Democrats, Republicans—or Both? An Empirical Analysis of the Effects of the Composition of State Governments on FDI, 1977–2004

Abstract

Does the partisan composition of state governments in the United States influence the location decisions of foreign multinational companies? This article argues that it does. We contend that partisan differences over state economic development policies still exist. Whereas Republicans tend to prefer an investment-driven (supply-side) growth model, Democrats favor a consumption-driven (demand-side) path to growth. Both sets of policies are of value to foreign direct investment; thus, multinationals do not favor one party over the other. A useful blend of policy measures is sought by foreign firms, making split state government preferable over unified government. Our arguments are comprehensively tested in a time-series cross-section analysis covering the period 1977–2004, with results supporting our claims.

Keywords:

• foreign direct investment (FDI)
• panel data
• state governments
• United States

Acknowledgments

The authors thank Jonathon W. Moses for his helpful comments on a prior draft of this paper. The perceptive comments of two anonymous referees and the journal editor were also tremendously important. The usual disclaimer applies.

Notes

¹A caveat is nonetheless in order. For all states, but in particular for the smaller ones, state politicians' attempts to boost demand will also result in leakages of spending onto goods manufactured in other states. Given this, active demand management by the state governments will likely be less effective than demand-side policies instituted by the federal government.

²Investments from seven countries into the 50 states over a 28-year period implies a data set with 9,800 observations on the dependent variable, but due to missing data for some years in some states there are 9,148 observations on the dependent variable. Missing observations on some of the independent variables decreases the dataset further to around 8,300 complete observations. The analysis in Table 3, moreover, only considers unified governments, thereby reducing the effective number of observations to 3,744.

³Earlier we have seen that unified Republican governments to a much greater extent than unified Democratic governments practice unitary taxation. To the extent that unitary taxation is seen by MNCs as an investment disincentive clearly linked with unified Republican governments, it could possibly balance out the positive impact of lower business and personal taxes. In an alternative, unreported model we therefore introduced unitary taxation in a separate step, but the effect and significance level of unified Republican government stayed basically unaltered from results reported in Table 3.