Annotated bibliography:1 paragraph for each article and this is what the paragraph for each article need to include- How the data was collected and from whom- small and straight forward summary- finds/ result
Microsoft Word - blalock_devaro_leventhal_simon.doc 1 Gender Bias in Power Relationships: Evidence from Police Traffic Stops Garrick Blalock, Jed DeVaro, Stephanie Leventhal, and Daniel H. Simon† First Version: December 31, 2006 This Version: February 22, 2007 We test for the existence of gender bias in power relationships. Specifically, we examine whether police officers are less likely to issue traffic tickets to men or to women during traffic stops. Whereas the conventional wisdom, which we document with surveys, is that women are less likely to receive tickets, our analysis shows otherwise. Examination of a pooled sample of traffic stops from five locations reveals no gender bias, but does show significant regional variation in the likelihood of citations. Analysis by location shows that women are more likely to receive citations in three of the five locations. Men are more likely to receive citations in the other two locations. To our knowledge, this study is the first to test for gender bias in traffic stops, and clearly refutes the conventional wisdom that police are more lenient towards women. † Blalock, Leventhal, and Simon are at the Department of Applied Economics and Management, Cornell University. DeVaro is at the School of Industrial and Labor Relations, Cornell University. We thank Nicola Persico at the University of Pennsylvania for graciously providing the data. We also thank Kate Antonovics at the University of California, San Diego, and Brian Withrow at Wichita State University for providing documentation for some of the data. Please direct correspondence to Garrick Blalock, 346 Warren Hall, Ithaca, NY 14853, +1 (607) 255-0307, garrick.blalock@cornell.edu. 2 I. INTRODUCTION For many decades the subject of differences between men and women in their treatment and labor market outcomes has been a core research topic in labor economics. A large fraction of this literature concerns gender differences in wages. A smaller fraction concerns gender differences in other labor market outcomes such as promotions. Questions about wages and promotions concern the extent to which worker productivity in the labor market is rewarded or compensated. In this paper we consider an alternative approach, to study punishments rather than rewards, and we do so in a non-workplace setting. That is, we consider the question of whether identifiable transgressive behaviors are more likely (or less likely) to be punished if the perpetrator is female rather than male. Specifically, we examine whether male drivers are more likely than female drivers to be ticketed when stopped by the police. Although the distinction between rewards and punishments may be somewhat artificial, in the sense that any reward can be reinterpreted as the negative of a punishment, and vice versa, we believe that approaching the study of gender differences in treatments from the angle of punishments rather than rewards offers some noteworthy advantages. Further, for the purpose of understanding gender differences in punishments, studying the treatment of traffic violations is particularly revealing. The notions of transgression and punishment in this context are simple and clearly defined. A transgression involves either a moving violation or a visible failure to maintain a vehicle, such as an expired inspection sticker or burned-out headlight. Punishment is defined as the receipt of a citation. A further advantage of the context is that the nature of the social interactions that precede and influence the decision about whether or not to punish is simple and clear, as well as constant across sample observations. Typically the incidents in question involve a focused interaction between only two parties (a single observer and a single perpetrator) for a relatively short period of time.1 In contrast, if we were to consider instead transgressions and punishments in the typical workplace setting, the nature and scope of transgressions would be quite diverse (e.g., arriving late for work, arguing with the boss, stealing from the company, using company phones for personal calls, shirking, or general poor performance), and the 1 We have data on the duration of stops in two locations. In Wichita, Kansas, more than two thirds of all stops take 15 minutes or less, while in Bloomington, Illinois, more than three fourths of all stops take less than 15 minutes, with the median stop taking 10 minutes. 3 punishments could range from mild rebukes to small or non-existent raises, denials of promotion, or firings. Furthermore, the nature of the social interactions that surround both transgressions and punishments is far more complex in the typical workplace, and incidents often involve multiple perpetrators and/or observers. All of these issues would complicate an analysis of gender differences in punishments. A simple conceptual model is helpful for organizing one’s thoughts about gender differences in the probability of punishment. Consider two individuals, called the perpetrator and the observer. The perpetrator is defined as an individual who might commit a transgressive action, and the observer is defined as an individual who can potentially observe a transgression and punish it. Let “Transgress”, “Caught”, and “Punish” denote the three events that the perpetrator commits a transgression, that the transgression is identified by the observer, and that the perpetrator is actually punished. Under the assumptions that the perpetrator is caught only when a transgression has been committed and that the perpetrator is punished only when a transgression has been committed and caught, we can decompose the probability that the perpetrator is punished into the product of three probabilities, as follows: Prob(Punish) = Prob(Punish,Caught,Transgress) = Prob(Punish | Caught,Transgress) × Prob(Caught | Transgress) × Prob(Transgress) From this expression, it is clear that an observed gender difference in the probability of punishment could arise from a gender difference in any of the three probabilities in the product. That is, if women are more likely to be punished, this might be because women are more likely to commit a transgression, or because they are more likely to be caught after having committed a transgression, or because they are more likely than men to be punished after having committed a transgression and been caught, or because of any combination of these possibilities. From the standpoint of thinking about whether punishment probabilities differ between women and men and whether such differences are due to gender discrimination, the three probabilities in the product are not all equally interesting. In particular, Prob(Transgress) is not of interest, since it concerns only the behavior of the perpetrator and not the interaction between the perpetrator and the observer, so it sheds no light on whether female perpetrators and male perpetrators are treated differently by observers. Furthermore, Prob(Caught | Transgress) will usually be of limited interest. While this probability is potentially affected by the behaviors of 4 both the perpetrator and the observer, in many cases the perpetrator’s gender is not known by the observer until after the perpetrator is caught, which limits the usefulness of this probability as an indicator of gender differences in treatment. Of greatest interest is Prob(Punish | Caught,Transgress), and this probability is the focus of this paper, since the perpetrator’s gender is virtually always known conditional on being caught, and whether or not punishment occurs frequently hinges on the nature of the interactiontypically occurring only after the perpetrator is caughtbetween the perpetrator and the observer. Given that the perpetrator commits a transgression and is caught, is the probability of punishment related to whether the perpetrator is male or female? That is the fundamental question that concerns us. Our particular context is that of traffic violations in five distinct regions of the United States: Bloomington, Illinois; Highland Park, Illinois; Wichita, Kansas; Boston, Massachusetts; and the entire state of Tennessee. In each incident involving a stopped vehicle, the perpetrator is the driver who committed a violation, and the observer is the police officer who stopped the driver. Conditional on having committed a violation and been stopped, a driver may or may not be punished by receiving a citation. Our main empirical research question is: Conditional on having been stopped, does the probability of getting a ticket depend on the driver’s gender?2 According to conventional wisdom, the answer to this question is that men are more likely to be ticketed than women. To document the conventional wisdom, we conducted a poll of students in various classes at Cornell University and elsewhere. In each class, on a given day we distributed the following question, randomizing the order of the three responses to avoid issues of framing associated with the order and collecting the anonymous written responses before the students left the classroom:3 2 We do not study Prob(Trangress); while it is quite possible that the probability of committing a violation differs between men and women, our focus in this paper is on potential gender differences in treatment. Similarly, we do not study Prob(Caught | Transgress), though it is quite possible that this probability also differs by gender. This might happen because of a gender difference in behavior (for instance, women might spend more time driving in school zones when school is in session, and police might tend to patrol in such areas and enforce the speed limit with particular zeal) or because in some cases the police officer might observe the driver’s gender in advance, and this might influence the officer’s decision to stop the driver, though in many cases the driver’s gender may not be observed by the officer until after the decision to stop is made. This would