Large areas of northern Oklahoma were hit by severe storms and flooding in the first two weeks of June 2008. Property damage was estimated to total more than $759 million, an astounding $198 for every resident of Oklahoma. (1) That same summer Missouri also experienced severe storms and flooding. Property damage totaled $17 million, or about $3 per capita. Crop damage totaled an additional $58 million, bringing the total damage to roughly $14 per capita. Counties in both states were declared to be "federal disaster areas" by President George W. Bush, making affected jurisdictions eligible for federal reimbursement of 75% for eligible recovery and rebuilding expenses. President Bush increased the federal cost share for Missouri to 90% of eligible costs a few weeks after the flooding while Oklahoma did not receive this extra assistance.
President Bush's decision to increase the federal reimbursement percentage sent an extra $8 million to Missouri. Why did Missouri receive more favorable treatment than Oklahoma? The Oklahoma storms had a much greater financial impact on the state, and they definitely could have used the additional assistance they would have received in reimbursement at the 90% level. Partisan politics provides a plausible explanation. Missouri was widely expected to be a pivotal swing state in that fall's presidential election, and this proved true with only 3,903 votes separating John McCain and Barack Obama. There was no danger of Republicans losing Oklahoma, which McCain ended up carrying with a 31% margin over Obama.
I study the effects of presidential electoral politics on the federal government's financial response to disasters. Specifically I ask whether states supportive of the president or "swing states" are more likely to receive additional disaster aid through presidentially ordered increases in the federal reimbursement rate after specific disasters. To answer this question I construct an original data set including information about presidential decisions regarding disaster aid for 945 disasters spanning 19 years and three presidential administrations (Clinton, George W. Bush, and Obama). I find that presidents are more likely to send additional aid to swing states and most significantly that the presidents in this study used their power in different ways, likely due in part to contrasting electoral strategies. These findings demonstrate the presence of partisan political calculations in the distribution of disaster aid, a power delegated to the president by Congress precisely to avoid such occurrences.
This article begins with a theoretical overview of distributive politics and the executive branch. Two theories are borrowed from the congressional literature as ways in which policy makers may wield policy power in a partisan way. Swing voter theory expects policy makers to focus on swing states while coalition maintenance expects policy makers to focus on rewarding their base states as a means of ensuring high turnout. Additionally, the impact of co-partisan presidents and governors, as well as proximity of the next presidential election, are tested as additional partisan factors with the potential to influence the distribution of disaster aid.
Theorizing about Presidents and Distributive Politics
Weingast, Shepsle, and Johnsen (1981) define distributive politics as policy areas with "project-by-project orientation, the geographic concentration of benefits, and the diffusion of costs" (681). In this article distributive politics implies disproportionate benefits relative to the burdens to geographic constituencies. This fits the nature of disaster relief, as disasters are inherently local or regional events, yet the federal government routinely provides relief and is expected to do so.
Distributive Politics and Congress
The vast majority of the early work related to distributive politics focuses on the ability of senators and representatives in Congress to direct funds to specific constituencies. Two basic theories developed. Swing voter theory indicates that parties (or legislators) will target marginal constituencies, defined as the states or districts most likely to change hands in an upcoming election (Bickers and Stein 1996; Lindbeck and Weibull 1987; Stein and Bickers 2009). The other often-tested theory posits that risk-averse incumbents will direct attention toward their strongest supporters and will be referred to as coalition maintenance (Cox and McCubbins 1986). (2)
Distributive Politics and the Executive Branch
Evidence of geographic distributive politics in the executive branch was first put forth by Wright (1974) who found that during the New Deal, Works Progress Administration jobs went disproportionately to electorally competitive states. More recent work has continued to investigate the prevalence of distributive politics within executive agencies. Mebane and Wawro (2002) found a modest increase in federal expenditures for basic programs toward supporters after midterm elections (coalition maintenance). Berry, Burden, and Howell (2010) found that congressional districts transitioning from being represented by a member of the opposition to a member of the president's party enjoy increased federal expenditures and Gersen and Berry (2010) find this affect to be larger for agencies with a higher proportion of political appointees (maintenance of a new coalition).
The above research indicates that executive agencies appear to use distributive politics for partisan advantage but does not put forth evidence that these decisions are being made in the White House. Gersen and Berry (2010) as well as Gordon (2011) indicate that political appointees likely play a role, but these appointees may be acting on explicit or implicit direction of the president or simply demonstrating their own partisan biases. This article is focused on presidential influence in this process. The next section identifies specific methods by which the president may exert this influence.
Direct Presidential Distributive Powers
Although Congress holds the "power of the purse," the president and executive branch wield considerable power over distributive allocations (Fisher 1975). As the leader of the executive branch, the president has the ability to influence agency decision making through executive orders (Mayer and Price 2002) and involvement in the rule-making process (DeMuth and Ginsburg 1986; Krent 2005). For example, Presidents John E Kennedy, Lyndon Johnson, and Richard Nixon all used this power to require presidential approval of all U.S. Agency for International Development distributions over $5 million (McKeown 2005).
More recently Congress has taken to delegating power to the president and executive agencies as the demands on its own time increase. Throughout the 1990s and 2000s, this has been most apparent in the areas of Medicaid (Thompson and Burke 2007) and environmental policy (Bowers 2010). The issuance of Medicaid waivers is the responsibility of the secretary of Health and Human Services, but the authority to grant waivers of environmental regulations varies from the Environmental Protection Agency administrator to the president himself. (3) The issuance of waivers and other administrative decisions delegated to the executive branch by Congress are the areas ripest for presidential involvement in distributive politics. This power is likely to be relatively unchecked, as there is a reason Congress delegated this power in the first place. These decisions may be political risky (Arnold 1992), too technical for Congress to handle given its limited time (Bowers 2010), or require a speedy response to an unexpected situation (Wamsley and Schroeder 1996).
Presidential Power and Disaster Response
The disaster relief system that has developed over the past 60 years is relatively unique and reflects the politically sensitive, technical, and time-critical elements of disaster response. Since 1950, presidents have possessed the power to issue national disaster declarations that then allow states to apply for federal reimbursement for a percentage of the losses caused by the disaster. The funds for this aid are appropriated by Congress into the Presidential Disaster Relief Fund and are not earmarked for any specific disaster (Sylves 2008). Congress makes "deposits" into this account and the Federal Emergency Management Agency (FEMA) distributes the funds as reimbursement for qualifying expenses in states and counties designated as federal disaster areas.
The Disaster Relief Act of 1950 (EL. 81-875) limited the use of federal funds for public assistance to the repair of public structures
(no rebuilding and no private structures). Over the course of the last 60 years, the scope of this aid has gradually increased to cover almost any disaster-related damages (Sylves 2008). Although the scope has increased considerably the process is unchanged. After a disaster strikes, state and local governments are responsible for the immediate response. Once the extent of the damage is known the governor, with the assistance of local FEMA officials, requests a national disaster designation from the president. (4) Further information is compiled by FEMA at the national level, and then a recommendation is made to the president if the disaster declaration should be approved or denied. If the declaration is approved, the state is then eligible for federal reimbursement of 75% of eligible expenses. (5)
Others have examined disaster response for political motivations and consequences with mixed results. This research into the politicization of disasters has exclusively focused on the issuance of presidential disaster declarations. Garrett and Sobel (2003) find that states that are more electorally important are slightly more likely to receive more declarations, and that more declarations are issued in election years. In contrast, Sylves and Buzas (2007) examined an extended period of time and did not find...