June 27, 2001
Testimony Before the Committee on Rules and Administration, United States Senate
Russell Senate Office Building, Room 301
by Abigail Thernstrom
Mr. Chairman and members of the Committee, thank you for inviting me to testify today. My name is Abigail Thernstrom. I am a political scientist by training, a senior fellow at the Manhattan Institute in New York, and a member of the state board of education in Massachusetts, where I live. I am also a commissioner on the U.S. Commission on Civil Rights, a position to which I was appointed by Congress this past January.
I am the author of a multiple-award winning book, Whose Votes Count? Affirmative Action and Minority Voting Rights, and the co-author of America in Black and White: One Nation, Indivisible, a history of race relations and racial change in the decades since World War II.
I am delighted that the Committee is holding this hearing. The Election 2000 story—and particularly the allegation of black disfranchisement in Florida—is one that needs straightening out. Considerable misinformation has been floating about, and unfortunately, the U. S. Commission on Civil Rights has now contributed to it.
It has been well-established that the Florida election suffered from a variety of imperfections. But the only question before the Commission was this: Was the Florida election marred by discrimination that impeded the ability of American citizens, particularly African Americans, to freely and fairly exercise their right to vote?
Sadly, the very body charged with conducting an objective fact-finding investigation into this matter, the Commission, squandered its credibility by responding with a report that is highly partisan in tone.
The report rests heavily on a statistical analysis performed by Dr. Allan Lichtman of American University. We believe that his statistical analysis is deeply flawed, and we offer an alternative one that is broader and more sophisticated—one that reaches completely different conclusions. This was conducted by Dr. John Lott of Yale Law School, in consultation with Professor Stephan Thernstrom of Harvard University, who are both here today.
But before explaining further, let me say, that no statistical analysis of rates of ballots spoilage or voter error can be precise and certain, so we can’t bet the farm on any of them. Ballots are secret in the United States. We cannot know how any individual voted nor exactly who cast a ballot that did not count for some reason. Thus, we cannot be at all sure of the race or any other characteristics of voters whose ballots were rejected. We believe Dr. Lott has done a much better job than Dr. Lichtman, but we want to underscore the point that all of the numbers in dispute are mere estimates, and have a range of error that is unknown but might be very large.
As I read the evidence, the Florida election was hampered by unintentional and unanticipated problems that were not motivated by racial bias and did not disfranchise minority voters. The assertion that black voters were nine times more likely than whites to have their ballots discarded simply does not withstand close scrutiny.
And it should be noted that the Commission’s report does not even recognize the concept of voter error—as opposed to disfranchisement.
Using all the variables that Dr. Lichtman used or might have used (had he done a thorough analysis), Dr. Lott was not able to find a consistent statistically significant relationship between the share of voters who were African American and the ballot spoilage rate.
Furthermore, there appears to be little relationship between racial population change and ballot spoilage. That is, as the black population in a county went up in recent years, the rate of ballot spoilage did not—suggesting that race per se is not the explanation for ballots that did not count.
Indeed, in looking at the variation in ratios of ballot spoilage among counties, it is clear that race has very little explanatory value. Dr. Lott’s more sophisticated regression analysis suggests that three quarters of all the county variation in the rate of ballot rejection can be explained with equations that do not use racial information at all—that factor in, for instance, the poverty rate.
Dr. Lichtman also ignored the conflict between his statewide estimate that blacks were nine times more likely than whites to cast spoiled ballots, and his arguably more accurate precinct data that show racial disparities only one third as large.
The report blames the governor and secretary of state for spoiled ballots, but elections in Florida are the responsibility of the county supervisors, and all but one of the twenty-five counties with the highest rate of ballot rejection had supervisors who were Democrats—the one exception being an official with no party affiliation.
It is difficult to see how the local Democratic officials could have been tempted to suppress the black vote, which is almost entirely Democratic.
I could continue at some length with the discussion of the flaws in the majority report but I do not have the time to run through all the major findings in our 50-page dissent. Other flaws in the report, however, include:
- The almost complete omission of Hispanics—Florida’s largest minority group, as well as all other non-black minorities, such as Asians and Native Americans.
- The failure to look at the impact on ballot error rates of median income and percentage of the population living in poverty.
- A complete failure to distinguish between actual discrimination and bureaucratic problems—problems caused by bureaucratic inefficiencies, inexperienced voters, and other technical problems unrelated to race.
- A misinterpretation of the 1965 Voting Rights Act.
- A misleading analysis of errors in the “purge list” of convicted felons ineligible to vote. Their own data on the "purge list" show the opposite of what they claim: whites were erroneously placed on the purge list at almost twice the rate of blacks.
- Procedural irregularities that marred the production of the majority report. Most importantly, I asked for Dr. Lichtman's multiple regressions and the data that went into them. The Commission has said that they do not exist—that the data as he organized it for purposes of analysis is literally unavailable. If so, on what basis did he draw his conclusions? I assume that, in fact, Dr. Lichtman's statistical analyses are on the hard drive of his computer. If so, why the secrecy? Sharing the details of regression analyses is routine among scholars; they and their critics both benefit. Indeed, we have been the beneficiary of data recently sent to us by other scholars across the political spectrum.
Misinformation is dangerous. The Commission majority report positively sets us back in our progress on the long road to racial and ethnic equality. Real civil rights problems stir the moral conscience of Americans. Inflated rhetoric depicting crimes for which there is no evidence undermine public confidence in civil rights advocates and the causes to which they devote themselves. And, it increases the alienation on blacks from the political process. Do we really want black Americans to believe there is no reason to get to the polls; elections are always stolen; and they remain disfranchised? The answer is clearly no, because those conclusions are simply not true.