- The Washington Times - Tuesday, July 31, 2001

Were votes by black Americans systematically disallowed during the 2000 elections? Mary Frances Berry, the chairman of the U.S. Commission on Civil Rights, told the press in early June that this indeed occurred. Miss Berry even felt it necessary for the Justice Department to investigate whether the discrimination was intentional.

It is hard to think of more inflammatory charges. The civil rights commission's Democrats asserted that "Statewide, based upon county-level statistical estimates, African-American voters were nearly 10 times more likely than white voters to have their ballots rejected." In addition, blacks were said to be erroneously included on a list of voters who were disqualified from voting because of a felony conviction at a greater rate than whites.

So-called "non-voted" ballots are ballots where either no vote was recorded or votes are made for multiple candidates. The Democrats' claim is simple: Counties or precincts with a higher percentage of blacks have a higher percentage of non-voted ballots, and this is interpreted as discrimination.

At the request of the commission's minority members, Russell Redenbaugh and Abigail Thernstrom, I agreed to examine the report. What I found contradicted the majority's claims.

The majority relied upon regressions, a statistical process that allows one to separate how much of the change in non-voted ballots could be correlated with different factors. This is exactly what one should do. However, the report never provided the precise analysis, only their conclusions. Despite repeated requests from the commission's minority members, the minority has still not been provided with the regression analysis, even though a month-and-a-half has passed since the report was first leaked to the press. The commission has refused to tell a House oversight committee how they had produced their estimates.

In fact, comparing counties in the Florida 2000 elections, the period of time examined by the commission, reveals that estimates showing that more black voters result in more non-voted ballots are quite sensitive to the specification used. When I also examined data from the 1992, 1996 and 2000 presidential elections, the share of voters who were black was never an important factor. Even with the estimates most supportive of the majority's claims, I found no more than 3 percent of the change in spoiled ballots could possibly be explained by the share of voters who were black.

Yet, even in the selected regressions suggesting such a slight relationship, there was a surprising twist: The link between the percent of voters who were black and non-voted ballots was stronger when the county election supervisor was a Democrat than a Republican. This more or less proves that any possible amount of bias against blacks cannot have been the result of intentional discrimination; for it is not plausible that black, Democratic election supervisors would consciously keep fellow blacks, who are also overwhelming Democrats, from voting.

Now to the commission's "evidence" of blacks more often being erroneously included on the ineligible voter list. The majority's own evidence shows that this claim is incorrect, and the evidence is irrelevant for drawing the conclusions that they want from it. The evidence that blacks win a greater share of appeals does not account for the fact that blacks make up an even much greater share of the list of ineligible voters to begin with. In fact, the rate that whites are removed from the list because they were incorrectly included to begin with is almost twice the rate of blacks (9.9 percent to 5.1 percent).

In any case, examining only those who successfully appealed says nothing about how many people who didn't appeal, but could have successfully done so.

Finally, the majority has consistently acted to stifle debate. First, the commission report was leaked to the New York Times before minority Republican members were even given a copy. Then the majority refused to provide their data to the minority before a Senate hearing last month. As noted earlier, they are still unwilling to explain how they calculated their results. Despite promises to the Senate that the minority report would be disseminated by the commission, the majority decided two weeks ago to allow only the portions of the minority report that omitted any references to the research that I had done.

Because the majority refused to provide the minority commissioners with any resources for a statistical analysis, I agreed to do such work without compensation. Despite numerous precedents to the contrary, the majority now argues that only work by consultants paid by the commission can be included in any dissent. Amazingly, the minority is not even allowed to cite my work regarding the Florida elections. How then is the minority supposed to be able to produce any contrary evidence? What is the majority afraid of?

John R. Lott Jr. is a senior research scholar at the Yale University Law School, and his work on the Florida elections can be obtained at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=276276.



Click to Read More

Click to Hide