Bush, Vote Buying and the 2004 Elections

Did George W. Bush learn not to repeat his father's mistakes?

Two months before the 1992 presidential election, the Federal Emergency Management Agency was caught off-guard as Hurricane Andrew ravaged parts of Florida. The situation turned into a public relations fiasco for Bush the senior and some pundits blamed his electoral loss partially on FEMA's bungled effort.

Fast forward 12 years.

The hurricane season of 2004 was one of the worst on record with four strong storms hitting Florida, killing 117 people and causing $21 billion in damages. This time around FEMA was ready. Perhaps a little too ready. The agency doled out some $800 million in aid before the elections including $31 million to residents in largely unscathed Miami-Dade County. But the next year, FEMA was censured by Department of Homeland Security for failing to perform any inspections to ensure damages before handing out the money .

The Bush administration was accused of distributing excess funds in order to gain an edge in the all-important battleground state, which it won in November. Now a new study presents some startling evidence showing that FEMA's largess was directed specifically at Republican precincts.

Using geographical data from 2.6 million aid applications, Jowei Chen, a PhD candidate at Stanford University, identified which voting precinct each one of the applications came from. Then, based on Governor Jeb Bush's vote share in the 2002 gubernatorial election, Chen tagged each precinct as either Republican, Democrat or moderate.

Chen found that, keeping all else equal including the severity of damage and household income, a relief applicant from a Republican precinct received 19 percent more aid than an applicant from a moderate precinct and 40 percent more aid than an applicant from a Democratic precinct.

"While it is impossible to prove whether the Bush administration acted with electoral motives in mind," Chen wrote. "FEMA's overwhelmingly pro-Republican bias in its distribution of hurricane disaster aid is consistent with a strategy of buying the votes of core partisan supporters."

(This 2005 story from the Sun-Sentinel, however, provides evidence that "electoral motives" were a concern for Bush associates.)

Chen calculates that Republican precincts received $200 million more in aid than they would have had no targeting taken place. But did these extra payments to Republicans change the outcome of the 2004 elections? It's highly unlikely.

Chen estimates that 19,000 extra Republican voters turned out that November because of the payments, but Senator John Kerry lost Florida by some 380,000 votes.

If nothing else, the study demonstrates that when it comes to capturing votes on election day, a candidate is better off targeting his or her base rather than swing voters. Those who were approved for aid and were registered with the Republican party were much more likely to vote in November than those who were approved but had no party affiliation.

But in the end the strategy -- if there was one -- was very costly. Chen estimates that it cost about $17,000 in FEMA aid to add one new Bush voter in a Republican neighborhood. That means that the 19,000 extra votes cost over $300 million.