FDI and Income Inequality: Evidence from a Panel of U.S. States

Article excerpt

I. INTRODUCTION

The challenges that foreign direct investment (FDI) might involve for high-income countries such as the United States are typically discussed with respect to the domestic repercussions of foreign activities of U.S.-based multinational enterprises (MNEs). In particular, the effects of offshoring and international sourcing on domestic labor demand and relative wages are subject to intense debate, with Slaughter (2000) and Feenstra and Hanson (1996) as prominent contributions.

However, the United States is not only the most important source of FDI but also tops the list of host countries of foreign-based MNEs. (1) U.S. states compete aggressively for FDI inflows, especially for new manufacturing plants (Head, Ries, and Swenson 1999), and recent empirical evidence by Keller and Yeaple (2009) suggests that inward FDI raises productivity and income even though U.S. firms often produce at the technological frontier. Possible drawbacks of FDI attracted by U.S. states, however, have received hardly any attention. This is in striking contrast to FDI in less advanced locations where it has been shown that FDI-induced growth effects typically went along with rising income inequality.

The effects of FDI on income distribution in advanced economies are theoretically ambiguous, and evidence is limited and inconclusive (Section II). (2) Against this background, we explore the relationship between inward FDI stocks and (several measures of) income inequality at the level of U.S. states for the period 1977-2001. We employ panel co-integration techniques that allow for cross-sectional heterogeneity and cross-sectional dependence

(Section III). Perhaps surprisingly, we find that FDI significantly reduces income inequality in the United States in the long run. However, there is considerable heterogeneity across U.S. states, with several states exhibiting a positive relationship between FDI and income equality (Section IV).

II. THEORETICAL BACKGROUND AND PREVIOUS FINDINGS

In contrast to outward FDI from the United States (and other developed countries) in less advanced host countries, the distributional consequences of inward FDI in the United States have received scant attention. Endowment-driven FDI models such as Feenstra and Hanson (1997) provide important insights into the distributional effects in developing host countries. They are, however, less useful for inward FDI in advanced host countries. According to Markusen (1995), advanced countries hosting most of the worldwide FDI have similar relative factor endowments.

In North-South models a la Feenstra and Hanson (1997), cost-oriented and vertical types of FDI result from large income and wage differentials between the source and the host country. For instance, Mexico's average per-capita income was less than 20% of the per-capita income in the United States; it is a major source of FDI. (3) By contrast, the major source countries of inward FDI in the United States typically have a somewhat lower per-capita income than the United States. Markusen (1995) therefore posits that the United States tends to attract mainly market-oriented, horizontal types of FDI. In the words of Blonigen and Slaughter (2001, 364), "it seems unlikely that inward FDI into the United States has occurred because ... other countries are now outsourcing unskilled-labor-intensive activities to the United States" so that "it is quite difficult to assess theoretically whether and how this inward FDI should affect wages." This does not necessarily imply that the variation of per-capita income across U.S. states is irrelevant for location choices by foreign firms within the United States. (4) However, the simple correlation between inward FDI stocks and per-capita income is weakly positive across U.S. states (0.18 in 2001). The ten poorest U.S. states hosted just 13% of total inward FDI stocks. This suggests that the location of inward FDI in the United States has been motivated more by considerations of market access than by cost considerations. …