The New York Times
September 15, 2005
Section B, Page 2

At FEMA, Disasters and Politics Go Hand in Hand
By ALAN B. KRUEGER

"AS we are all aware," James Lee Witt told Congress in April 1996, when he was director of the Federal Emergency Management Agency, "disasters are very political events."

There is much truth to Mr. Witt's statement. Research on the spending patterns of the emergency management agency shows that, to a significant degree, the agency is influenced by political concerns that are distinct from the suffering and destruction wrought by natural disasters, under both Democratic and Republican administrations.

Although declaring a disaster should be clear-cut and above politics, the legislation that governs FEMA gives the president much discretion to decide whether an event is a disaster that qualifies for assistance. Upon receiving a request by a state's governor, the president may declare a "major disaster" if a natural catastrophe "causes damage of sufficient severity and magnitude to warrant major disaster assistance."

While no one would doubt that a disaster of the magnitude of Hurricane Katrina deserves the full commitment of the federal government, the language in the FEMA law is vague enough to count two feet of snow in Ohio as a major disaster, as was the case last December.

Indeed, the law specifically prohibits the use of an "arithmetic formula or sliding scale" to deny assistance. So, disaster requests are not evaluated based on standard quantitative evidence; instead, declarations involve subjective judgment.

Not surprisingly, in this vacuum presidents have displayed a tendency to declare more disasters in years when they face re-election. Mary W. Downton of the National Center for Atmospheric Research and Roger A. Pielke Jr. of the University of Colorado, Boulder, for example, looked at the flood-related disasters that were declared from 1965 to 1997 in an article published in "Natural Hazards Review" in 2001. Even after accounting for the amount of precipitation and flood damage each year, they found that the average number of flood-related disasters declared by the president was 46 percent higher in election years than in other years.

The tendency to declare more disasters during election years is not limited to floods. President Bill Clinton set a record by declaring 73 major disasters in 35 states and the District of Columbia in 1996, the year he was up for re-election.

When George W. Bush faced re-election in 2004, he declared 61 major disasters in 36 states - 10 more than in 2003 and tied for the second highest number of major disaster declarations ever, according to data provided by FEMA.

The increase from 2003 to 2004 was particularly sharp in the 12 battleground states in which the election was decided by 5 percent or less; these states had 17 major disasters declared in 2004 but only 8 in 2003, and, therefore, accounted for 90 percent of the increase.

In perhaps the most thorough econometric analysis of the politics of FEMA disaster funds to date, Thomas A. Garrett of the Federal Reserve Bank of St. Louis and Russell S. Sobel of West Virginia University concluded in a study published in "Economic Inquiry" in 2003 that Congressional as well as presidential politics play a role in allocating payments. Specifically, they examined factors related to the number of disasters declared in each state from 1991 to 1999, and the dollar amount of assistance provided.

Of course, a state may be particularly unlucky in any given year and be struck by an unusually large number of destructive storms. To account for this effect, they controlled statistically for the value of private insurance property claims from disasters in a state each year and the amount of Red Cross assistance provided. States that incur greater insurance losses and receive more aid from the Red Cross in a particular year are undoubtedly in greater need of emergency assistance.

It is reassuring to note that FEMA payments are substantially greater when the Red Cross provides more aid or private insurance losses are greater, suggesting that more generous FEMA payments do reach states that suffer more severe damage from disasters.

Nonetheless, Mr. Garrett and Professor Sobel conclude that Congressional and presidential politics play an important role in allocating FEMA assistance. Most important, they find that the amount of disaster relief provided per incident increases with the number of representatives a state has on one of the FEMA oversight committees in the House of Representatives. Having an additional representative on either of the two main House oversight committees is associated with an extra $36.5 million of assistance from FEMA, they found.

This figure may seem to overstate the role of politics because representatives from states prone to be hit by disasters probably seek out seats on FEMA oversight committees. But recall that the analysis simultaneously adjusts for the amount of Red Cross assistance and private insurance losses from disasters each year. So, having Congressional representation on an oversight committee appears to matter even when compared with disasters in other states that cause roughly the same amount of damage and suffering.

Interestingly, representation on a Senate oversight committee does not have a detectable effect on FEMA payments, perhaps because disasters are often local events that more intensely affect House members' constituents.

All told, Mr. Garrett calculates that a third of FEMA payments are directly attributable to representation on one of the nine FEMA Congressional oversight committees, independent of the disaster's severity.

When Congress turns to evaluating the Hurricane Katrina disaster, it should also consider sharpening FEMA's mission. One helpful step would be to define more precisely the requirements necessary for the president to declare a disaster. For example, disaster payments could be restricted to events that cause damage exceeding a specified threshold or significant loss of life. Long before Hurricane Katrina, it should have been apparent that FEMA needed to focus more on alleviating and preventing suffering from major catastrophes and less on delivering pork to voters at election time.

Alan B. Krueger (www.krueger.princeton.edu) is the Bendheim professor of economics and public affairs at Princeton University.


Source: http://www.nytimes.com/2005/09/15/business/15scene.html?pagewanted=print