« The Afghan Election: Is Voting Really Not a Paradox? | Main | Does George W. Bush Care about Black People? »

21 September 2005

Electoral politics and FEMA's poor performance in the aftermath of Hurricane Katrina

The aftermath of Katrina has brought about a critical examination of the Federal Emergency Management Agency (FEMA). Many commentators have suggested that only the military is equipped to handle the aftermath of natural disasters (see a recent article by Mark Sappenfield in the Christian Science Monitor). This is an interesting observation since prior to the end of the Cold War (up until 1993), FEMA directors typically had military backgrounds and the agency was often criticized for its preoccupation with military objectives (see the 1994 Mother Jones article by Ted Grup, "How the Federal Emergency Management Agency learned to stop worrying -- about civilians -- and love the bomb". )

Fast forward to 2005 and the aftermath of Katrina. How did FEMA stop worrying about the bomb, and what exactly does it love these days?

Some say that FEMA stopped worrying about the bomb when President Clinton appointed James Lee Witt to head and reorganize FEMA. Although Witt is generally praised for his management of FEMA, he along with Clinton were criticized that the agency was being converted to a reelection slush-fund to essentially buy votes (see " Money! Come & Get It!; Reinventing disaster, Bill Clinton has turned the decrepit FEMA agency into another arm of his permanent campaign" from The American Spectator by James Bovard).

A paper that I've been working on sheds some light on this question. In it, I examine presidential disaster declarations from 1981 through 2004 (note that Katrina is not included). Among the bullet points (so bulleted) are:


  • there is a relationship between the electoral importance of a state (as measured by the competitiveness of previous statewide presidential contests and the number of electoral votes) and how many presidential disaster declarations a state receives;
  • this relationship is present from 1993-2004, but not from 1981-1992; and,
  • voters reward presidents at the ballot box for disaster declarations throughout the entire period.

Reassuringly, the very best predictor of whether a state gets a disaster declaration is actual need (measured using data from a private company that collects insurance data). But the analysis suggests that Clinton and W. Bush have allowed their presidential disaster declarations to be shaded by electoral concerns.

The figure below displays presidential disaster declarations and private disaster declarations (a measure provided by the ISO and based on insurance data -- see the paper for more details).

pres.dis.jpg

What if Clinton had maintained (or W. Bush restored) the military leadership of FEMA but (somehow) restructured the mission of FEMA? Would the federal response to Katrina have been better? I only examine presidential disaster declarations -- but would Louisiana have received more preventative disaster aid had it been more electorally important (or if only need were considered)?

Posted by Andrew Reeves at September 21, 2005 11:41 AM