According to two articles previously published in this journal, fiscal policy exerts a predictable effect on the outcome of presidential elections in the United States (Cuzan and Heggen 1984; Cuzan and Bundrick 1992). It appears that, independent of economic conditions, fiscal restraint is rewarded but fiscal expansion is rejected at the polls. These findings are consistent with those of William Niskanen (1975, 1979) and Sam Peltzman (1990, 1992), the only other scholars, both economists, who have explored the impact of fiscal policy on American presidential elections.(1)
This article updates, deepens, and extends the earlier work. The 1992 and 1996 elections are added to the set, for a total of thirty-two elections (or thirty, depending on data availability--see below) extending across a period of well over a century, one of the two longest data series that, to the best of our knowledge, is accounted for by any one presidential elections model.(2) An improved specification of the multiple regression model, one that uses the percentage of the two-party vote, rather than the percentage of the total vote, and which adds the number of consecutive terms that presidents of the same party have occupied the White House, results in a respectable fit with the data. Finally, analysis of the relationship between fiscal policy, presidential incumbency, and election outcome shows that it is not incumbency but fiscal policy that accounts for the success of sitting presidents. This last finding raises intriguing questions about the role of presidential leadership in democratic government.
Fiscal Policy and Presidential Elections: Testing an Implausible Hypothesis
Following Cuzan and Heggen's original article, published in this journal (see also Cuzan and Heggen 1985; Cuzan and Bundrick 1996), it is hypothesized that the outcome of presidential elections is contingent on fiscal policy. If fiscal policy is expansionary, incumbents are defeated; if it is cutback, they are reelected. By incumbents we mean the president or his party's candidate, along with his team. Fiscal policy is expansionary if F, the ratio of federal expenditures to gross national product (GNP) between election years, rises at a rate that is equal or greater than during the previous presidential term; it is cutback if this ratio declines or increases at a slower rate than in the previous administration (for a formal definition, see Table 1). There is a third possibility: a steady-state fiscal policy, where the ratio stays the same for two consecutive administrations. However, there is not one such case since 1872. Thus, de facto, fiscal policy or FISCAL is a binary variable (more about this below). Note that what defines fiscal policy are not the changes in the absolute amount of federal expenditures but in the percentage of GNP spent. It is in this relative sense that terms such as federal spending or expenditures and size of the budget are used here.
Definitions and Measurements of Variables
VOTE2 Percentage of the two-party
vote won by the
incumbent party candidate
(adapted from Fair 1996a).
ELECT ELECT = 1 if incumbents (the president
or his party's nominee) win
reelection; ELECT = -1
if incumbents are defeated.
GROWTH (g3) The annualized rate of growth
of real per capita
gross domestic product
(GDP) through the first three
quarters of the
presidential election year
(Fair, 1996a, 1996b).
INFLATION (p15) The annualized rate of growth
of the GDP price
index in the first fifteen
quarters of the presidential
term (Fair 1996a, 1996b).
TERMS (T) The number of consecutive
terms by presidents
of the same party