MAIN FOCUS OF THE ARTICLE

This empirical research investigates state based factors affecting the inward FDI employment among fifty states of the United States, uses an­nual data for the period of time from 1997 to 2007 and identifies several state-specific determinants of FDI employment.

Literature Review and Hypotheses

The empirical literature related to the state based determinant of FDI employment in the US is limited. In evaluating the effects of FDI on the local economies, economists focus primarily on the performance of foreign-owned subsidiaries operating in the US It is already known that the establishment of a new foreign subsidiary or the expansion of an already existing one leads to higher employment and wages (Axarloglou, 2005). Reserchers identified link between job growth in the US economy during a period of increasing foreign direct investment flow. The economic impact on US employment due to FDI is evident, as are linkages among the various benefits due to the inward flow of FDI (Craig, 2008).

According to Axarloglou & Pournarakis in the last two decades, various US states offered strong economic incentives in an effort to attract FDI inflows, with the hope that FDI would stimulate local economies (Axarloglou & Pournarakis, 2005). Axarloglou, Casey and Han analyzed the effects of FDI inflows in local economies across US states. The empirical results point out that the US economy benefits from FDI inflows in manufacturing both in terms of employment and real wages.

The studies by Borstorff, Collum and Newton relate to FDI in the southern US, specifically automobile FDI in Alabama and describe state­specific features of southern states in recruiting foreign investment bringing the employment op­portunities (Borstorff, Collum & Newton, 2007). Ajaga and Nunnen a analysis complements the regression analysis of Mullen and Williams and the Markov chain approach of Bode and Nunnen- kamp and presents strong evidence of favorable FDI effects on output and employment at the level of US states (Ajaga & Nunnen, 2008).

Alfaro examined the effect of foreign direct in­vestment on growth in the primary, manufacturing, and services sectors. Foreign direct investments in the primary sector, however, tend to have a negative effect on growth, while investment in manufac­turing a positive one. Evidence from the service sector is ambiguous (Alfaro, 2003). Blomstrom, Fors and Lipsey compared the relation between foreign affiliate production and parent employ­ment in US manufacturing multinationals with that in Swedish firms.

Methodology

Data Sources and Variables

In order to test the implications of our models, we collected a panel of aggregate data on foreign direct investment on all US states, excluding the District of Columbia. The entire data set includes 50 states for which foreign direct investment and all other relevant variables are reported over the 1997-2007 period.

In the United States, the Bureau of Economic Analysis (BEA) (BEA, 2013a), a section of the US Department of Commerce, is responsible for collecting economic data related to FDI flows. Monitoring this data is very helpful in trying to de­termine the impact of FDI on the overall economy, but is especially helpful in evaluating states and industry segments. The data on stock of FDI are from the US Department of Commerce, Bureau of Economic Analysis (BEA) (BEA 2013a),

The real per capita disposable income is mea­sured as the nominal per capita disposable income deflated by the GDP deflator in constant (2000) US dollars. The real per capita taxes is measured by dividing the real state tax revenue by the state population.

The share of scientists and engineers in the workforce, a proxy for labor quality, is collected from the National Science Foundation, Division of Science Resources Statistics, Science and En­gineering Indicators 2010 (NSF 2013) The data on FDI related employment are collected from the Bureau of Economic Analysis while the data on state employment are collected from the US Department of Labor, Bureau of Labor Statistics (NSF, 2013). The information on real research and development expenditure is collected from the National Science Foundation, Division of Science Resources Statistics, Science and Engineering Indicators 2010 (NSF, 2013)

The data on the average wage and total state employment are collected from the US Depart­ment of Labor, Bureau of Labor Statistics. Fol­lowing Coughlin, Terza, and Arromdee (1991), the manufacturing density variable is measured as the manufacturing employment per square mile of state land excluding federal land. The data on manufacturing employment are collected from the US Department of Labor, Bureau of Labor Statistics (BLS, 2013)). The information on union membership is collected from http://www. unionstats.com/(Unionstats, 2013) maintained by Barry Hirsch (Georgia State University) and David Macpherson (Trinity University).

Model Specification

Drawing on the existing empirical literature in this area, we specify the following model:

where:

FDIEMP: FDI Related Employment EDU: Educational Attainment RFDI: Real FDI Stock

GSPGR: Real GSP Growth Rate PCEXP: Real Per Capita Exports PCTAX: Real Per Capita Taxes RWAGE: Real Wage

UNION: Union Membership (Share of Workers who are Members of Labor Unions)

SAE: Share of Scientists and Engineers in the Labor Force

MANDEN: Manufacturing Density

RND: Real Research and Development Expen­diture

HWY: Highway Mileage

FDIEMP_it represents FDI related employment in state i in year t; EDU_it is the real per capita expenditure on education in state i in year t; RF- DI_it represents real FDI Stock in state i in year t; GSPGR_it stands for real gross state product growth rate in state i in year t; PCEXP_it is real per capita exports in state i in year t; PCTAX_it symbolizes real per capita taxes in state i in year t; RWAGE_it is real wage in state i in year t; UNION_it repre­sents share of workers who are members of Labor Unions in state i in year t; SAE_it stands for share of scientists and engineers in the labor force in state i in year t; MANDEN_it represents manufacturing density in state i in year t; RND_it relates to real research and development expenditure in state i in year t; HWY_it stands for highway mileage in state i in year t.

Our first variable, the real per capita expendi­ture on education is expected to have a positive effect on foreign direct investment employment. Therefore, we would expect that β₁ > 0. Our second variable, the real FDI stock is expected to have a positive effect on FDI employment.

The eight variables, the share of scientists and engineers in the workforce, a proxy for labor quality is expected to have a positive effect on foreign direct investment employment. Therefore, we would expect that β₈ > 0. The manufacturing density is expected to be related positively to foreign direct investment employment. There­fore, we would expect that β₉ > 0. As Coughlin, Terza, and Arromdee (1991) and Head, Ries and Swenson (1995, 1999) point out, manufacturing density could also be used as a proxy for agglom­eration economies. States with higher densities of manufacturing activity is expected to attract more foreign direct investment because the foreign in­vestors might be serving existing manufacturers. Our tenth variable, the real research and develop­ment expenditure is expected to have a positive effect on foreign direct investment employment. Therefore, we would expect that β₁₀ > 0. Highway mileage is a indicator of infrastructure is expected to be positively correlated with foreign direct investment employment. Therefore, we would expect that β₁₁ > 0.

Empirical Results

The results of our empirical analysis are presented in Table 9., in addition to the eleven independent variables included in Equation (1). All the vari­ables presented in Table 9 are expressed in loga­rithm and the coefficient of each variable can be interpreted as elasticities. The results of the study imply that FDI employment in the US is strongly influenced by the state spending on education. The coefficient of this variable is positive and statisti­cally significant at the 1% level of significance.

The real stock of FDI has a positive and statistically significant effect on FDI related employment. The results of the study suggest that FDI employment is strongly correlated with the real FDI stock in the US This could be due to the fact that the states with high level of FDI employment also have larger FDI stock. Real per capita exports have the positive sign and are statistically significant at the 1% level. Most of the time higher FDI stock and employment result in higher state exports.

Real GSP growth rate has the unexpected negative sign and it is statistically significant at the 1% level. It can be explained by the fact, that many foreign investors choose the southern part of the US as a desirable location for their FDI. The southern US states has become more aggressive in recruiting foreign investment by providing incen­tives to attract investments and communicating the unique advantages they offer to foreign companies. The real per capita taxes has the expected negative sign and it is statistically significant at 1% level. This finding is also consistent with the findings of previous studies. Real wages have the positive sign and it is statistically significant at 1% level. It is known that foreign companies investing in US not only provide jobs, but relatively high-paying jobs what constitutes important determinant of FDI employment. Unionization variable has an expected positive sign and it is statistically significant at the 1% level of significance. Sur-

Table 9. Determinants of FDI related employ­ment in the United States, 1990-2007. Panel least squares estimates, dependent variable: FDI related employment.

Note: *** indicates the statistical significant at the 1% level.

prisingly, the share of scientists and engineers in the workforce has an unexpected negative sign. It can be related to the fact that the labor force is relatively more productive and skilled in urban than in rural areas. Manufacturing density vari­able has the expected positive sign. This variable is also expected to capture the agglomeration economies and we can guess that the more dense the manufacturing activity is in a given state, the more likely higher foreign direct investment em­ployment will be. However, current results reveal that the Southeast region in the US stems from relatively high manufacturing density. Highway Mileage represents infrastructure level in the state and it is definitely positively correlated with the FDI employment at the 1% level of significance.