Statement of David C. Baldus

Prof. Baldus Statement Date: June 11, 2001 To: The Honorable Russell D. Feingold
Committee on the Judiciary, U. S. Senate
716 Hart Senate Office Building
Washington D.C. 20510-4904 From:

David C. Baldus
Joseph B. Tye Distinguished Professor of Law
College of Law, University of Iowa Re: DOJ report on the Federal Death Penalty System (June 6, 2001)

I have read U.S. Department of Justice, The Federal Death Penalty System: Supplementary Data, Analysis and Revised Protocols for Capital Case Review (June 6, 2001) ("the report"), which supplements the DOJ report of September 12, 2000. The following comments explain why in the face of the findings and data in the DOJ September 2000 report, the latest DOJ report utterly fails to convince me that there is no significant risk of racial unfairness and geographic arbitrariness in the administration of the federal death penalty. I believe there is still the just as much reason to be concerned about these issues as there was when the September 2000 report was issued.

1. The report completely overlooks the evidence of race-of-victim discrimination documented in the September 12, 2000 report.

A main theme of the latest report (p. 10) is that the death penalty authorization rate is higher for whites (.38) than it is for blacks (.25) and Hispanics (.20). These are the same figures that appeared in the September 2000 report. The latest report's emphasis on these statistics appears to suggest that white defendants are actually treated more punitively than minority defendants.

A more plausible explanation for the higher authorization rates for the white defendants is plainly documented in the September report ? (1) white defendants are more likely to have killed whites1 and (2) the U.S. Attorney charging and DOJ authorization rates are much higher in white-victim cases than they are in minority-victim cases. For example, data in the September 2000 report indicate that the Attorney General (AG) authorization rate for capital prosecutions is .37 (61/167) in white-victim cases and .21(81/383) in minority-victim cases -- a 16 percentage point difference that is statistically significant at the .001 level. The more punitive treatment of white-victim cases is a plausible alternative explanation for the higher authorization rates in white-defendant cases that the new DOJ report does not even recognize, let alone dispel.

The September 2000 report also documents race-of-victim disparities in the actual imposition of death sentences in the federal system. Among all death-eligible offenders, those data indicate that the death-sentencing rate from 1995 to 2000 is twice as high in white victim cases as it is in minority victim cases. Nationwide, the rates are .05 (10/198) for the white-victim cases versus .02 (10/446) for the minority-victim cases; in the eleven states in which death sentences were actually imposed, the rate in the white-victim cases was .17 (10/59) versus .08 (10/119) in the minority-victim cases -- a nine percentage-point difference.2

These are the same kinds of race-of-victim disparities documented in McCleskey v. Kemp .3 The latest report simply ignores the data on race-of-victim disparities in the charging and authorization process, and in the actual imposition of federal death sentences.

2. The report confounds the issue of "regional disparities" in the administration of the federal death penalty with the issue of racial disparities in the distribution of death eligible cases.

The report argues that we should not expect the proportions of black, white, and Hispanic offenders among death-eligible cases that are accepted for federal prosecution to correspond to "the racial and ethnic proportions in the general population." (p.13) Perhaps, but that is not the question. The real issue in this regard is the racial composition of the pool of death-eligible cases that are not accepted for federal prosecution. The report offers no data on that question. As a result, we do not know to what extent the death-eligible cases that were prosecuted in federal court are representative of all homicides that could have been charged as federal capital crimes, in the districts that are discussed in the report (pp.14-18) and in the country as a whole.

More importantly, the report seeks to equate its arguments concerning geographic disparities in the racial distribution of death-eligible cases with an explanation for clearly documented geographic and regional disparities in the administration of the death penalty. (Pp. 17-18) This is extremely misleading. The patterns that need to be studied are differences between regions in the rates at which death sentences are (a) sought by United State's Attorneys, (b) approved by the Attorney General, and (c) imposed by juries.

The September 2000 report clearly shows that in practice the federal death sentencing system is largely a Southern program. Twelve of the 19 men on federal death row as of September were sentenced in the South, including 6 from Texas and 4 from Virginia. The new report focuses on regional differences in the racial composition of the pools of potential capital cases that the districts have generated (p. 17). This has nothing to do with regional disparities in the rates at which death eligible defendants in the system are capitally charged and sentenced to death.

3. The report presents no data or other compelling reasons to dispel concerns about the exercise of discretion by U.S. Attorneys in the post-authorization stage of the process.

One the most striking findings of the September 2000 report is that in the period after the AG has approved a capital prosecution, 48% of white defendants avoid the risk of a death penalty by entering a plea agreement to a non-capital charge, while the rates that blacks and Hispanics enter such agreements are 25% and 28% respectively. (p.19) The department is obviously concerned about this issue because it plans to limit the power of U.S. Attorneys to enter such agreements without AG approval. (p. 22)

The report seeks to dispel concerns created by these data by pointing out first that it "takes two to make a plea agreement" and the data do not reflect racial differences in the rates at which the government offered post-authorization plea agreements. This argument raises an empirical question about the 62 cases (as of the September 2000 report) in which a post-authorization plea agreement was not reached. Was a plea bargain offered by the prosecution in these cases and rejected by the defense, or was none offered? It would have been easy for the DOJ to ask its own prosecutors whether they offered plea agreements in these cases. Apparently, it was not done.

The report further argues that even if differential acceptance rates by white and minority defendants did not explain the race disparities in the post-authorization guilty pleas, the September 2000 report's findings on this issue "would not be suggestive of bias by the U.S. Attorney's offices." (p. 20) The argument is that the detection of discrimination by U.S. Attorneys must rest on an analysis of "what happens in the process as a whole" and that decisions taken "at the final plea stage are uninformative as possible indications of bias by the U.S. Attorney offices." (p.20) Certainly it is important to view the system as a whole, but prior research demonstrates that race disparities may operate at discrete stages in a decision making process that overall appears to be evenhanded. There is serious cause for worry here, and the report makes no attempt to address it.4

The claim that no differential treatment exists in the post-authorization plea stage is a mere assertion with no evidence whatever to support it. Without data on the comparative culpability of the offenders (and the race of the victims) in the cases affected by these post-authorization pleas bargaining decisions, one has no idea the extent to which similarly situated defendants were in fact treated comparably.

4. The report provides no compelling reason for the DOJ's failure to authorize a comprehensive state of the art study of fairness in the administration of the federal death penalty system.

The report notes a meeting of "researchers and practitioners on January 10, 2001" in Washington D.C. to consider the feasibility of conducting a comprehensive empirical study and evaluation of fairness in the administration of the federal system. (p.11) I was one of the researchers at that meeting.

The report correctly states that there was general agreement at the January meeting that the conduct of such a study would entail a "multi-year research initiative." Two years would be the likely time line. In the meantime, half a year has passed since that meeting, and nine months since the release of the initial report, and neither the NIJ nor any other agency of the Department of Justice has taken any visible step to begin to make such a study possible. Quite the opposite. Attorney General Ashcroft's testimony last week suggested that he believes that the idea should be abandoned.
The report also states that "discussion" at the January 10 meeting "indicated," that such a study "could not be expected to yield definitive answers concerning the reasons for disparities in federal death penalty cases." This was certainly not the consensus of the researchers at the January 10 meeting. On the contrary, the consensus was that such a study would provide the best possible evidence on the question. Certainly the results of such a study would yield far more definitive answers to the issue of racial fairness in the system than the arguments presented in the department’s latest report.

The new report offers no reason at all why such a study should not be conducted even if it would require up to two years to complete. It also offers no reason why the DOJ appears unwilling to identify by defendant name and docket number the more than 700 death-eligible cases that make up the database for its latest study. With this information independent researchers could collect data on the cases in the DOJ database and conduct the kind of study that would provide the best evidence available on the question of fairness in the federal death sentencing system.

5. The report misconceives the nature of race discrimination in the administration of the federal death penalty.

A main theme of the report is that the core issue of racial fairness is whether U.S. Attorneys are consciously engaged in "favoritism towards White defendants." (p. 11) In other words, are their decisions based on "invidious" racial reasons (p.12) or motivated by "bias" (p. 20) or a "particular desire to secure the death penalty for minority defendants." (p. 17) This states the issue far too crudely. No one with an understanding of the system suggests that it is driven by such a conscious and blatant animus against minority defendants or defendants whose victims are white.

The concern about racial unfairness in the system is whether defendants with similar levels of criminal culpability and deathworthiness are treated comparably or differently because of their race or the race of their victims. The reasons for differential treatment by U.S. Attorneys - and by agents of the FBI, the DEA and other are federal law enforcement agencies - are almost certainly nonconscious. More importantly, the reasons for the differential treatment of similarly situated offenders on the basis of their race or the race of the victim are irrelevant. It is the fact that differential treatment cannot be explained by legitimate case characteristics that makes it morally and legally objectionable, when it exists. Without a systematic study based on full information concerning the criminal culpability and the race of the victims of all of the death eligible offenders, we will remain in the dark about whether unexplained differential treatment based on the race of the defendant and victim exists in the federal death penalty system, and if so, what causes it.

1. For the cases for which both race-of-defendant and race-of-victim data are available, 92% (109/119) of the white defendant cases involved a white victim.

2. The race-of-victim disparity nationwide is significant at the .06 level while the disparity in the states in which death sentences have been imposed is significant at the .09 level. The states in which death sentences were imposed between 1995 and 2000 are Arkansas, Georgia, Illinois, Kansas, Louisiana, Missouri, North Carolina, Oklahoma, Pennsylvania, Texas, and Virginia.
Of particular relevance are the race-of-victim disparities in case involving black defendants. Nationwide, in black defendant/white victim cases, the death-sentencing rate was .11 (6/55) while in the black defendant/minority victim cases, the rate was .03 (7/253), an 8 percentage-point difference significant at the .01 level. In the eleven death-sentencing states, the death-sentencing rate in the black defendant/white victim cases was .24 (6/25) while in the black defendant/minority victim cases, the rate was .07 (7/95), a 17 percentage-point difference significant at the .02 level.

3. 481 U.S. 279 (1987).

4. The report's argument also overlooks the fact that many of the post-authorization plea agreements are made in cases in which the U.S. Attorney's initial recommendation to waive the death penalty was overruled by the AG, a circumstance that needs to be factored into any analysis of the post-authorization decisions.

Return to Testimony, Resolutions, Statements, & Speeches on the Death Penalty