Reducing criminal record discrimination through banning the box: the importance of timing and explanation in the reveal of a drug conviction

ABSTRACT

There has been a recent push toward placing restrictions on when and how employers can ask about job applicants’ criminal records. In our research, we asked hypothetical employers to evaluate job applicants so we could examine whether certain ‘ban the box’ practices increase the chances that formerly incarcerated individuals find jobs. Our results showed that an applicant with a drug conviction was more likely to be hired if his record was revealed after an interview rather than on the job application and if he explained the unusualness of his offense rather than if he provided no explanation (Studies 1 and 3). Further, we found that an interviewed applicant who explained his record was similarly likely to be hired regardless of whether he volunteered information about his record or it was discovered through a background check (Study 2). Finally, when equally qualified applicants were being considered at the initial application stage, the applicant without a criminal record was preferred even if the applicant with a record explained it (Study 4). These results indicate the importance of restricting access to criminal record information until after an applicant has been interviewed and in allowing the applicant to explain the unusualness of the offense.

KEYWORDS:

• Ban the box laws
• criminal record
• discrimination
• stereotypes

Open Scholarship

This article has earned the Center for Open Science badges for Open Materials and Preregistered. The materials are openly accessible at https://osf.io/y94n8/ and Study 3: https://aspredicted.org/eh5yk.pdf, Study 4: https://aspredicted.org/pa6mp.pdf.

Acknowledgements

This article was based on research conducted for the first author's senior honors thesis at Cornell University. We would like to thank Rachel Schlund and Margaret Diaz for their helpful comments on earlier versions of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 We use male pronouns when describing our studies and results to align with the usage of male applicants in our studies.

2 Our OSF page can be found at the following link: https://osf.io/y94n8/

3 A few students took the survey more than once. Throughout the paper, when participants took the study multiple times, we used their first set of responses in the analyses.

4 In Studies 1–3, we asked participants to provide their initial assessment of how likely they would be to hire Thomas after seeing his resume on a 7-point scale. This was asked before the manipulation and was thus expected to be high across conditions. We had no plans to include this variable in our analyses to test our hypotheses, and therefore, we left it out of the results sections.

5 Some best practices for BTB laws recommend that employers wait to conduct a background check until after a conditional offer of employment has been made (Rodriguez, 2015). Therefore, our telling participants that the applicant was the most qualified lines up well with this practice.

6 Participants were also asked to report when they would feel comfortable hiring Thomas following his release from prison, whether they knew what ‘banning the box’ on employment applications was, and whether they knew of someone who had, or they themselves had, previously been incarcerated. Because these questions were not related to our hypotheses, the results from them are not reported in this paper. All additional items appear on our OSF page.

7 We had a variety of memory/attention check questions throughout our studies to make sure participants were following along. For consistency, across all studies, we excluded data from participants who incorrectly answered a memory question at the end of the study that asked them to report which condition they were in. Although some minor details of the results change depending on which attention check questions we use to exclude people, the overall conclusions of our research remain the same across various subsets of the data.

8 We included only three conditions instead of a full factorial design because we knew our sample size would be limited.

9 In previous studies, there was an indication that, since Thomas was previously convicted of a drug offense, he should not be hired to work at CVS, a store filled with various types of drugs. In order to eliminate the place of hiring as a potential factor in participants’ decision of whether or not to hire Thomas, the hiring company was changed from CVS to Michaels for this study.

10 We were originally planning on also excluding people who incorrectly answered a question about whether they saw the full explanation or not, but due to a typo in the study materials, we were not able to do this.

11 Participants failed the attention checks at different rates across conditions χ² (3, N = 199) = 12.02, p = .007, V = .246. They were the most likely to fail the attention check questions in condition 2, where Thomas had a criminal record and explained it (27.1% v. 11.7% across the rest).

12 Our counterbalancing factor of whether Thomas or Jeffrey was the one with the criminal record did not interact with our other manipulations, so we collapsed across this variable to simplify our analyses.

13 Note that over half of our participants in Study 3 had participated in the hiring of an employee in their real lives.