You like numbers?
Here's how to estimate the effect of spoilage on the election outcome. For fun, let's take Florida 2000. We know from comparison of census tracts to precincts that 54% of the 179,855 ballots “spoiled” were cast by African-American voters, that is, 97,000 of the total.
Every poll put the Black vote in Florida for Al Gore at over 90%. Reasonably assuming “spoiled” ballots matched the typical racial preferences, Gore lost more than 87,000 votes in the spoilage pile. Less than 10% of the African-American population voted for Mr. Bush, i.e. Bush lost no more than 10,000 votes to spoilage. The net effect: Gore had a plurality of at least 77,000 within the uncounted ballots cast by Black citizens.
OK, then, what about “Non-Black” voters, whose votes made up the remaining 46% of the spoilage pile? Well, frankly, you can ignore these, as these voters split their vote somewhat evenly between Gore and Bush. Sticklers wanting a closer exam would note that Gore probably won a majority of these votes as well. Moreover, the only large group of spoiled votes in a wealthy white county occurred in Palm Beach (due to “butterfly” ballots), a rare, rich white group of strongly Democratic voters.
Now you can dazzle your Republican friends with the real vote count — Gore won Florida by 77,000 in 2000. But they figure you already spend too much time concerned about issues like democracy and too little time concerned about guns and tax cuts.
GREG PALAST, adjunct professor of statistics at Indiana University (Gary) many, many years ago, is more recently the author of the New York Times bestsellers, The Best Democracy Money Can Buy and Armed Madhouse: Sordid Secrets and Strange Tales of a White House Gone Wild. www.GregPalast.com
You like numbers?