She and Luke Stein conducted a randomized controlled field experiment testing for racial discrimination in secondary markets. The yearlong experiment consisted of approximately 1,200 online classified advertisements placed in over 300 areas throughout the United States, both small towns and large cities, with the intention of testing for racial bias among buyers. The experiment featured photographs of a person's hand holding an iPod. Doleac and Stein randomly varied the skin color of the person's hand: dark-skinned, light-skinned, or light-skinned with a tattoo. In all other respects, the photographs were nearly identical. Consistent with other studies on racial bias in the marketplace, Doleac and Stein found that black sellers did considerably worse than white sellers along numerous dimensions. For instance, black sellers got 13 percent fewer responses and 18 percent fewer offers of purchase from interested buyers. When they did receive an offer, that offer was around 11 percent lower. These effects were similar to those associated with the white tattooed hand which was the placebo group in the study.
DNA databases
Doleac has also written on issues related to contemporary crime control, such as the use of DNA databases to deter and capture criminals, as well as the role of outside lighting. Her work on lighting with Nicholas J. Sanders used a 2007 natural experiment where Congress increased daylight saving time by four weeks. They added three weeks in the spring and one week in the fall. Doleac and Sanders used this as an opportunity to study the effect that daylight, separate from other seasonal factors, might have on crime, and found that when daylight saving time began in the spring, robbery rates for the entire fell 7 percent with an additional 27 percent drop during the evening hour that gained extra sunlight. This study is relevant for broader consideration in crime control that focuses on expanding lighting to high crime areas with external lighting. Doleac has also studied the role of DNA databases on reducing crime both through deterrence and capturing criminals. Several states have passed laws allowing DNA testing of arrestees. Following famous work by Nobel Laureate Gary Becker, Doleac hypothesizes that the availability of these DNA databased would deter criminals as well as catch criminals insofar as criminals are sensitive to the risk of arrest. Violent offenders who have had their DNA collected and catalogued in digital databases were 23 percent more likely to be convicted of another crime implying that profiled offenders were caught more often than those who hadn't been stored in the database. Given the extraordinarily high costs of incarceration, Doleac suggests that DNA databases are considerably more cost-effective than other ways of reducing crime such as hiring more police or through incarceration.
Ban the Box
Doleac has also written extensively both an academic study and general interest writings on prisoner reentry. Prisoner reentry is a considerable challenge in the United States due to the ubiquity of background checks and the myriad of obstacles prisoners face reentering the labor force upon release. An increasingly popular policy at the federal, state and municipality level is the "ban the box" initiative. Employers often will ask applicants for jobs if they have any prior arrests or convictions by checking Yes or No on a "box". The ban the box initiative are laws, oftentimes at the city level, banning employers' right to ask this information of applicants during the earliest stages of the application process. In a 2018 article published in the Journal of Labor Economics, Doleac and Benjamin Hansen from the University of Oregon found that the laws had unintended consequences on the hiring of low skill minorities. The authors found that ban the box policies lowered the chances of employment by 5 percent for young, low-skilled black men and almost 3 percent for young, low-skilled Hispanic men. The authors argue that this effect was caused by rampant statistical discrimination in labor markets. Without information on a candidate's criminal history, employers make extreme racially biased assumptions that the applicant is similar to his or her demographic. Thus seeing a black applicant, even one without a criminal history, the firm "fills in" the missing information by assuming he or she is like the group average, and thus is more likely to believe the applicant has a criminal history, even when they do not.
