The Leadership Conference on Civil and Human Rights

The Nation's Premier Civil and Human Rights Coalition

The Leadership Conference on Civil and Human Rights  & The Leadership Conference Education Fund
The Nation's Premier Civil and Human Rights Coalition

Chapter Two: Race and Prosecutorial Discretion

Criminal Justice Report Banner
Table of Contents
grey arrow Executive Summary
grey arrow Introduction
grey arrow Chapter One: Race and the Police
grey arrow Chapter Two: Race and Prosecutorial Discretion
grey arrow Chapter Three: Race, Sentencing and the "Tough Crime" Movement
grey arrow Chapter Four: Willful Judicial Blindness
grey arrow Chapter Five: Race and Juvenile Justice System
grey arrow Chapter Six: Consequences of Too Little Justice
grey arrow Chapter Seven: Recommendations
Chapter II

Race and Prosecutorial Discretion

            Racial profiling and other enforcement strategies begin the insidious process by which minorities are disproportionately caught up in the criminal justice system. But such disparities do not end at the point a suspect is arrested. At every subsequent stage of the criminal process -- from the first plea negotiations with a prosecutor, to the imposition of a prison sentence by a judge -- the subtle biases and stereotypes that cause police officers to rely on racial profiling are compounded by the racially skewed decisions of other key actors. This chapter examines the role of the prosecutor in perpetuating racial inequality in our criminal justice system.

            Prosecutors occupy a central role in American criminal justice. They represent the public in the solemn process of holding accountable those who violate society’s rules. That task carries with it substantial unchecked discretion. The threshold decision of whether to bring charges against a suspect, and, if so, which charges are appropriate, is almost never subject to review by a court. The subsequent decision to enter into a plea agreement is reviewable at the margins because courts may reject plea agreements under certain circumstances. But in practice, prosecutors decide who will be granted the leniency that a plea bargain represents.

            Prosecutorial discretion is most dramatically exercised in the area of sentencing. Traditionally, sentencing has been a judicial prerogative, but, as will be explained, the advent of mandatory minimum sentencing laws and sentencing guideline systems has shifted in large measure the power to determine punishment from judge to prosecutor. Even where judges retain ultimate authority to impose sentence, a prosecutor’s sentencing recommendation will carry great weight. Prosecutors today enjoy more power over the fate of criminal defendants than at any time in our history.

            Regrettably, the evidence is clear that prosecutorial discretion is systematically exercised to the disadvantage of black and Hispanic Americans. Prosecutors are not, by and large, bigoted. But as with police activity, prosecutorial judgment is shaped by a set of self-perpetuating racial assumptions.

A.          The Decision to Prosecute

            The first and most basic prosecutorial decision is whether to pursue a particular criminal case at all. Prosecutors have the authority to decline prosecution altogether, or to authorize diversion, under which completion of drug treatment or community service results in the dismissal of the charges. But such displays of prosecutorial mercy appear to be exercised in a manner that disproportionately benefits whites.

            In 1991 the San Jose Mercury News reviewed almost 700,000 criminal cases from California between 1981 and 1990 and uncovered statistically significant disparities at several different stages of the criminal justice process. Among the study’s findings was that six percent of whites, as compared to only four percent of minorities, won “interest of justice” dismissals, in which prosecutors dropped a criminal case entirely. Moreover, the study found, 20 percent of white defendants charged with crimes providing for the option of diversion received that benefit, while only 14 percent of similarly situated blacks and 11 percent of similarly situated Hispanics were placed in such programs.

            Related to the decision to decline prosecution is the decision to charge the defendant in state or federal court. This choice is presented when jurisdiction over the crime resides concurrently in state and federal court, as in many drug cases. The police may make an initial decision of whether to bring the evidence to state or federal prosecutors (a discretionary call that may itself be influenced by racial considerations), but the authority to determine that a defendant will be federally prosecuted rests with federal prosecutors. Typically they will decline to prosecute the defendant in federal court if the case does not seem “serious” enough or if the defendant does not seem to pose a significant threat to public safety. Obviously these are not scientific judgments – rather they are exercises of discretion informed by predictions, hunches and preconceptions, some of which are racially tinged.

            The decision of whether to prosecute a drug case in federal court has important consequences for the defendant because federal sentences are notoriously harsher than state sentences. Federal parole was abolished in 1987, and federal drug convictions frequently result in lengthy, mandatory sentences. Moreover, if the prosecutor includes in the indictment charges carrying mandatory penalties and then refuses to permit a plea to other charges, the defendant has no opportunity to undergo drug treatment as an alternative to imprisonment, since federal law does not offer judges that option.

            According to the United States Sentencing Commission, federal courts in 1990 sentenced drug traffickers to an average of 84 months in prison, without possibility of parole. By contrast, state courts in 1988 sentenced drug traffickers to an average maximum sentence of 66 months, resulting in an average time served of only 20 months. Thus, the decision of a prosecutor to bring federal charges, rather than letting the case proceed in state court, can result in a prison term that is years longer than the sentence that would likely result from state prosecution.

            That the prosecutorial decision to bring charges in federal, rather than state, court is often exercised to the detriment of minorities is best demonstrated by statistics on crack cocaine prosecutions. In 1986 Congress enacted especially harsh mandatory minimum penalties for these offenses. From 1988-1994, hundreds of blacks and Hispanics – but no whites – were prosecuted by the United States Attorney’s office with jurisdiction over Los Angeles County and six surrounding counties. The absence of white crack defendants in federal court could not be ascribed to a lack of whites engaged in such conduct; during the 1986-1994 period, several hundred whites were prosecuted in California state court for crack offenses.

            National statistics tell the same story: From 1992-1994, approximately 96.5 percent of all federal crack prosecutions were of non-whites. A 1992 U.S. Sentencing Commission Report determined that only minorities were prosecuted for crack offenses in over half of the federal judicial districts that handled crack cases. And during that period, in New York, Texas, California, and Pennsylvania combined, eight whites were convicted of crack offenses; the number of white crack defendants convicted in Denver, Boston, Chicago, Miami, Dallas and Los Angeles combined was zero, compared to thousands of convictions of black and Hispanic crack offenders.

            These discrepancies are remarkable because the crack epidemic knew no racial bounds. Despite stereotypes perpetuated by the media and popular culture, government statistics show that more whites overall used crack than blacks. According to the National Institute on Drug Abuse, between 1991 and 1993 whites were twice as likely to have used crack nationwide than blacks and Hispanics combined. Crack use was somewhat more concentrated in minority communities, but in Los Angeles, for example, whites comprised more than 50 percent of those who had ever used crack, and about one-third of those who could be termed “frequent users.”

            The exercise of prosecutorial discretion in favor of white crack defendants and against black crack defendants is illustrated by the parallel cases of two men – one white, the other black – charged with cocaine trafficking in Los Angeles. Stephen Green, a black man, was arrested with 70 grams of crack and sentenced in federal court to 10 years in prison – the mandatory minimum federal sentence for selling more than 50 grams of crack. Daniel Siemanowski, a white man, was arrested with 67 grams of crack, and was also therefore eligible for the 10-year mandatory sentence. But he was tried and convicted in state court, and received a jail sentence of less than a year.

            Federal law enforcement authorities have disputed that the wide discrepancy in federal crack prosecutions reflects differential treatment based on race, arguing that authorities target high-volume traffickers of whatever race. But since it is empirically true that more whites than non-whites used crack during this period, this argument presumes that whites largely bought their crack from non-whites. Studies, however, suggest the contrary – that drug users tend to purchase their drugs from individuals of the same race as the user, and that drug seller racial breakdowns are similar to drug user racial breakdowns. Of course some suburban residents drive into the inner city to buy drugs on the street, and of course those open-air street markets present an easy enforcement target for the police. But most drug deals occur behind closed doors, in offices and in private homes that the police don’t patrol. Just because most of the dealers hawking crack vials on the street corner were not white in the late 1980’s does not mean whites did not participate aggressively in the crack distribution network.

            In any event, the reality is that many black defendants prosecuted in federal court are not high-volume traffickers. According to Los Angeles federal district judge J. Spencer Letts, “those high in the chain of drug distribution are seldom caught and seldom prosecuted.” Instead, federal prosecutorial efforts have focused predominantly on low-volume dealers and low-level couriers who happen to be black and Hispanic. U.S. District Court Judge Consuelo B. Marshall has observed: “We do see a lot of these [crack] cases and one does ask why some are in state court and some are being prosecuted in federal court . . . and if it’s not based on race, what’s it based on?”

B.          Charging Decisions and Plea Bargaining

            Once a prosecutor decides to bring charges against an individual, plea negotiations present the next opportunity for a prosecutor to grant some degree of leniency to a defendant, or to insist on maximum punishment. Prosecutors have virtually unlimited discretion to enter into an agreement by which the defendant will plead guilty in exchange for the dismissal of certain charges or a reduced sentence, and once again the exercise of discretion is characterized by racially disparate results.

            The San Jose Mercury News report discussed above revealed consistent discrepancies in the treatment of white and non-white criminal defendants at the pretrial negotiation stage of the criminal process. During 1989-1990, a white felony defendant with no criminal record stood a 33 percent chance of having the charge reduced to a misdemeanor or infraction, compared to 25 percent for a similarly situated black or Hispanic. Between 1981 and 1990, 50 percent of all whites who were arrested for burglary and had one prior offense had at least one other count dismissed, as compared to only 33 percent of similarly situated blacks and Hispanics. Blacks charged with a single offense received sentencing enhancements in 19 percent of the cases, whereas similarly situated whites received such enhancements in only 15 percent of the cases.

            Over the course of 700,000 cases, these discrepancies establish a clear pattern of unfair treatment for thousands of black and Hispanic criminal defendants. The extent of disparate treatment in individual cases can be stark indeed. Consider the fates of two individuals – one black, one white (unnamed in the Mercury News report) – each charged with four criminal counts: three counts of burglary, and one count of receiving stolen property. Neither man had been to jail before. Neither had used a weapon in the offense. Drugs were involved in neither crime. Both men entered into plea bargains. But the black man was required to plead guilty to all four criminal charges and received an eight-year sentence. The white man was permitted to plead guilty to a single burglary charge and received a sentence of 16 months.

            Statistics from other jurisdictions confirm that prosecutorial discretion may result in disparate treatment of minorities and whites. The State of Georgia has a “two strikes, you’re out” law, under which a life sentence may be imposed for a second drug offense. Under the Georgia scheme, the State’s district attorneys have unfettered discretion to seek this penalty. As of 1995, life imprisonment under the “two strikes” law had been imposed on 16 percent of eligible black defendants, while the same sentence had been imposed on only one percent of eligible white defendants. Consequently, 98.4 percent of those serving life sentences under Georgia’s “two strikes, you’re out” law are black.

            Statistics in federal court mirror the experiences in these states. A United States Sentencing Commission report found that, for comparable behavior, prosecutors offered white defendants plea bargains that permitted the imposition of sentences below what would otherwise be the statutory minimum more often than they offered such deals to blacks or Hispanic defendants. Moreover, federal prosecutors have sole authority to grant a departure below the mandatory minimum level based on substantial assistance to authorities, another means by which leniency may be offered to some defendants and not others.

C.          Bail

            Another turning point in the criminal justice process, one that can mean the difference between freedom and incarceration for criminal defendants, is the bail determination. While the decision to set bail is ultimately a judicial function, prosecutors play an important role in determining whether a criminal defendant will be released on bail or detained in jail prior to trial by recommending detention or release.

            A New York State study examined the extent to which black and Hispanic criminal defendants were treated differently from similarly situated white criminal defendants with respect to pretrial detention, and concluded that statewide, minorities charged with felonies were detained more often than white defendants charged with felonies. Indeed, the study found that 10 percent of all minorities held in jail at felony indictment in New York City, and 33 percent of all minorities held in jail at felony indictment in the rest of New York State, would be released before arraignment if minorities were detained as often as comparably situated whites.

            Another study reviewed bail determinations for criminal defendants in New Haven, Connecticut, and concluded that the bail rates set for black defendants exceeded those set for similarly situated white defendants. In short, the study concluded, lower bail rates could have been set for black defendants without incurring the risk of flight that bail rates are designed to avoid. And federal statistics indicate that while non-Hispanics are likely to be released prior to trial in 66 percent of cases, Hispanics are likely to be released only 26 percent of the time.

            Bail status not only determines whether the defendant is to be incarcerated before trial, it also bears on the likelihood of conviction. Although jurors are not supposed to know whether the defendant has been jailed before trial, they can often discern the defendant’s bail status and are more likely to convict a defendant who has already been incarcerated. Here again, one racial disparity begets disparity further along in the justice system.

D.          The Death Penalty

            Thirty-eight states and the federal government authorize capital punishment. In each of those jurisdictions it is the prosecutor who makes the critical decision of whether or not to seek death. That decision is guided somewhat by statutory aggravating and mitigating factors, but many of these factors, such as the heinousness of the crime, are subjective. Judges and juries may eventually reject a prosecutor’s request that the death penalty be imposed, but prosecutors alone decide whether death is an option.

            The importance of race as a factor in the imposition of capital punishment is well documented. First, the evidence reveals disparity in the application of the death penalty depending on the race of the victim. Individuals charged with killing white victims are significantly more likely to receive the death penalty than individuals charged with killing non-white victims. Of numerous studies of death penalty outcomes reviewed by the congressional General Accounting Office (GAO), 82 percent found that imposition of the death penalty was more likely in the case of a white victim than in the case of a black victim. One of the most thorough death penalty studies, conducted by Professors David Baldus, Charles Pulaski, and George Woodworth, found that defendants charged in Georgia with killing white victims were 4.3 times more likely to receive the death penalty than defendants charged with killing black victims. The Baldus study also found that more than 50 percent of those sentenced to death for killing a white person would not have received the death penalty had they killed a black person. According to the GAO, the effect of the victim’s race on the sentencing outcome appears to be particularly pronounced at the earlier stages of the judicial process, such as the prosecutor’s decision to charge the defendant with a capital offense and then whether or not to accept a guilty plea to lesser charges.

            Second, while some of the evidence concerning the death penalty reveals that the race of the defendant alone does not result in unwarranted disparity, other evidence is to the contrary. It is at least true that the race of the defendant, when combined with the race of the victim, yields significant disparities in the application of the death penalty. The Baldus study concluded that blacks who killed whites were sentenced to death 22 times more frequently than blacks who killed blacks, and seven times more frequently than whites who killed blacks. Again, this discrepancy appears to hinge on the exercise of prosecutorial discretion. Georgia prosecutors sought the death penalty in 70 percent of the cases involving black defendants and white victims, while seeking the death penalty in only 19 percent of the cases involving white defendants and black victims, and only 15 percent of the cases involving black defendants and black victims.

            In short, black defendants charged with killing white victims were the group most likely to receive the death penalty. Until 1991, when Donald Gaskins, a white man, was executed in South Carolina for the murder of a black victim, no white person had been executed for the murder of a black person since the Supreme Court’s 1976 decision in Furman v. Georgia holding that capital punishment is not necessarily unconstitutional. In all, since 1976, only 11 whites have been executed for the murder of a black victim, while 145 blacks have been executed for the murder of a white victim, and 80 percent of those currently on death row are there for killing a white person. And the Baldus Study revealed that of the seven individuals executed in Georgia between 1976 and 1986, all were convicted of killing whites, and six of them were black, despite the fact that of all homicides in Georgia during that period, only 9.2 percent involved black defendants and white victims, and 60.7 percent involved black victims.

            Still other studies indicate that capital punishment is disproportionately applied on the basis of the race of the criminal defendant, irrespective of the race of the victim. Professors Baldus and Woodsworth, co-authors of the groundbreaking Georgia study, conducted a statistical analysis of the death sentence in Philadelphia between 1983 and 1993, and concluded that for similar crimes, black defendants were almost four times more likely to receive the death penalty as white defendants, and that 38 percent of black defendants sentenced to death would not have been so sentenced had they been white.

            Statistics on the imposition of the federal death penalty are similarly disturbing. In 1988, Congress enacted the first federal death penalty provision in the aftermath of Furman. The 1988 law authorized the death penalty for murders committed by those involved in certain drug trafficking activities under 21 U.S.C. §848. From 1988 to 1994, 75 percent of those convicted under 21 U.S.C.§848 were white. However, of those who were the subject of death penalty prosecutions under that law in the same period, 89 percent were Hispanic or black (33 out of 37) and only 11 percent (four out of 37) were white. And, indeed, the first defendant scheduled to be executed under the 1988 law is Hispanic.

            A congressional subcommittee studied the application of the 1988 law and concluded that “some of the death penalty prosecutions under [21 U.S.C. §848] have been against defendants who do not seem to fit the expected ‘drug kingpin’ profile,” including, in several cases, “young inner-city drug gang members and relatively small-time traffickers,” or individuals who committed murder at the behest of a higher-up who received a lesser sentence.

            Under federal procedures, the personal written authorization of the Attorney General is required before a capital prosecution may proceed. Thus, the application of the federal death penalty involves the exercise of prosecutorial discretion at the highest levels of the United States Government. Nobody suggests that the current Attorney General or her predecessors who authorized federal death penalty prosecutions are or were motivated by impermissible racial factors. But this act of prosecutorial decision-making, like others at all levels of government, is subject to institutional and community pressures that may have racial overtones.

            As an empirical matter, it is undeniable that prosecutors exercise their discretion in ways that have racially disproportionate impacts, even if their intent is race-neutral. Such unfairness may ultimately be more dangerous than explicitly racist behavior, since it is harder to detect (both by victim and perpetrator) and harder to eradicate than the blatant racism most Americans have learned to reject as immoral. In any event, prosecutorial decision-making, in tandem with police tactics, contribute to the criminalization of race and the racialization of crime, a vicious cycle that is having a devastating effect on minority communities throughout the nation.

Bennett L. Gershman, “The New Prosecutors,” 53 U. Pitt. L. Rev. 393, 448 (1992)(“The American prosecutor, owing to a variety of social and political factors, has emerged as the most pervasive and dominant force in criminal justice.”). See also “Prosecution and Race” at 20-25 (noting the “extraordinary, almost unreviewable, discretion and power of prosecutors” and the necessity of such discretion, but also that “[t]he deficiency of prosecutorial discretion lies not in its existence, but in the randomness and arbitrariness of its application.”).

“Prosecution and Race” at 21-22 and nn. 28-33.

Christopher Schmitt, “Plea Bargaining Favors Whites, as Blacks, Hispanics Pay Price,” San Jose Mercury News, December 8, 1991 (Mercury News Report), p. 1A.

United States Sentencing Commission, Special Report to Congress: Cocaine and Federal Sentencing Policy (February 1995), p. 138 and nn. 186-188.

Dan Weikel, “War on Crack Targets Minorities Over Whites,” Los Angeles Times, May 21, 1995, p. A1.


Id. See also United States Sentencing Commission, Special Report to Congress: Cocaine and Federal Sentencing Policies at xi (Blacks and Hispanics accounted for 95.4 percent of crack cocaine distribution offenses in 1993).



See Ch. I, n. 23. Director of National Drug Control Policy Barry McCaffrey has also observed that youths purchasing drugs tend to purchase from members of their own race. Patricia Davis and Pierre Thomas, “In Affluent Suburbs, Young Users and Sellers Abound,” Washington Post, December 14, 1997, p. A20.

Weikel at A1.

Data analyzed by the U.S. Sentencing Commission reveals that nationwide, more than 94.5 percent of federal crack defendants in 1992 were either low- or middle-volume dealers or couriers, and only 5.5 percent were high-volume dealers. U.S. Sentencing Commission, Cocaine and Federal Sentencing Policy at 172.

Weikel at A1.

Most criminal cases end in a plea bargain. See Rebecca Hollander-Blumoff, “Getting to ‘Guilty’: Plea Bargaining as Negotiation,” 2 Harv. Negotiation L. Rev. 115, 117 n. 7 (1997) (discussing studies showing that as much as 90 percent of all criminal cases end in plea bargains).

See “Prosecution and Race” at 23-25 and nn. 41-59 (discussing importance of, and prosecutorial discretion inherent in, charging decisions and plea bargaining).

53 Mercury News Report at 1A.

54 Id.

No Equal Justice at 143. The Georgia Supreme Court initially held that these statistics presented a prima facie case of discrimination, and invalidated the “two strikes, you’re out” statute. Stephens v. State, 1995 WL 116292 (Ga. 1995). The court, however, reversed itself less than two weeks later upon being presented with a petition signed by all of Georgia’s 46 district attorneys claiming that the court’s approach, because it would apply to so many other areas of prosecution, such as the death penalty, would “paralyze the criminal justice system” in Georgia. No Equal Justice at 143 (citing Stephens v. State, 456 S.E.2d 560 (Ga. 1995)).

Race to Incarcerate at 139. The Justice Department contends that these disparities were due to legally relevant case-processing factors. Id. (citing Dale Parent, et al., National Institute of Justice, Mandatory Sentencing (January 1997), at p. 4).

18 U.S.C. § 3553(e). Similarly, the federal sentencing guidelines require a prosecutor to certify substantial assistance before the court may depart below the applicable guideline range. United States Sentencing Guidelines, Section 5K1.1.

Office of Justice Systems Analysis, New York State Division of Criminal Justice Services, “Disparities in Processing Felony Arrests in New York State, 1990-1992” (September 1995) (New York Felony Study), pp. v-vi, xi.

Ian Ayres and Joel Waldfogel, “A Market Test for Race Discrimination in Bail Setting,” 46 Stan. L. Rev. 987 (1994).

Bureau of Justice Statistics, United States Department of Justice, Compendium of Federal Justice Statistics (January 1999), p. 25.

See generally Bail Reform Hearings, 1982: Hearings on S. 1554 Before the Subcomm. on

Constitutional Rights of the Senate Comm. on the Judiciary, 97th Cong., 1st Sess. (1982) (statement of Professor Daniel Freed).

“Prosecution and Race” at 24 and nn. 49-51.

U.S. General Accounting Office, Death Penalty Sentencing: Research Indicates Pattern of Racial Disparities (1990) (GAO Study), p. 5 (noting that ‘[t]his finding was remarkably consistent across data sets, states, data collection methods, and analytic techniques.”).

David C. Baldus, Charles Pulaski, and George Woodworth, “Comparative Review of Death Sentences: An Empirical Study of the Georgia Experience,” Journal of Criminal Law and Criminology 74 (fall 1983) (Baldus Study), pp. 661-753.

McCleskey v. Kemp, 481 U.S. 279, 321 (1987) (Brennan, J., dissenting) (citing Baldus study).

GAO Study at 5.

The Baldus study concluded that black defendants were only 1.1 times more likely to receive the death penalty than white defendants. See also GAO Study at 6 (“The evidence for the influence of the race of defendant on death penalty outcomes was equivocal”).

McCleskey v. Kemp, 481 U.S. at 327 (Brennan, J., dissenting) (citing Baldus study).

Death Penalty Information Center, “Race of Defendants Executed Since 1976,” DPIC website, www.essential.org/dpic, accessed March 15, 2000.

David Cole, “Race, Life and Death,” Washington Post, January 11, 2000, p. A17.

McCleskey v. Kemp, 481 U.S. at 327 (Brennan, J., dissenting) (citing Baldus Study).

Richard C. Dieter, “the Death Penalty in Black & White: Who Lives, Who Dies, Who Decides,” Death Penalty Information Center (June 1998), p. 5 (citing Philadelphia study). See also David C. Baldus, George Woodworth, David Zuckerman, Neil Alan Weiner & Barbara Broffit, “Racial Discrimination in the Post-Furman Era: An Empirical and Legal Overview, with Recent Findings from Philadelphia,” 83 Cornell. L. Rev. 1638 (1998).

Staff Report by the Subcommittee on Civil and Constitutional Rights, House Judiciary Committee, “Racial Disparities in Federal Death Penalty Prosecutions 1988-1994,” 103rd Congress, 2d Sess. (March 1994), p. 2. As the Staff Report noted, “[a]lthough the number of homicide cases in the pool that the U.S. Attorneys are choosing from is not known . . . the almost exclusive selection of minority defendants for the death penalty, and the sharp contrast between capital and non-capital prosecutions under [21 U.S.C. §848] indicate a degree of racial bias in the imposition of the federal death penalty that exceeds even pre-Furman levels.”Id.


See “Prosecution and Race” at 34.

Our Members