Fahrenheit Fact no. 32: 20 Counties ignored voter removal list

One of the segments in "Fahrenheit 9/11" concerns the fiasco surrounding the removal of convicted felons from the voter rolls in Florida. From the film:
NARRATOR: Second, make sure the chairman of your campaign is also the vote count woman. And that her state has hired a company that's gonna knock voters off the roles who aren't likely to vote for you. You can usually tell 'em by the color of their skin.
Dave Kopel writes:
"According to Fahrenheit, Bush cronies hired Data Base Technologies to purge Florida voters who might vote for Gore, and these potential voters were purged from the voting rolls on the basis of race. ("Second, make sure the chairman of your campaign is also the vote count woman. And that her state has hired a company that's gonna knock voters off the rolls who aren't likely to vote for you. You can usually tell 'em by the color of their skin.") As explained by the Palm Beach Post, Moore's suggestion is extremely incomplete, and on at least one fact, plainly false. The 1998 mayoral election in Miami was a fiasco which was declared void by Florida courts, because--in violation of Florida law--convicted felons had been allowed to vote. The Florida legislature ordered the executive branch to purge felons from the voting rolls before the next election. Following instructions from Florida officials, Data Base Technologies (DBT) aggressively attempted to identify all convicted felons who were illegally registered to vote in Florida. There were two major problems with the purge. First, several states allow felons to vote once they have completed their sentences. Some of these ex-felons moved to Florida and were, according to a court decision, eligible to vote. Florida improperly purged these immigrant felons. Second, the comprehensive effort to identify all convicted felons led to large number of false positives, in which persons with, for example, the same name as a convicted felon, were improperly purged. Purged voters were, in most cases, notified months before the election and given an opportunity to appeal, but the necessity to file an appeal was in itself a barrier which probably discouraged some legitimate, non-felon citizens from voting. According to the Palm Beach Post, at least 1,100 people were improperly purged."
The voter roll removals seem to have been more aggressively pursued than was necessary, resulting in a large number of people who were not allowed to vote, despite not being convicted felons. But the key things to keep in mind is that the problem was well-known months before the election, and many (if not most) of those people improperly purged were able to appeal the decision. Also 20 counties completely ignored the felon list entirely. The American Sociological Review notes that felons vote approximately 69% democratic. Thus, allowing them to vote in these 20 counties most likely increased Gore's overall vote count. But how many felons voted? The dissenting opinion of the US Civil Rights Commission (pdf) states that "Approximately 5,600 felons voted illegally in Florida on November 7.". Assuming that the 69% figure holds true, then somewhere around 3,864 democratic votes were garnered for Gore by allowing felons to vote. This number, however, is pure conjecture, and is completely based on the 69% figure given by the American Sociological Review. The point is that in a race decided by somewhere around 500 votes, allowing 5,600 felons to vote undoubtadly swayed the election. As for the comment about "you can know them by the color of their skin", Dave Kopel shows us that race did not play into the infamous voter removal list:
Regardless, Moore's suggestion that the purge was conducted on the basis of race was indisputably false. As the Palm Beach Post details, all the evidence shows that Data Base Technologies did not use race as a basis for the purge. Indeed, DBT's refusal to take note of a registered voter's race was one of the reasons for the many cases of mistaken identity. DBT's computers had matched these people with felons, though in dozens of cases they did not share the same name, birthdate, gender or race...[A] review of state records, internal e-mails of DBT employees and testimony before the civil rights commission and an elections task force showed no evidence that minorities were specifically targeted. Records show that DBT told the state it would not use race as a criterion to identify felons. The list itself bears that out: More than 1,000 voters were matched with felons though they were of different races. The appeals record supports the Palm Beach Post's findings. Based on the numbers of successful appeals, blacks were less likely to have been improperly placed on the purge list: of the blacks who were purged, 5.1 percent successfully appealed. Of Hispanics purged, 8.7 percent successfully appealed. Of whites purged, 9.9 percent successfully appealed. John R. Lott, Jr., "Nonvoted Ballots and Discrimination in Florida," Journal of Legal Studies, vol. 32 (Jan. 2003), p. 209. Of course it is theoretically possible that the appeals officials discriminated against blacks, or that improperly purged blacks were not as likely to appeal as were people of other races. But no one has offered any evidence to support such possibilities.
In regards to various recount scenarios, Dave Kopel writes:
As USA Today summarized, on May 11, 2001: Who would have won if Al Gore had gotten manual counts he requested in four counties? Answer: George W. Bush." "Who would have won if the U.S. Supreme Court had not stopped the hand recount of undervotes, which are ballots that registered no machine-readable vote for president? Answer: Bush, under three of four standards." "Who would have won if all disputed ballots � including those rejected by machines because they had more than one vote for president � had been recounted by hand? Answer: Bush, under the two most widely used standards; Gore, under the two least used."
The New York Post has an opinion column that addresses what black voters experienced. Among their analysis, we find this:
"The error rate was 9.9 percent for whites, 8.7 percent for Hispanics, and only a 5.1 percent for African-Americans."
Also, Moore fails to note that of the 25 counties with high "vote spoilage" (votes not counted for lack of clarity), 24 had democratic election officers in charge. RealClearPolitics reports:
Lott found that among the 25 Florida counties with the greatest rate of vote spoilage, 24 had Democratic election officers in charge of counting the votes. He concluded that "having Democratic officials in charge [of county elections] increases ballot spoilage rates significantly, but the effect is stronger when that official is an African-American." Commissioner Abigail Thernstrom argued that "it is very difficult to see any political motive that would lead Democratic local officials to try to keep the most faithful members of their party from the polls and to somehow spoil the ballots of those who did make it into the voting booth."
(All emphasis in quoted articles mine) -a_sdf