Does Presidential Primary and Caucus Order Affect Policy? Evidence from Federal Procurement Spending

Article excerpt

This article examines the contention that states with early presidential primaries or caucuses receive disproportionate distributive policy benefits. The basic theory is that presidential candidates pledge more federal spending per capita to these states because doing well in their contests is critical to capturing the nomination. Candidates then deliver on these promises if they win the White House. Using by-state procurement per capita data from 1984 to 2004, four conditional hypotheses derived from this thinking are tested. The results show that primary or caucus order matters only during competitive nominations when the ultimately victorious presidential candidate won the state's contest.

Keywords: presidency research; elections and voting behavior

There is a great deal of evidence to show that the results of the early state contests for delegates to national party conventions do much to determine who wins presidential nominations. A number of scholars argue that victory or strong performances in the first primaries and caucuses, especially those in Iowa and New Hampshire, can provide momentum that elevates a candidate in public opinion polls, generates favorable media coverage, and assists in fund-raising. Indeed, the benefits are significant enough so as to make doing well in early contests critical to winning the nomination (Bartels 1985, 1988; Mutz 1995; Norrander 2006; Orren and Polsby 1987; Steger, Dowdle, and Adkins 2004).1 Early nominating contests are also important because they exert a winnowing effect, and those who do poorly in them are often forced to drop out (Norrander 2000, 2006; Steger, Hickman, and Yohn 2002).2

Although much research has been done on the effect of primary and caucus order on nomination outcomes, none has been undertaken on its influence over policy outcomes. On the surface, this makes sense. To get to the White House, a presidential candidate must also persuade the general electorate to vote for her. Moreover, the nomination frequently goes uncontested-generally, when a sitting president is running for reelection. On such occasions, the nominee does not have to be sensitive to primary and caucus order.

But there is anecdotal evidence that nomination contest order matters to policy outcomes. Presidential aspirants, it seems, attempt to outdo each other with promises to promote the policy interests of states with early contests. In 2004, for instance, Democratic candidates, such as Howard Dean and John Kerry, continually iterated their support for ethanol subsidies. Ethanol is an energy product made from corn, a crop that is important to the economy of Iowa, the state with the first contest.3 In 1992, Bill Clinton was criticized as a "pander bear" by opponent Paul Tsongas for promises of favorable policy he made to voters residing in early primary states, such as New Hampshire and Florida.

The recent "front-loading" of the nomination contests-that has occurred as states have brought their primaries and caucuses toward the beginning of the process so as to greatly shorten it (W. Mayer and Busch 2004)-seems to be at least partially the product of policy considerations. The push to move the Florida contest toward the beginning of the 2008 process was predicated on the idea that it would force the presidential candidates to take favorable positions on issues of importance to the residents of the state (Bosquet 2007). The same can be said for California. The state's assembly speaker, Fabian Nunez, put it this way: "Right now, California gets back 79 cents for every dollar we send to D.C. Early primary states . . . do much better. I think the prospective nominees should be vetted on their views on this issue" (Skelton 2007, 3B). Jim Brulte, a Republican leader in the state, remarked, "When you come and campaign in California, you learn about California. And when you learn about California, that makes you a better president for California" (Steinhauer 2007, A15). …