As the COVID-19 crisis was exploding in late March 2020, Michigan Gov. Gretchen Whitmer claimed that medical supply “vendors are being told [by the federal government] not to send stuff here to Michigan.” She asserted that her state had “been uniquely singled out” because of her criticizing President Trump’s response to the pandemic.

We don’t know if her allegation is correct. However, a recent study does suggest that states that supported Donald Trump in the 2016 election appear to have received small-business loans as a percentage of eligible payroll that were much larger than the coastal states that he lost. Rural Midwestern states like the Dakotas and Nebraska got funds that covered 70% or more of their eligible payroll. The comparable numbers for New York and California were 40% and 38% respectively, or about half of what the Trump-supporting states got.

For many, such favoritism is seen as a unique feature of the Trump administration. However, the politicization of resource allocation in a crisis is nothing new. Here we provide a brief overview of research finding that political considerations influence the allocation of aid during crises.

The New Deal / Economic historian Gavin Wright examined the disparate distribution of loans and relief during the New Deal era with an eye to explaining the generous aid directed to western states and the paucity of aid flowing to the South. Wright used simple cross-state regression models to compare the influence of economic variables (e.g., unemployment) and political variables on both the allocation of spending and work relief jobs across states. His results suggest that political factors contributed to the pattern of relief and he points to the high variability of Democratic voting in western states as a key factor in the generous level of aid directed to those states. The implication is that the Roosevelt administration steered aid to states where it would be of most use in getting the president reelected. By contrast, the South had a strong tendency to vote Democratic so there was less need to steer aid to those states.

Wright was not the only one to make this observation. Robert Fleck revisited the cross-state allocation of aid and concluded that “electoral variables do matter [in the allocation of New Deal spending] and their influence is substantial.” Disaster aid can be understood as a way to buy votes in states where the marginal voter matters for electoral success. Politicians are smart enough not to waste the opportunity to use disaster aid this way.

Gary Anderson and Robert Tollison also analyzed New Deal–era spending with an eye on congressional influences such as budget allocations and oversight. They estimated several models that included both measures of economic hardship and political influence in each state. They found that “spending went partly to the needy and partly to those with political clout.” In particular, their results showed that states with lawmakers on congressional appropriations committees received larger amounts of spending.

Jim Couch, Keith Atkinson, and William Wells took a more granular look at New Deal aid, focusing on the allocation of agricultural aid in Alabama. They documented that average spending per farmer across Alabama counties ranged from $16 to $90 and found that aid per farmer was positively correlated with Roosevelt’s vote share, the vote share of pro–New Deal Gov. Bibb Graves, and being represented on the agriculture committee in the Alabama legislature.

Disaster declarations / In recent decades, the federal government has assumed a bigger role in responding to natural disasters, largely through the Federal Emergency Management Agency. As explained by Thomas Garrett and Russell Sobel, political influence can affect FEMA aid through both presidential disaster declarations and relief spending authorized by Congress. To analyze those potential influences, Garrett and Sobel examined FEMA disaster spending across all 50 states for the period 1991–1999. While acknowledging that some disaster spending in the 1990s was clearly tied to events such as 1992’s Hurricane Andrew, 1993’s Mississippi River flooding, and 1994’s Northridge, CA earthquake, Garrett and Sobel found that disaster declarations increased in election years (with a larger increase in 1996 than 1992), that disaster declarations were more frequent in electorally important (swing) states, and that states with greater representation on subcommittees overseeing FEMA received higher disaster relief allotments. Taking the increases in disaster declarations and FEMA spending together, they concluded that “nearly half of all FEMA disaster relief is explained by political influence rather than actual need.”

Unlike his predecessors, Trump may be more likely to let his feelings of personal offense override his political self interest, so Michigan may get short-changed.

Andrew Reeves also looked at presidential disaster declarations. Examining the period 1981–2004, he found that declarations increased in election years and that electorally competitive states received more disaster declarations than uncompetitive states. Reeves also looked into the effect of declarations on voter behavior and concluded that each additional declaration led to a one percentage point bump in vote share.

The federal government also makes agricultural disaster payments to areas experiencing crop failures because of droughts, floods, or other conditions. Garrett, Thomas Marsh, and Maria Marshall analyzed agricultural disaster payments made within the 48 contiguous states between 1992 and 1999. After controlling for factors such as precipitation changes that might lead to floods or droughts, they found that states represented on relevant subcommittees of the House and Senate Agriculture committees as well as the House subcommittee overseeing agricultural appropriations received higher levels of agricultural disaster relief. The authors then used their estimation results to calculate that as much as 30% of agricultural disaster relief payments were allocated based on political influence rather than crop losses.

The analysis most pertinent to the current COVID disaster is Matt Ryan’s study of H1N1 vaccine allocations. The H1N1 (swine flu) virus spread across the United States in early 2009, but fortunately a vaccine for it was developed by late summer. The first doses were available in October 2009, but initially were in limited supply. Vaccines were distributed by the Department of Health and Human Services, and the House Committee on Energy and Commerce has oversight authority over HHS. Ryan’s results indicate that states received an additional 60,000 doses of vaccine for each additional Democratic member (the majority party in 2009) the state had on the House Energy and Commerce Committee. No such influence was found for states represented on the Senate Committee on Health, Education, Labor, and Pensions, which also has oversight over HHS.

Conclusion / To reiterate, we don’t know for certain whether Michigan or any other state has been shortchanged in the current crisis because of political favoritism. However, past findings about electorally important states suggest that, if anything, Michigan is likely to be treated well; it is expected to be an important swing state in the 2020 presidential election. However, unlike his predecessors, Trump may be more likely to let his feelings of personal offense override his political self-interest, so Michigan may indeed get shortchanged as a result of Whitmer’s criticisms of the Trump administration.

Existing research repeatedly shows that political factors influence resource allocation in disasters, so it should come as no surprise if it is ultimately determined that politics affected the response to the COVID-19 crisis. These outcomes are not the product of particular personalities or partisan affiliations, but of the incentives created by the institutional structure of politics. Politicized disaster relief is nothing more than politics as usual.

Readings

  • “Allocating Infection: The Political Economy of the Swine Flu (H1N1) Vaccine,” by Matt E. Ryan. Economic Inquiry 52(1): 138–154 (2014).
  • “Congressional Influence and Patterns of New Deal Spending, 1933–1939,” by Gary M. Anderson and Robert D. Tollison. Journal of Law and Economics 34(1): 161–175 (1991).
  • “Inter-Party Competition, Intra-Party Competition, and Distributive Policy: A Model and Test Using New Deal Data,” by Robert K. Fleck. Public Choice 108(1–2): 77–100 (2001).
  • “New Deal Agricultural Appropriations: A Political Influence,” by Jim F. Couch, Keith E. Atkinson, and William H. Wells. Eastern Economic Journal 24(2): 137–148 (1998).
  • “Political Allocation of U.S. Agriculture Disaster Payments in the 1990s,” by Thomas A. Garrett, Thomas L. Marsh, and Maria I. Marshall. International Review of Law and Economics 26(2): 143–161 (2006).
  • “Political Disaster: Unilateral Powers, Electoral Incentives, and Presidential Disaster Declarations,” by Andrew Reeves. Journal of Politics 73(4): 1142–1151 (2011).
  • “The Political Economy of FEMA Disaster Payments,” by Thomas A. Garrett and Russell S. Sobel. Economic Inquiry 41(3): 496–509 (2003).
  • “The Political Economy of New Deal Spending: An Econometric Analysis,” by Gavin Wright. Review of Economics and Statistics 56(1): 30–38 (1974).