Completing the hand recount of ballots now underway in southeast Florida would lead to a less accurate vote count in the state. That would be true even if those conducting the hand recount are conscientious and strictly impartial.
This result, which could determine our next president, is the consequence of conducting the hand recount only in those large counties in which the vote, as measured by the machine counts, strongly favored Vice President Gore. The only alternatives to assure that the electors chosen by the State of Florida reflect the preferences of a majority of voters in that state are to reject the results of this selective hand recount or to conduct a time-consuming hand recount for the entire state.
Consider the following example: Assume that the probability of an "undervote" by those who prefer Vice President Gore was the same as by those who prefer Gov. Bush. And assume that those conducting the hand recount can accurately determine the preferences of those voters whose vote for president was not registered by the machine count and recount. In that case, the proportion of the undervotes added to the vote for Gore will be the same as the proportion of the machine count of the votes for Gore. This will not change the proportion of the votes for Gore in these counties but will increase the vote for Gore in the state totals.
Moreover, assume that those who conduct the hand recount proportionally allocate all of the estimated undervotes to either Gore or Bush. In that case, the change in the vote for Gore and Bush in the three southeast counties would be as follows:
County | Undervotes | Gore | Bush | Net for Gore |
Miami-Dade | 10,000 | 5,318 | 4,682 | 636 |
Palm Beach | 10,000 | 6,376 | 3,624 | 2,752 |
Broward | 6,000 | 4,113 | 1,887 | 2,226 |
Total | 26,000 | 15,807 | 10,193 | 5,614 |
In this example, the vote for Gore in the state totals would be increased by 5,614, overwhelming the small margin for Bush in the machine counts and the expected vote difference in the absentee ballots. This would be the general result even if those who conduct the hand recount allocate only a small proportion of the total undervotes to Gore and Bush.
One should not be surprised that the Gore campaign chose to focus its demand for a hand recount of Palm Beach County, the one county with the greatest potential for changing the state-wide vote balance. There is more reason to be surprised that James Baker, speaking for the Bush campaign, made his case against the hand recount primarily on the basis that it would be more subjective and vulnerable to partisan mischief. Baker's argument, however accurate, was both unnecessary and distracted attention from a stronger case against the hand recounts: A strictly impartial hand recount in a biased set of counties would lead to a bias in the statewide vote totals.
Unless the State of Florida and the contesting parties are prepared to bear the cost and time of a statewide hand recount, the selective hand recounts in the southeast Florida counties should be dismissed out of hand. In that case, the State of Florida should assign its Electoral College votes to the major party candidate with a majority of the votes in the machine recount plus those in the absentee ballots.
—William A. Niskanen, an economist, is chairman of the Cato Institute.
