.adobe.com/view/04243d7e-5a9a-4bd8-9d97-1bb1ce77b9c5

37.

Lundgren

L. M.

Curtis

M. A.

Oettinger

(2010). Postincarceration policies for those with criminal drug convictions: A national policy review. Families in Society, 91(1), 31–38. https://doi.org/10.1606/1044-3894.3952

38.

Lutze

F. E.

Rosky

J. W.

Hamilton

Z. K.

(2014). Homelessness and reentry: A multisite outcome evaluation of Washington State’s reentry housing program for high risk offenders. Criminal Justice and Behavior, 41(4), 471–491. https://doi.org/10.1177/0093854813510164

39.

Makarios

Steiner

Travis

L. F.

III . (2010). Examining the predictors of recidivism among men and women released from prison in Ohio. Criminal Justice and Behavior, 37(12), 1377–1391. https://doi.org/10.1177/0093854810382876

40.

Metraux

Culhane

D. P.

(2004). Homeless shelter use and reincarceration following prison release. Criminology & Public Policy, 3(2), 139–160. https://doi.org/.pallas2.tcl.sc.edu/10.1111/j.1745-9133.2004.tb00031.x

41.

Mitchel

M. N.

(2012). Interpreting and visualizing regression models using Stata. Stata Press.

42.

Nagin

Waldfogel

(1998). The effect of conviction on income through the life cycle. International Review of Law and Economics, 18(1), 25–40. https://doi.org/10.1016/S0144-8188(97)00055-0

43.

Nelson

Deess

Allen

(1999). The first month out: Post-incarceration experiences in New York City. Vera Institute of Justice. https://www.vera.org/publications/the-first-month-out-post-incarceration-experiences-in-new-york-city

44.

O’Brien

(2001). “Just like baking a cake”: Women describe the necessary ingredients for successful reentry after incarceration. Families in Society: The Journal of Contemporary Social Services, 82(3), 287–295. https://doi.org/10.1606/1044-3894.200

45.

Oyama

(2009). Do not (re) enter: The rise of criminal background tenant screening as a violation of the Fair Housing Act. Michigan Journal of Race & Law, 15, 181–222.

46.

Pager

(2003). The mark of a criminal record. American Journal of Sociology, 108(5), 937–975. https://doi.org/10.1086/374403

47.

Pager

Quillian

(2005). Walking the talk? What employers say versus what they do. American Sociological Review, 70(3), 355–380. https://doi.org/10.1177/000312240507000301

48.

Pager

Shepherd

(2008). The sociology of discrimination: Racial discrimination in employment, housing, credit and consumer markets. Annual Review of Sociology, 34, 181–209. https://doi.org/10.1146/annurev.soc.33.040406.131740

49.

Pager

Western

(2012). Identifying discrimination at work: The use of field experiments. Journal of Social Issues, 68(2), 221–237. https://doi.org/10.1111/j.1540-4560.2012.01746.x

50.

Petersilia

(2003). When prisoners come home: Parole and prisoner reentry. Oxford University Press.

51.

Pogarsky

Roche

S. P.

Pickett

J. T.

(2017). Heuristics and biases, rational choice, and sanction perceptions. Criminology, 55(1), 85–111. https://doi.org/10.1111/1745-9125.12129

52.

Poulos

(2020). Criminal record based housing discrimination harms public safety. Seattle Journal for Social Justice, 19, 399–407.

53.

Pratt

T. C.

(2008). Rational choice theory, crime control policy, and criminological relevance. Criminology and Public Policy, 7(1), 43–52. https://doi.org/10.1111/j.1745-9133.2008.00489.x

54.

Reosti

(2020). “We go totally subjective”: Discretion, discrimination, and tenant screening in a landlord’s market. Law & Social Inquiry, 45(3), 618–657. https://doi.org/10.1017/lsi.2019.81

55.

Ricciardelli

Mooney

(2018). The decision to disclose: Employment after prison. Journal of Offender Rehabilitation, 57(6), 343–366. https://doi.org/10.1080/10509674.2018.1510866

56.

Roman

C. G.

Travis

(2004). Taking stock: Housing, homelessness, and prisoner reentry. Urban Institute. http://webarchive.urban.org/publications/411096.html

57.

Saint-Mont

(2015). Randomization does not help much, comparability does. PLOS ONE, 10(7), Article e0132102. https://doi.org/10.1371/journal.pone.0132102

58.

Schneider

(2018). The prison to homelessness pipeline: Criminal record checks, race, and disparate impact. Indiana Law Journal, 93, 421–455.

59.

Shadish

W. R.

Cook

T. D.

Campbell

D. T.

(2002). Experimental and quasi-experimental designs for generalized causal inference. Wadsworth Cengage Learning.

60.

Silva

L. R.

(2015). Criminal histories in public housing. Wisconsin Law Review, 2015, 375–397.

61.

StataCorp. (2019). Stata statistical software: Release 16.

62.

Travis

(2005). But they all come back: Facing the challenges of prisoner reentry. Urban Institute Press.

63.

U.S. Census Bureau. (2012). Genealogy data: Frequently occurring surnames from census 2000. https://www.census.gov/topics/population/genealogy/data/2000_surnames.html

64.

U.S. Census Bureau. (2019). QuickFacts: Ohio. https://www.census.gov/quickfacts/fact/table/OH/PST045219

65.

Urban Extension. (2021). Ohio’s urban, suburban, and rural connection. Ohio State University. https://urban-extension.cfaes.ohio-state.edu/about-us/urban-ohio

66.

Van Olphen

Eliason

M. J.

Freudenberg

Barnes

(2009). Nowhere to go: How stigma limits the options of female drug users after release from jail. Substance Abuse Treatment, Prevention, and Policy, 4(1), 1–10. https://doi.org/10.1186/1747-597X-4-10

67.

Vuolo

Uggen

Lageson

(2016). Statistical power in experimental audit studies: Cautions and calculations for matched tests with nominal outcomes. Sociological Methods & Research, 45(2), 260–303. https://doi.org/10.1177/0049124115570066

68.

Vuolo

Uggen

Lageson

(2018). To match or not to match? Statistical and substantive considerations in audit design and analysis. In Gaddis

M. S.

(Ed.), Audit studies: Behind the scenes with theory, method, and nuance (pp. 119–140). Springer.

69.

Walter

R. J.

Viglione

Tillyer

M. S.

(2017). One strike to second chances: Using criminal backgrounds in admission decisions for assisted housing. Housing Policy Debate, 27(5), 734–750. https://doi.org/10.1080/10511482.2017.1309557

70.

Wasserstein

Schirm

Lazar

(2019). Moving to a world beyond “p < 0.05.” The American Statistician, 73, 1–19. https://doi.org/10.1080/00031305.2019.1583913

71.

Weiss

(2016). Housing access for people with criminal records. Advocates’ Guide. https://www.nlihc.org/sites/default/files/AG-2018/Ch06-S06_Criminal-Records_2018.pdf

72.

Western

Braga

A. A.

Davis

Sirois

(2015). Stress and hardship after prison. American Journal of Sociology, 120(5), 1512–1547. https://doi.org/0002-9602/2015/12005-0006$10.00

73.

Williams

(2012). Using the margins command to estimate and interpret adjusted predictions and marginal effects. The Stata Journal, 12(2), 308–331. https://doi.org/10.1177/1536867X1201200209

74.

Zannella

Clow

Rempel

Hamovitch

Hall

(2020). The effects of race and criminal history on landlords’ (un)willingness to rent to exonerees. Law and Human Behavior, 44(4), 300–310. https://doi.org/10.1037/lhb0000419

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.17 MB

Criminal History,Race,and Housing Type: An Experimental Audit of Housing Outcomes

Abstract

Keywords

Obtaining Housing With a Criminal Record 3

Criminal History, Private Housing, and Rational Choice/Expected Utility Theories

Current Study

Method

Study Context

Sample

Key Independent Variables

Dependent Variable

Robustness Checks 13

Control Variables

Response Subcategories

Analytic Strategy

Quantitative Analyses

Qualitative Analysis

Quantitative Results

Descriptive Statistics

Bivariate Analyses 18

RQ1: Criminal Record

RQ2: Race

RQ3: Housing Type

RQ4: Multiple Regression With the Three-Way Interaction 19

Qualitative Results

Discussion

Supplemental Material

sj-docx-1-cjb-10.1177_00938548221082086 – Supplemental material for Criminal History, Race, and Housing Type: An Experimental Audit of Housing Outcomes

Footnotes

Authors’ Note:

ORCID iD

Supplemental Material

Notes

References

Supplementary Material

Obtaining Housing With a Criminal Record³

Robustness Checks¹³

Bivariate Analyses¹⁸

RQ4: Multiple Regression With the Three-Way Interaction¹⁹