Once arrested for possession that crosses this threshold, the penalty is up to six months of jail time and a $500 fine. Similar laws vary by state, but most states consider possession of marijuana beyond a certain limit to be a misdemeanor. Nationwide, many drug arrests are still related to marijuana possession. Based on data disclosed by the FBI’s Uniform Crime Reporting program, 40% of drug arrests are marijuana-related, and in 2018, 92% of these arrests were for possession-related offenses. Thus, it is crucial to pay attention to misdemeanor records. Second, expunging felony records is more complicated, and automatic expungement may not be feasible for marijuana-related felonies. There are limited studies exploring the impact on employment access of having a misdemeanor record. Uggen et al. and Leasure , in particular, touch on the economic perspective, focusing on labor-market outcomes. Uggen et al. find that low-level misdemeanor records have statistically non-significant negative effects on hiring outcomes and Leasure find that a misdemeanor conviction significantly hinders hiring outcomes,regardless of the job applicant’s racial background. These two studies enriched the literature and are the foundation for this paper. Both studies answer several essential questions but with mixed results. First, Uggen et al.’s findings suggest that job applicants with misdemeanors should not be too concerned because the negative effect on employment outcomes of having a misdemeanor is statistically small.
However, the findings of Leasure suggest that the negative effect of a misdemeanor on job applicants’ employment outcomes can be quite significant. The present paper is motivated by this disparity because if having a misdemeanor record has no negative impacts on employment, as Uggen et al. suggest, plant grow table then it is unnecessary to expunge the records from an economics perspective. Second, both studies are field experiments with fictitious job applicants. Uggen et al. conduct an audit study, whereas Leasure use fictitious resumes for online job applications. Both methods are suboptimal for the following reasons: Currently, most advertisements for entry-level jobs are posted online and do not allow in-person application, so audit studies limit the data collection pools; the Covid-19 pandemic has resulted in job-search and hiring processes being more complex and time consuming, and it may take longer than usual to collect data through field experiments; and there is insufficient data to motivate additional states to enact legislation automatically expunging marijuana-related records and data collection through field experiments usually takes longer compared to laboratory experiments. To avoid these problems, I collected data through laboratory experiment, which allow the collection of large quantities of data in a short period and the easy modification of the independent variable.Employment is a key factor preventing recidivism ; it is also the primary source of legitimate income for ex-offenders and directly associated with housing and food security.
In addition, public support for expungement is high for persons convicted of substance-related offenses, who “signal” their reform through stable employment . However, many studies find that having a criminal record significantly lowers the chances of getting a job . Previous research studies show that employers are reluctant to hire ex-offenders because they are worried they will revert to criminal behaviors . The dilemma is that the general public supports expungement when reform is signaled through employment, but ex-offenders face difficulties finding jobs without expungement of their records. How to leave their criminal background in the past and start a new life with a job becomes one of the most important challenges for ex-offenders. Those with misdemeanor convictions are no exception. The punishment for their crimes may be light, but the effect on employment outcomes of having a misdemeanor record can be as severe as those for having a felony record. As Jacobs points out in The Eternal Criminal Record, it has become increasingly difficult to escape the mark of a criminal record . In most states in the U.S., misdemeanor records follow ex-offenders their entire lives and rarely get expunged. In California, background-check agencies are permitted to disclose crimes up to seven years after conviction. For example, if a college undergraduate is arrested and convicted at the age of 22 for excessive possession of marijuana in California , their record will follow them for the remainder of their lives, and a background check will disclose this record until they are 29 years old.
If they are only arrested and not yet convicted, the background check can disclose their arrest record for as long as the trial is ongoing. The trial can go on for months or, in some cases, a year or two. If they plead not guilty, this will prolong the process, and the arrest record will show up until the case is resolved. To expunge this misdemeanor record, the individual would need to file a petition and go through numerous other processes. However, with automatic expungement, this college undergraduate can pay the necessary fines and live their life without further concern about the record. However, it is worth asking whether there exists discrimination in relation to marijuana-related misdemeanors. The automatic expungement of marijuana related misdemeanors is only meaningful if it can mitigate any negative effects. Without negative impacts on employment, marijuana-related misdemeanor records may not need to be automatically expunged. Researchers who focus on misdemeanors have mixed results. Uggen et al. focus on drug-related arrests using an audit experiment. Subjects were paired by race, and one subject per pair was assigned to the treatment condition each week. The treatment condition was a sole arrest for disorderly conduct, with no resulting charge or conviction. Each week, the subjects applied for jobs in person and tried to have conversations with hiring authorities or other employers. If they could not meet with the employers directly, they maximized their in-person contact by asking to speak to a manager. Most of the communication on the subject’s end was scripted, such as asking questions about the pay and expressing their interest in the job. If and when treatment subjects were asked for more information by employers, they also divulged their criminal history and indicated that they had never been convicted. As is typical of such studies, the focus is the callback rate. The results suggest that although misdemeanor arrests lower applicants’ callback rates, the effect is minimal. For both white and black Americans, the difference between the callback rate of applicants with clean records and applicants who have been arrested for misdemeanors is only 4%. In this case, automatic expungement seems unnecessary. However, Leasure argues that the audit method limits the pool of employers considered because job openings are now increasingly posted online. In-person applications are rare. Uggen et al. also focus on individuals who had been arrested but not convicted; also, being convicted could have a different effect on employers’ decisions. Leasure pays attention to those convicted of drug-related misdemeanors and conducts a correspondence experiment, sending 1,000 fictitious resumes to employers in Columbus, Ohio, through online job application sites such as Craigslist. The data in the project are a sub-sample of the 1,000 resumes sent. The control group comprised applicants with clean records, and the treatment group included fictitious job applicants with drug-related convictions; other characteristics were controlled. The treatment groups were further divided into two sub-groups, one with applicants with felonies and the other with misdemeanor drug convictions. Similar to Uggen et al. , Leasure further study the racial disparity in results. The correspondence study focuses on callback rates, hydroponic table and the results suggest that a misdemeanor record significantly hinders an applicant’s chances of receiving a callback; indeed, applicants with misdemeanor records had a callback rate 13% lower than that of applicants with clean records.
This finding supports automatic expungement. Interestingly, Leasure does not find any significant difference between white and black applicants with or without criminal records, contrary to the findings of racial disparity in many previous studies. Uggen et al. and Leasure may be the only studies that primarily focus on misdemeanors from an economic perspective. The difference between their results is thus intriguing and requires further investigation. Furthermore, neither study focuses on expunging misdemeanor records. Thus, the present paper contributes the following: First, I focus on marijuana-possession-related misdemeanors to offer a new perspective on the automatic expungement of marijuana-related misdemeanor records from the perspective of employment outcomes. Second, this is a study of both misdemeanor-conviction and arrest records rather than only one or the other in order to check whether there are differences in discriminatory attitudes towards them. Third, I incorporate both racial and gender differences when studying the effects of a misdemeanor record on job applicants’ employment outcomes. In 1980, women comprised only 14% of misdemeanor arrests; this number jumped to 25% in 2016 . More women are arrested for misdemeanors, and the employment access of these female ex-offenders deserves more attention. Finally,the present study takes advantage of previous studies and uses a 10-point-scale evaluation that was designed in response to Heckman’s critique, discussed below.Vignettes have long been used in laboratory experiments to study gender discrimination in the labor market. They are also used in other studies, such as those focusing on ethnic and age discrimination and discrimination on the basis of physical appearance . When studying discrimination in hiring processes using vignettes, researchers usually change the variable of interest – for example, gender is used in gender discrimination studies – and use the correspondence method to check whether employers or laboratory subjects decide to hire or promote based on the characteristics reflected in fictitious resumes. Heckman and Heckman and Siegelman challenge the validity of such correspondence studies and offer several criticisms. They argue that discrimination can be identified through experiments, but these cannot differentiate between taste-based or statistical discrimination. They are also concerned about variance in the unobserved productivity variables, which can result in biased estimates of discrimination . Neumark further elaborate this second point and propose using a heteroskedastic probit model. In light of Heckman’s critique, K¨ubler, Schmid, and St¨uber offer a 10-point-scale experimental design in their gender-discrimination study. They argue that instead of using a binary invitation decision , using a 10-point evaluation scale allows for robust results, as long as the scale covers all the evaluations the employers might wish to make . Rather than measuring callback rates, they asked employers to express, on a scale of 1 to 10, how likely they were to select the fictitious resumes for the next round of interviews.The experiment uses oTree software . The study’s subjects were UC Irvine undergraduate students from the UCI Experimental Social Science Laboratory. The experiment’s show-up payment was $7, and subjects earned additional money during the experiment. The experiment proceeded as follows: first, subjects reviewed 21 fictitious resumes for an entry-level position as a customer service representative. They knew the resume format was standardized and that the applicants had already passed initial screening. This was done to guide the subjects to not focus on the job experience reflected on the resumes. The detailed instructions to subjects are set out in the Appendix. After reviewing the resumes, subjects rated the likelihood of each applicant being hired on a 10- point-scale . The subjects were also asked to leave brief comments on each resume to explain their rating decision. Resumes are divided into seven groups based on their characteristics. In the control group , the job applicants are white males. I chose white-sounding names to indicate race. Names were randomly chosen from the 2010 U.S. census data. This method has been used in many studies, including Bertrand and Mullainathan’s study . The resumes contained similar elements with minor modifications. For example, all applicants were high school graduates from different cities in Orange County, California, and all had entry-level, customer-service-related job experience, but in different fields or at different companies. The applicants were in their early 20s. All criminal records were recent because, in California, criminal convictions can only be reported for seven years unless another law requires employers to look deeper into an applicants’ background. These minor changes were necessary because a 10-point evaluation assumes that the characteristics reflected in the resumes cover all aspects employers care about. The resume template follows Leasure . The resumes in the first treatment group have all the features of the control group, but with an added misdemeanor record and contain the following statement to signal this criminal history: “the job applicant has a misdemeanor record and allows us to run a background check on him/her.” This conveys a message that the job applicants chose to self-disclose their criminal history, and this was noted by human resources staff .