Thomas J. Miles
Thomas J. Miles, Racial Disparities in Wiretap Applications Before Federal Judges, 41 Journal of Legal Studies 419 (June, 2012) (31 footnotes)
The criminal justice system is marked by sharp racial disparities. A black person is more likely to be searched, to face serious criminal charges, to be denied bail, to receive a sentence of incarceration, and to receive a longer prison sentence. [fn1] An important and perennial question is whether discrimination by decision makers in the criminal justice system causes these disparities. The question is difficult to answer for two reasons. Observed racial disparities usually reflect the discretionary choices by multiple actors. For example, a police officer's decision whether to arrest, a prosecutor's choice whether to pursue charges and which charges, and a judge's choice of punishments may all contribute to the disparity in sentences. But these individual decisions are often not directly observed, which makes it impossible to relate the contribution of each to the disparity. In addition, quantitative researchers do not observe all of the information on which criminal justice actors rely in rendering their decisions. Any discrimination must be disentangled from unobserved defendant and offense characteristics that might alternatively account for the disparity. A central concern of the existing literature is determining whether discrimination or omitted variables account for the disparities.
This paper takes a different approach. It studies a situation in which the activity observed is the decision of a single actor in the criminal justice system and in which the immediate outcome is certain. The actor studied here is the assistant U.S. attorney. The federal prosecutor exercises vast and largely unreviewable discretion in choosing which charges to bring and against whom. [fn2] In recent decades, the growth of federal criminal statutes and the establishment of the (formerly) mandatory Federal Sentencing Guidelines further enhanced the federal prosecutor's power to obtain plea bargains. Some see the federal prosecutor as an adjudicator in addition to an investigator and enforcer (Stuntz 2001; Barkow 2009). At the same time, substantial racial disparities persist in federal sentences and in outcomes directly within the purview of the prosecutor, such as the likelihood that a defendant receives a sentencing reduction for cooperation (Albonetti 1987; Mustard 2001; Johnson, Ulmer, and Kramer 2008). It is therefore important to understand how race influences federal prosecutors.
The activity studied here is the prosecutor's choice of which federal judge will review an application to conduct wiretap surveillance. Before instituting a wiretap for a suspect, federal investigators must obtain a warrant from a judge, and in most judicial districts the individual prosecutor has discretion to choose which judge within the district will review the wiretap application. Upon making that choice, the prosecutor appears directly before the judge (usually in the judge's chambers), delivers the application for a wiretap warrant, and gives an oral summary of it, highlighting the nature of the investigation and the supporting evidence. The prosecutor answers any questions the judge asks about the application. In essence, the prosecutor must choose one judge from the local bench and have a relatively brief but direct personal interaction with her.
Even before seeking judicial approval for a wiretap, a federal prosecutor must obtain permission from the Department of Justice (DOJ). The DOJ's internal review involves multiple steps and multiple reviewers, who apply standards for the use of wiretaps that are higher than the statutory and constitutional requirements that courts apply. In so doing, the DOJ screens out marginal applications before they reach the stage of judicial review. Consequently, all federal judges invariably approve the wiretap applications that reach them.
The certainty of judicial approval implies that a prosecutor need not select a reviewing judge with an eye to maximizing the likelihood of approval. That is, the prosecutor has no incentive to shop for a judge who is favorably disposed to law enforcement. In selecting a reviewing judge, a prosecutor may then pursue other objectives, such as minimizing the cost of interacting with a judge. These costs may include a preference for interacting with persons with certain characteristics, a preference that may operate consciously in the manner of Becker (1957) or unconsciously, as with implicit discrimination (Bertrand, Chugh, and Mullainathan 2005). The indulgence of the prosecutor's personal preferences is an agency cost, but with the legal adequacy of the application assured, it seemingly does not impair the immediate objective of receiving approval.
The process of obtaining judicial approval for wiretaps has not previously been empirically examined. Yet the prosecutor's choice of reviewing judge is intriguing because the legal and institutional arrangements governing that choice resemble the experimental conditions in tests for implicit discrimination (see Bertrand, Chugh, and Mullainathan 2005). The prosecutor's task is a relatively simple one, the selection of one judge to approach personally for review of the application. The prosecutor does not have to articulate a justification of this choice, and the criteria for this choice are ambiguous and entirely within the prosecutor's discretion. He usually makes the choice under time pressure. In addition to managing other cases, the prosecutor hopes to establish wiretap surveillance before the targeted criminal activity terminates. The conditions under which a prosecutor chooses the reviewing judge thus appear conducive to the operation of implicit discrimination, but, importantly, these conditions arise not in laboratories with experimental subjects but in actual courthouses with active prosecutors.
This paper tests these predictions using a data set of all wiretaps used in federal criminal investigations during the years 1997-2007. The results show that, consistent with the absence of an incentive to seek favorable judges, judicial characteristics such as ideology and prior professional experience do not influence the number of wiretap applications a judge receives. But the estimates also show that African American judges receive about 25 percent fewer wiretap applications than do other judges, even after numerous judicial and district characteristics are controlled for. The racial disparity in wiretap applications is robust to a variety of controls for district-level differences, including district-level trends and district-by-year fixed effects. The disparity persists across the type of phone tapped, the offense investigated, and even the day of the week and season of the year in which surveillance occurs.
The paper considers--and ultimately discounts--a variety of explanations for the racial disparity in wiretap applications. Differences in the location of judges' chambers and in their educational and experiential backgrounds do not explain the disparity. Similarly, variation in caseloads and attorneys' perceptions of judicial performance fail to account for the disparity. Nor do a judge's past sentencing practices or ideology correlate with the disparity. In addition, the results are not consistent with an account in which prosecutors reserve the strongest wiretap applications for the most skeptical judges. Judicial characteristics do not correlate with the duration of a wiretap's surveillance or the rates at which wiretaps result in arrests, convictions, and motions to suppress. In fact, review by an African American judge correlates with a lower rate of intercepting incriminating conversations.
The possibility that race may influence the decisions of prosecutors will be unsurprising to those who perceive the racial disparities of the criminal justice system as evidence of discrimination (McAdams 1997, pp. 642-52). Others may grant that few prosecutors are consciously discriminatory (Davis 2007, p. 203), and still others may doubt that sophisticated and closely monitored decision makers, such as federal prosecutors, could be even unconsciously discriminatory. Prosecutors are well versed in antidiscrimination law, trained to base their judgments on law and fact, and face the risk of appeal and even professional sanctions for allowing arbitrary factors such as race to influence their decisions. Yet the psychological literature shows that individuals, even those espousing egalitarian values, may suffer from implicit biases (see, for example, Bertrand, Chugh, and Mullainathan 2005; Greenwald and Krieger 2006; Jolls and Sunstein 2006; Rachlinski et al. 2009; but also see Mitchell and Tetlock 2006).