The Local Labor Market Effects of Korean Automotive Investments in the United States

Abstract

This study investigates the local labor market effects of automotive foreign direct investment (FDI) in Alabama in the wake of the North American Free Trade Agreement (NAFTA). Of particular interest is the effect of FDI originating from a nontraditional source country in comparison to investments by European and Japanese firms. Using 2005–2011 American Community Survey (ACS) data, we compare changes in employment rates and median weekly wages in three area types: areas with Korean FDI, areas with German and Japanese FDI and areas without FDI. Results show that Korean FDI leads to increased employment rates but decreased median weekly wages. We attribute these findings to price competitive strategies of Korean firms.

Keywords

North America Free Trade Agreement (NAFTA)economic development foreign direct investment automotive manufacturing labor market employment rates wage differentials

Introduction

Debates over whether the North American Free Trade Agreement (NAFTA) of 1994 has destroyed manufacturing jobs in the U.S. reached a peak in the months leading to the U.S. Presidential election of 2016. While its impacts on the U.S. labor market remain controversial, foreign automobile assemblers and component manufacturers increasingly engaged in massive investments in North America as a result of this economic shock (Johnson 1993; Lopez-de-Silanes, Markusen, and Rutherford 1994; Tuman and Erlingsson 2019). Specifically, auto sector provisions under NAFTA triggered the earliest entry of German and Japanese assemblers into the U.S., such as BMW (1994), Mercedes-Benz (1997), and Toyota (1999), and most recently the South Korean (hereafter Korean) assemblers Hyundai (2005) and Kia (2009). These automotive firms show similar investment behaviors of choosing the U.S. over Canada or Mexico and also selecting locations in the South rather than traditional auto states in the Midwest. However; within the same state, foreign automakers show disparities in their county-level locational preferences and corporate strategies. In this economic context, we explore the impact of post-NAFTA automotive foreign direct investment (FDI) on the local labor market in the U.S.

Many countries, irrespective of their level of development, attract FDI to promote economic development (see Blomström and Kokko 2003). From earlier research based on dependency theory and world system theory (Wallerstein 1974; Bornschier, Chase-Dunn, and Rubinson 1978; Evans 1979) to later studies criticizing those theories (Firebaugh 1992; Firebaugh and Goesling 2004), scholars have given a great deal of attention to assessing developmental effects of FDI and its implications for host populations. The earliest literature on the role of FDI in economic development theorized that FDI decelerates the periphery’s economic growth because it creates labor market inequalities through generating employment in a particular sector (Baran 1956; Bornschier, Chase-Dunn, and Rubinson 1978; Evans 1979; Evans and Timberlake 1980). However, later research contradicts the findings in prior studies, supporting that foreign capital spurs economic growth and mitigates income inequalities by industrializing poor regions (Firebaugh 1992; Firebaugh and Goesling 2004).

Although these studies in sociological literature extensively discuss labor market changes resulting from FDI, they primarily focus on FDI’s effects on less developed countries. In addition, the majority of this research has focused on the distributional effects of FDI at country and global levels without much consideration of its potential impacts at the regional and local levels. However, as major receivers (as well as senders) of FDI, advanced economies also have relied on foreign capital for economic growth and revitalization (Markusen and Venables 2000; Yeaple 2009). In addition, many local governments participate in competitive bidding to promote foreign investments. Once a firm decides to enter a foreign market and select locations, it is not only the FDI attraction policies at the national level, but also the incentive agreements negotiated at the local level that matter to foreign investors (Brown and Raines 2000; Phelps and Raines 2003; Wren and Taylor 1999).

On the other hand, economic geographers have expanded discussion on the local and regional effects of FDI in developed countries. There is a long tradition of research in economic geography literature concerning the regional development impacts of FDI in developed countries since the 1970s. One of the key concepts in this literature is branch plant syndrome, which refers to the regional effects of branch plants established in economically disadvantaged regions within developed economies (Firn 1975; Watts 1981). As part of multi-plant firms, branch plants are involved in the production activities, often at the lower end of the commodity chains. Researchers have found that FDI inflows in the form of branch plants in less-developed regions improve local economic conditions in the short term but do not lead to sustainable development (Hallwood 1986; Harri, Lloyd, and Newlands 1988). Recent studies in economic geography also emphasize foreign investors’ roles in the geographic dispersion of global commodity chains and their impacts on regional economic development (Crescenzi and Iammarino 2017; Monastiriotis and Jordaan 2010; Zhang and Zhao 2007).

Such a trend is expected to be accelerated by recent economic stagnation as well as a rapid growth and expansion of multinational enterprises (MNEs) from emerging markets (see Guillen and Garcia-Canal 2009; Sonn and Lee 2012). In this sense, this study intends to explore whether and how foreign capital inflows from nontraditional source countries influences the local labor market in established economies. We examine changes in local labor market outcomes driven by FDI in the U.S., taking the case of automotive investments made after NAFTA’s ratification in 1993. We also discuss whether and how entry of foreign firms may contribute to reducing existing labor market inequality at the local level. Of particular theoretical interest is the effect of reverse FDI, which Ramamurti (2009) refers to as a “man-bites-dog” case (p. 8) in which traditionally-defined FDI senders and receivers are switched. According to the U.S. Bureau of Economic Analysis (2014), the share of FDI from nontraditional sources is rapidly growing in the U.S. although its cumulative stock is still much lower than that of top investors from European countries (e.g., United Kingdom, Netherlands, Luxembourg, Germany, Switzerland, and France), Japan, and Canada. In 2009–2014, the total amount of FDI from Korea ($23.9 billion), China ($9.8 billion), and India ($4.3 billion) had been enormous. Under these circumstances, it is worthwhile to understand whether reverse FDI can be an alternative source of economic growth for the U.S. economy, bringing “good jobs” back to local populations.

In this paper, we explore changes in the local labor market outcomes that vary by whether the locales have experienced inward automotive FDI from a nontraditional source country. Then we compare differences in the labor market changes driven by FDI inflows from nontraditional sources with those driven by FDI inflows from traditional sources. Specifically, we estimate and compare changes in the employment rates and weekly wage levels in Alabama before and after post-NAFTA automotive investments by companies from different home countries: Germany, Japan, and Korea.¹ Finally, we discuss the results in relation to the foreign firms’ strategies.

Background and Expectations

FDI and Its Country of Origin

Competitors in the same industry implement different strategies depending on their market positions. First or early movers have sufficient time to accumulate necessary knowledge, develop sophisticated technology, and establish their brand reputation (Suarez and Lanzolla 2005). In contrast, later entrants lacking such resources find it difficult to differentiate their products and often rely on price competitiveness to increase market profitability. This logic applies to the process of international expansion as well. When investing in foreign markets, firms show variations in their approaches to achieving market success, and such differences are highly associated with where the firms come from.

Bartlett and Ghoshal (2000) argue that late movers originating from nontraditional source countries confront various challenges because of their countries of origin (or “liabilities of origin”). When operating in unfamiliar business environments, every MNE needs to overcome not being local—the so-called liability of foreignness. However, MNEs originating from countries that are not among the most developed in the world are also subject to disadvantages related to their late-mover position and domestic institutional constraints (Mathews 2002; Luo and Tung 2007). Specifically, the late movers have been relatively less successful in penetrating developed countries than developing countries. This is because the conditions of the most developed countries differ from their domestic institutional and economic environments and their firm-specific advantages (FSAs) are not well adapted to the needs in the advanced host markets (Cuervo-Cazurra and Genc 2008; Guillen and Garcia-Canal 2009; Meyer, Mudambi, and Narula 2011). The firms also have difficulties acquiring legitimacy for not possessing reputational capital that protects them from discrimination by competitors, consumers, and host governments (Yildiz 2014). These entry barriers require less experienced firms from nontraditional source countries to adopt strategies distinguished from those implemented by established multinationals (see Luo and Tung 2007; De Beule, Elia, and Piscitello 2013).

From previous findings, we conclude that foreign investors in the same industry entering the same host country may show different investment behaviors depending on their country of origin and market forces. While a majority of past studies focus on investment strategies themselves, we suppose that disparities in the firms’ strategies can potentially moderate their impacts on the labor market in FDI destinations, resulting in different labor market outcomes.

FDI and Labor Market Outcomes

FDI and employment

The consensus in the literature is that MNEs’ investments, regardless of their country of origin, have distributional impacts on the host economies, creating employment. Apart from attracting highly productive resources to the locations, FDI creates jobs not only in the relevant sectors but also in supportive industries (Aaron 1999; Iyanda 1999; Zhang and Felmingham 2002; Waldkirch, Nunnenkamp, and Bremont 2009). However, the extent of these effects may depend on the sources of investments (Little 1986; Bagchi-Sen 1991). Research shows that firms entering foreign markets follow different stages of expansion (Johanson and Vahlne 1977). They gradually increase their resource commitment in the host country as they accumulate the knowledge and experience required to overcome the liability of foreignness.

In addition, firms originating from nontraditional sources have various disadvantages (that is, lack of advanced technology and know-how, branding deficiencies, home-country-based constraints, and so forth) compared to local or foreign firms from established economies (see Bartlett and Ghoshal 2000; Luo and Tung 2007). In other words, these firms have fewer FSAs, and thus smaller sizes and lower performance, than mature multinationals do (Guillen and Garcia-Canal 2009). These different levels of FSAs may affect the firms’ behaviors such as their mode of entry and locational choices, subsequently resulting in differential employment effects (Harrington, Burns, and Cheung 1986; Bagchi-Sen 1995).

In this study, we expect that reverse FDI entering the U.S. may affect employment differently, yet still positively, than FDI from traditional sources. Due to a shorter history of investments in the U.S. and the latecomer disadvantages, firms engaging in reverse FDI may take a slower move in its expansion and subsequently have a smaller scale of production compared to early movers that have invested in the U.S. for a longer period. Thus, we expect that reverse FDI may have smaller effects on employment than FDI from traditional sources does.

Hypothesis 1a: Reverse FDI in the form of greenfield investments positively affects employment in the affected areas relative to the unaffected areas.

Hypothesis 1b: Reverse FDI in the form of greenfield investments positively affects employment in the affected areas to a lesser degree than traditional FDI.

FDI and wages

Research shows that foreign companies pay higher wages, on average, than domestic firms, eventually increasing overall income level in the relevant industries (Clerides, Lach, and Tybout 1998; Harrison and McMillan 2011). This is because the firms engaging in cross-border investments tend to be larger, more productive, more capital-intensive, and more technology-intensive. Although some highlight labor-exploitative behaviors of MNEs, the firms do not necessarily pay lower wages than local firms (see Harrison and Scorse 2010). Any wage difference can be caused by various structural characteristics (e.g., the cost of living, comparability with other firms’ wages, and sector-level collective agreements, Brown, Ingram, and Wadsworth 2004; Gerlach and Stephan 2006) in addition to the firm-specific characteristics (e.g., size, ownership, and performance, Groshen 1991; Card and De la Rica 2006). Due to such wage adjustment, MNEs, regardless of their sources, take industry-level standard wage rates of the host country into account in their pay policies.

Noteworthy about the U.S. automotive industry, however, is that domestic and foreign firms determined wage differently due to their relationships with the United Auto Workers (UAW), an American labor union that represents the interests of workers in the automotive sector. The American automakers, commonly referred to as the big three (i.e., General Motors, Ford, and Chrysler) are historically known to have accepted collective bargaining. Through continuous negotiations, the UAW had put pressure on the American firms for wage growth, which resulted in the firms’ agreement on the annual improvement factor, which was later replaced with profit sharing plans (De Gier 2010; Katz and Meltz 1991; Holmes and Rusonik 1991). On the other hand, foreign automakers have actively avoided unionization of their workers by choosing to settle in the Southern states, most of which have right-to-work (RTW) laws that allow workers to make their own decision of whether to join a union (Cole and Deskins 1988; Mair, Florida, and Kenney 1988; Reid 1990; Ulgado 1996). In addition to such a locational choice of foreign firms since the 1980s, other industrial changes (e.g., automation, failed international competition, U.S. firms’ increased offshoring activities) further debilitated unions in foreign (and eventually also in domestic) automotive manufacturing and allowed foreign firms to pay hourly wages lower than the industry average (Morgan 2008; Rubin 1996).

We expect that, even among foreign automakers, MNEs from nontraditional sources may bring different impacts on wages for the American workers than described in the earlier studies. MNEs engage in FDI largely for two motivations: market access (horizontal FDI) and comparative advantage (vertical FDI). Horizontal FDI is characterized by same production processes in different locations to save on transport and/or trade costs. Vertical FDI is intended to separate stages of production to exploit factor price differences and hence lower production costs (Yeaple 2003). While MNEs’ entry into advanced markets is usually associated with their market access motive, it is essential for the firms to save on their production costs because country-specific advantages (CSAs) in relatively “cheap” labor, which often determine competitive advantages of the late movers, are likely to be unavailable in more developed business environments. In order to maintain cost leadership relative to mature multinationals with strong ownership advantages, firms engaging in reverse FDI may attempt to minimize increasing costs (see De Beule, Elia, and Piscitello 2013). Based on this, we advance the following hypotheses on the level of wages:

Hypothesis 2a: Reverse FDI negatively affects the level of wages in the affected areas relative to the unaffected areas.

Hypothesis 2b: Reverse FDI negatively affects the level of wages in the affected areas to a greater degree than traditional FDI does.

Research Context: Korean Automotive Investments in Alabama

To test these hypotheses, this paper focuses on Korean automotive FDI inflows into the U.S. The integration of the North American automobile industry attempted in earlier intergovernmental negotiations such as the 1965 Canada-U.S. Auto Pact, as well as the 1988 Canada-U.S. Free Trade Agreement, was finally achieved through NAFTA’s ratification (Johnson 1993). While this agreement eliminated tariffs and other barriers to trade in the automotive sector among the NAFTA countries, it also increased protectionist barriers (that is, tariffs and local content requirements) against imports of automobiles and auto components from other third countries (Lopez-de-Silanes, Markusen, and Rutherford 1994). The NAFTA required 62.5 percent of regional value content for automobiles to be imported without tariffs, which later increased to 70 percent under United States–Mexico–Canada Agreement (USMCA). As a result, many foreign automakers from Europe and Japan entered the U.S. for local production from 1994 onward. Korean automakers also had little choice but to shift production, and hence jobs, across the Pacific.

In addition, a growing U.S. trade deficit in the automotive sector, which was partially attributed to a declined dollar strength in the late 1990s to 2000s, resulted in the U.S. government’s protectionist actions against Korean auto exports to the North American market. Hyundai Motor’s market successes of the late 1990s in North America were perceived as a strong threat to American automobile producers. Thus, through United States–Korea Free Trade Agreement (KORUS FTA) negotiations, the U.S. government started putting economic pressure on the Korean government to increase sales of American automobiles in Korea and to encourage Korean automakers’ local production in the U.S. Moreover, the adverse movements of the exchange rate increased production costs in Korea and further motivated Korean automotive firms’ greenfield investments in the U.S. (Bank of Korea 1997–2007). In an effort to cope with the various trade barriers and changes in the real exchange rate, Hyundai decided to make tariff-jumping FDI in 2002. In 2005, Hyundai Motor Manufacturing Alabama (HMMA) produced its first “made-in-USA” vehicles in Montgomery, AL, a move that was followed by several waves of transplant investments of Hyundai’s suppliers, its sister company, Kia Motors, and its suppliers.

This study examines how Korean automotive firms’ entry into Alabama affected the local labor market. Also, this study compares the outcomes with those of traditional FDI in Alabama, expecting differential effects among them. Following the establishment of German-owned Mercedes-Benz assembly plants in Vance in 1994, Japanese-owned Honda entered Alabama and constructed its production sites in Lincoln in 2001. As mentioned earlier, unlike these MNEs from established economies, Korean multinationals may experience increased production costs in the U.S. In addition, Korean firms have latecomer disadvantages as well as a shorter history of investments in North America. These disadvantages outweigh those experienced by Japanese counterparts that created a substantial wave of transplant investments in 1980s or the German counterparts that have successfully maintained production since the 1990s. Such a challenging status may alter the ways in which Korean firms affect the host labor market.

In this study, we employ both qualitative and quantitative methodologies to explore our research questions because combining the two methods can provide a more comprehensive insight. As a first step of this study, we present our qualitative analysis on Korean automotive firms’ late mover strategies. Our second step is to conduct quantitative analysis to test our hypotheses. We define FDI types depending on their sources and explore differences in contextual-level conditions across areas accordingly. Second, we estimate and compare the effects of FDI from different sources on the changes in local employment rates and wage levels. Third, we examine whether FDI from different sources continues to produce a unique effect on the dependent variables once the changes in local demographics and labor market conditions are considered. Finally, we explore differences by sex in each of these analyses.

Qualitative Analysis of Interview Data

Having made trips to Alabama four times between 2012 and 2014, we conducted 10 in-depth interviews with corporate executives, managers, Alabama government officials, and staffing agencies’ managers. We spent, on average, two to three hours interviewing each of the interview participants. We also collected archival data concerning recruitment processes from both the firms and the state government.

Korean Automotive FDI Strategy

According to an interview with one of the executives at HMMA (25 February 2012), when Hyundai searched for a potential production site in 2000, the biggest consideration was in labor costs: availability of cheap, nonunionized labor. That is, Korean firms have systematically selected low-growth, low-income areas to invest in. On top of that, Korean firms do not share labor force with other foreign automakers in Alabama due to geographical distance, which enables the firms to monopolize the cheap labor (see Table A1). Accordingly, it may be a rational choice for them to continue to set low wages exercising strong bargaining power over labor force in the areas.

This theoretically supports that late movers adopt strategies distinguished from those implemented by early movers to boost price competitiveness. As mentioned previously, when competing against early movers with strong ownership advantages, late movers usually choose cost-efficiency as their strategy. The cost-effective strategies may become even more important when the firms engage in reverse FDI, as their CSAs are no longer available and they suffer from latecomer disadvantages. Thus, minimizing increasing costs of production is one of the primary agendas for these multinationals. Along this line, Hyundai has implemented cost-reduction strategy in two additional areas: assembler-supplier relations and selective production lines.

First, Hyundai chose to use its traditional Korean partners as a primary source of parts supply in the U.S. rather than partnering with local or foreign suppliers. Hyundai has created manufacturing geographical agglomeration in Alabama, inviting many of its Korean partners (see Jones 2017). According to the Alabama Department of Commerce (2013), 165 automotive firms have settled in Alabama since the settlement decision of Mercedes-Benz in 1994. The largest share of them serve Hyundai (70), followed by Mercedes-Benz (49) and Honda (42). When we restrict the firms to those with 50+ employees, the largest share still serves Hyundai (49), followed by Honda (31) and Mercedes-Benz (27). By having their old partners build local production facilities, Hyundai sought to control product quality and, more importantly, production costs (interview with a purchasing manager at HMMA, 25 February 2012).

This supplier management policy enabled Hyundai and its first-tier Korean suppliers in Alabama to maintain low labor costs by handling the management-labor relations together; the Korean automotive firms in the areas agreed to build production facilities with some distance from one another in order to avoid potential labor disputes (interview with a purchasing manager at HMMA, 25 February 2012). Having previously dealt with frequent labor strikes by strong unions in the domestic market, Hyundai was concerned that within-industry wage differential across Korean firms in Alabama may arouse antipathy among the laborers working at suppliers that pay lower wages. By attempting to keep apart the workers physically, the firms managed their labor costs. Such an effort to reduce costs may have had trickle-down effects, putting pressure on smaller vendors and suppliers who already pay much lower wages than a flagship company. That is, Hyundai’s investments may have, as the local government expected, created some high-paying jobs, but they could also have created much more low-paying jobs at the same time.

Moreover, the vehicles produced by each automaker in Alabama support Hyundai’s price competitive strategies. Both Mercedes-Benz and Honda manufacture mid- or full-size sport utility vehicles (SUV) in Alabama whereas Hyundai produces less sophisticated models. Though Mercedes-Benz added a compact (C-Class) production line in 2014, it primarily makes M-Class, R-Class, and GL-Class. Similarly, Honda focuses on Odyssey (a minivan), Pilot (a mid-size SUV), and Acura MDX (a mid-size luxury SUV). In contrast, Hyundai manufactures Sonata (a mid-size sedan) and Elantra (a compact car)—two of its most basic, price-sensitive car models. As Hyundai begrudgingly started local production in the U.S. in response to protectionism, it was necessary to seek to keep production costs as low as they were in Korea. Accordingly, Hyundai has relied on importing its high-end car models from Korea while only making best-selling, affordable car models on site (interview with a purchasing manager at HMMA, 25 February 2012).

Quantatative Analysis: Data and Methods

Our analytic strategy is to assess the local labor market effect of reverse FDI at an aggregate level. The focus of our analysis is on yearly changes to two main labor market outcomes: employment rates and logged median weekly wage levels.

The data for analysis comes from a seven-year sample of individuals (2005–2011) of the American Community Survey (ACS; Ruggles et al. 2010). In the models of employment rates, we include Census 2000 data in addition to the ACS data (2005–2011). Since the Census and the ACS ask respondents to report the previous year’s employment status and wage levels, using data of a single year provides a lagged value. We restrict the sample to the non-institutionalized, non-student population aged eighteen to sixty four to capture prime working ages. In the models of wage, we further restrict the sample to those who are employed. The sample includes residents of the thirty Public Use Microdata Areas (PUMAs)—an indicator of the local labor market—in Alabama, and the primary unit of analysis is the PUMA. The end product is panel data of thirty PUMAs spanning 2000 and 2005–2011 estimating employment rates and logged median weekly wage levels for the local area. This yields 480 observations and 420 observations in the models of employment and the models of wage, respectively.

Model Specification

As mentioned above, the dependent variables are yearly changes in employment rates and logged median weekly wage levels. We analyze the changes in the employment rates at seven time points (2005, 2006, 2007, 2008, 2009, 2010, and 2011) and changes in the wage levels at six time points (2006, 2007, 2008, 2009, 2010, and 2011). For analysis of changes in logged median weekly wages, earnings are deflated by the Federal Reserve Economic Data (FRED)’s Consumer Price Index and placed in 2009 constant dollars prior to log transformation.

The main explanatory variable of interest is yearly FDI penetration from varying sources in the respondents’ residential PUMA areas. Combining data from the 2013 Alabama Industrial Database with the Alabama automotive industry profile and supplier list published by Alabama Department of Commerce, we restrict the automotive firms to those with 50+ employees to allow for the supposed impacts on the labor market. Then we distinguish between two area types: PUMAs with no FDI penetration (reference; hereafter NDI) and PUMAs with FDI penetration. The latter area type received FDI from different sources, namely Korean direct investment (hereafter KDI) and other, or non-Korean, direct investment (hereafter ODI). FDI penetration is a continuous variable with numerical values scaled to represent the number of individuals employed by foreign investors of different origins divided by 1,000.

We measure FDI penetration using four dimensions: the firms’ location in Alabama, their starting year of production in Alabama, the assemblers’ country of origin, and the suppliers’ major client. The firms that are incorporated for definition have all arrived in Alabama since 2005 (the year that KDI penetrated), and they represent FDI penetration in particular areas only after their starting year of production. We also argue that the country of origin of each supplier’s main client is a better measure for the source of FDI penetration than the country of origin of each supplier itself. For example, supplier firms that settled in Alabama to serve a Korean assembler are from various countries (including Korea) but are considered a part of KDI for the analysis. This applies the same to supplier firms serving German and Japanese automakers. This is attributed to the characteristics of automobile industries whose commodity chains are subject to assembler-supplier relationships, specifically downward margin pressure based on their vendor contracts and pricing agreements. Thus, we consider the country of origin of the suppliers’ “main” client; we define KDI as investments made by a Korean automaker and any supplier firms whose main client is the Korean automaker and ODI as investments made by German and Japanese automakers and any supplier firms whose main clients are the German and Japanese automakers.

Statistical Specification

In order to evaluate the yearly effect of Korean and non-Korean investments on the labor market outcomes, we use the two-way fixed effects estimation of difference-in-differences (DID) method. With this approach, we compare the difference in the employment rates and wage levels before and after the FDI inflows in the areas affected by the investments to the same difference in unaffected areas. Average changes over time in localities without investments are then subtracted from average changes over time in localities with investments. By using DID estimation, we can substantially reduce the omitted-variable bias (that is, endogeneity) that can be created in the cross-sectional analyses. Formally, the specification is as follows:

δ = ({\bar{Y}}_{C - year} - {\bar{Y}}_{C - (year - 1)}) - ({\bar{Y}}_{T - year} - {\bar{Y}}_{T - (year - 1)})

(1)

where $δ$ is the DID causal effect of interest (that is, the amount of FDI penetration), $\bar{Y}$ is the average outcome in control (C) and treatment (T) groups in a year and its prior year.

Then we use OLS regression to estimate the two-way fixed effects models that control for time effects.

Y_{iyear} - Y_{i (year - 1)} = β + δ_{k} P_{ki} + δ_{o} P_{oi} + λ_{0} X_{1 i} + λ_{1} X_{2 i} + ε_{i}

(2)

where Y_i is the outcome in a year and its prior year in local area i; P_ki is a dummy variable of treatment group indicator that is equal to 1 if the area has attracted KDI inflows and 0 otherwise; P_oi is a dummy variable of treatment group indicator that is equal to 1 if the area has attracted ODI inflows and 0 otherwise; X_1i is a vector of local area controls which include prior conditions and changes between the two periods; X_2i is a vector of yearly controls; δ is the causal effect of inward FDI as specified in (1); and $λ$ are parameters to be estimated.

While fixed effects estimation controls for the effects of all the unobserved variables that do not change over time (e.g., PUMA-level dummy variables), our models also include contextual controls that change over time as well as the year dummy variables. The contextual controls are indicators of local area conditions that are constructed by aggregating the individual information to the PUMA level, using the ACS dataset (Ruggles et al. 2010). These include total population size (logged), the share of college graduates, industrial composition (the share of the population employed in the manufacturing industry and the share of the population employed in agricultural industry), the share of the foreign-born population, and the share of the black population. By including these PUMA-level variables, we controlled for overall labor availability (Bloom, Canning, and Fink 2010; Bloom and Freeman 1986), high-skilled labor availability (Peri 2016), skilled labor and wage structure (Dickens and Katz 1986; Barth et al. 2016), agglomeration of immigrant populations that increases cheap labor availability (Kerr et al. 2017; Ndofor and Priem 2011), and the native-born populations’ replaceability with immigrant labor that depresses wage (Borjas 1987; Llull 2018). Also included is an interaction term of whether the areas experienced automotive FDI penetration before 2005 (which is constructed using combined data of the 2013 Alabama Industrial Database and Alabama Department of Commerce’s supplier list) and how much the areas received KDI. Additionally, we added an interaction term of the share of the population employed in the manufacturing industry and KDI inflows, an interaction term of population size of PUMA and KDI inflows, and an interaction term of prior employment rates and KDI inflows. We include these three interaction terms to consider that KDI’s employment effects can become negligible due to other PUMA-level conditions, such as particular industrial activities, changing labor supply, and prior labor market conditions.

Results

Table 1 presents descriptive statistics for the dependent variables by FDI typology and sex of resident. Results show that both employment rate and median weekly wages are on average substantially lower in areas with KDI penetration. For instance, men in KDI areas are 9 percentage points less likely to be employed and average nearly $120 less than counterparts in NDI areas. This pattern is largely true for women, as well, but to a lesser degree. Noteworthy is that, in contrast to the general notion, results show that the residents of NDI areas are more likely to be employed and earn more than those of areas with FDI penetration.

Table 1.

Descriptive Statistics of Dependent Variables by FDI Typology and Sex of Resident, 2005–2011.

	Men			Women
	Areas with	Areas with	Areas with	Areas with	Areas with	Areas with
	No Investment	Korean Investment	Other Investment	No Investment	Korean Investment	Other Investment
Employed (%)	73.21	64.33	67.61	55.34	50.50	51.92
N	33,935	14,065	21,521	31,775	13,335	19,829
Wages (median)	$762	$641	$706	$494	$410	$463
N	20,527	7,608	11,467	14,953	5,656	8,302

Source: ACS (Ruggles et al. 2010).

Note: The numbers for wage levels represent medians of the total wage and salary income divided by weeks worked reported from 2006–2011.

Before attributing the differences observed in Table 1 to FDI penetration, we must also examine the extent to which local demographics and labor market conditions vary across locales. For example, if the less educated disproportionately reside in areas with KDI penetration, disparities in employment rates and wages might reflect contextual characteristics rather than the effects of FDI penetration. Table A1 reports descriptive statistics for contextual-level control variables. Based on the three FDI typologies, areas with KDI inflows have the least favorable labor market conditions. Specifically, KDI areas have a much smaller share of college graduates than NDI areas and areas with ODI inflows do. In addition, they have greater representation of people in the agricultural and mining industry (5.1%) and the manufacturing industry (23.2%) than other areas have. Moreover, KDI areas have experienced FDI penetration prior to 2005 to a lesser degree (36.0%) than ODI areas have (67.1%). This indicates that employment rates and wage levels may differ due to socioeconomic contexts irrespective of FDI inflows. Thus, we include changes in these indicators as control variables in the fixed effects estimation.

The following multivariate models analyze locational differences in changes in the employment rates and wage levels. We control for contextual variables and yearly effects.

Effect of KDI Inflows on Employment Rates

Table 2 presents results from DID models estimating the relationship between inward FDI from different sources and changes in employment rates. The first column includes only yearly controls while the second adds PUMA-level contextual predictors. The reference category is NDI. Focusing on the effect of KDI inflows and changes in the employment rates, results show that KDI inflows positively affect employment growth rates. After accounting for contextual predictors, changes in employment rates for KDI penetration average 10 percent higher, but not significantly, than those for NDI. However, results are statistically significant separately for men and women living in areas that attracted KDI inflows; men experience a higher employment growth by 12 percent (p < .01) and women by 2 percent (p < .01) for every increase of one on the scale of KDI penetration. Such positive effects appear only after controlling for various interaction terms and socioeconomic contexts. This suggests that KDI results in a higher growth rate of employment when other intervening local demographic and labor market conditions are held constant.

The pattern is in the opposite direction for ODI inflows; changes in employment rates for ODI penetration average 20 percent lower, but not significantly, than those for NDI. In other words, compared to NDI, FDI inflows from nontraditional source positively affect employment growth while FDI inflows from traditional source do not affect employment growth rate any differently.

Table 2.

OLS Estimates from Difference-In-Differences Models Predicting Change in the Employment Rates.

		Employment
		All				Men				Women
FDI types (ref = no investment)
	Korean Investment	0.09		0.10		0.09		0.12	**	0.09		0.02	**
		(0.24)		(0.32)		(0.34)		(0.04)		(0.34)		(0.01)
	Other Investment	−0.24		−0.20		−0.24		−0.28		−0.24		−0.13
		(0.22)		(0.21)		(0.31)		(0.32)		(0.31)		(0.34)
PUMA-level contextual controls
% College graduates				0.05				0.03				0.05
				(0.06)				(0.05)				(0.07)
Total pop (logged)				2.33				0.68				3.80
				(1.60)				(2.29)				(2.38)
% Foreign-born				0.11				0.03				0.27
				(0.10)				(0.13)				(0.19)
% Black				−0.09				−0.16	*			−0.01
				(0.05)				(0.07)				(0.07)
% Employed in mnfg industry				0.17	***			0.18	**			0.13
				(0.04)				(0.06)				(0.08)
% Employed in agri/mining industry				0.18	**			0.21	**			0.21
				(0.06)				(0.08)				(0.19)
FDI penetration before 2005 * Korean DI				0.60				0.55				0.65
				(0.88)				(1.22)				(1.24)
% Employed in mnfg industry * Korean DI				0.08				0.02				−0.01
				(0.22)				(0.32)				(0.01)
Population size of PUMA * Korean DI				0.00				−0.04				0.00
				(0.08)				(0.03)				(0.01)
Prior employment rates * Korean DI				0.00				−0.05				0.03
				(0.01)				(0.06)				(0.12)
Intercept		71.28	***	43.55	*	71.28	***	59.72	*	71.28	***	21.50
		(0.24)		(16.83)		(0.35)		(23.88)		(0.35)		(25.13)
N (observations)		480		480		240		240		240		240
N (groups)		60		60		30		30		30		30
Adjusted R-sq		0.32		0.37		0.32		0.38		0.32		0.36

Note: Year dummies are included in all equations. Standard errors in parentheses.

* p < .05; **p < .01; ***p < .001 (two-tailed).

Table 3 presents results from t-test for testing the significance of the difference between OLS estimates of KDI and ODI penetration. Results confirm that estimates of changes in employment rates for KDI inflows are not significantly different from those for ODI inflows. That is, the effects of FDI from each source on changes in employment rates are indistinguishable.

Table 3.

T-Test between OLS Estimates from Difference-In-Differences Models Predicting Change in the Employment Rates.

	Employment
	All		Men		Women
F	1.17	1.08	0.57	0.61	0.57	0.25
Prob > F	0.28	0.30	0.45	0.44	0.45	0.62

Effect of KDI Inflows on Wage Levels

Findings for the models predicting the relationship between inward FDI from different sources and weekly wages show that changes in logged median weekly wages for KDI penetration average smaller than those for NDI (Table 4). This holds true separately for men and women living in areas that attracted KDI inflows; men experience stagnated wage growth by 3.1 percent (p < .001) and women by 2 percent (p < .05) for every increase of one on the scale of KDI penetration. Such negative effects appear only after controlling for socioeconomic contexts. This suggests that a lower growth rate of wages may be attributed to KDI penetration rather than local demographic and labor market conditions. In the case of ODI, no relationship is found between FDI inflows and changes in wages. That is, there is no difference in growth of wage levels for ODI penetration and NDI across years. This applies to both men and women.

Table 4.

OLS Estimates from Difference-In-Differences Models Predicting Change in the Weekly Wage Levels.

		Wage
		All				Men				Women
FDI types (ref = no investment)
	Korean Investment	−0.002		−0.02		−0.01		−0.03	***	−0.01		−0.02	*
		(0.02)		(0.02)		(0.01)		(0.01)		(0.01)		(0.01)
	Other Investment	−0.004		−0.01		−0.001		−0.01		−0.005		0.002
		(0.03)		(0.02)		(0.02)		(0.01)		(0.02)		(0.01)
PUMA-level contextual controls
% College graduates				0.01	***			0.01	***			0.01	***
				(0.002)				(0.001)				(0.001)
Total pop (logged)				0.04				0.20	***			0.09
				(0.06)				(0.05)				(0.05)
% Foreign-born				−0.004				−0.01	*			−0.001
				(0.004)				(0.004)				(0.01)
% Black				−0.001				−0.002	*			0.000
				(0.001)				(0.001)				(0.001)
% Employed in mnfg industry				0.01	***			0.001				−0.003
				(0.002)				(0.003)				(0.003)
% Employed in agri/mining industry				0.01				−0.003				−0.01
				(0.004)				(0.004)				(0.01)
FDI penetration before 2005 × Korean DI				−0.04				−0.05	**			−0.02
				(0.03)				(0.02)				(0.02)
Intercept		6.27	***	5.53	***	6.50	***	4.29	***	6.04	***	4.94	***
		(0.04)		(0.62)		(0.04)		(0.58)		(0.04)		(0.56)
N (observations)		420		420		210		210		210		210
N (groups)		60		60		30		30		30		30
Adjusted R-sq		0.24		0.27		0.18		0.21		0.31		0.33

Note: Year dummies are included in all equations. Standard errors in parentheses.

* p < .05; **p < .01; ***p < .001 (two-tailed).

Table 5 presents results from t-test for testing the significance of the difference between OLS estimates of KDI and ODI penetration. Results confirm that estimates of changes in wage levels for KDI inflows are significantly different from those for ODI inflows, separately for men and women, when accounting for both yearly and contextual effects (p < .05). That is, the effects of KDI on changes in weekly wage levels are statistically smaller than those of ODI.

Table 5.

T-Test between OLS Estimates from Difference-In-Differences Models Predicting Change in the Weekly Wage Levels.

	Income
	All		Men		Women
F	0.01	0.02	0.31	4.39	0.04	4.95
Prob > F	0.94	0.88	0.58	0.04	0.85	0.03

The models of wage present interesting results. KDI inflows result in a lower level of wage growth for men and women on average than NDI and ODI inflows. Specifically, men are paid at least 3 percent less and women 2 percent less for every increase of one on the scale of KDI penetration. In other words, relative to their counterparts in areas without KDI penetration, men and women in KDI areas work for lower pay. The effect of interaction term of FDI penetration prior to KDI and KDI inflows is also noteworthy (Table 4). In addition to KDI penetration being negatively associated with a change in men’s wage levels, the effect is significantly negative for the interaction term. This means that if other foreign automakers have invested in the KDI areas before 2005, changes in wage levels for men average 4.5 percent lower than those for men in areas without KDI penetration (p < .01). That is, men suffer from even lower wages if their residential areas have experienced both KDI and ODI inflows. This also signals that inward KDI itself may be negatively associated with wage growth for both men and women, and in the presence of other foreign automakers, growth rates of wages for men can decrease to an even greater extent for KDI penetration than for ODI penetration.

Discussion and Conclusion

Despite public concern over losing manufacturing jobs to Mexico, NAFTA contributed to creating employment in the U.S. by triggering a wave of foreign automotive investments. Since European automakers entered the U.S. in the mid-1990s, Japanese and Korean producers also invested in the U.S. during the 2000s. This paper sought to investigate how post-NAFTA automotive FDI has impacted the U.S. labor market at the local level. Specifically, we investigated the labor market changes in Alabama between 2005 and 2011, focusing on the extent to which FDI from different sources affected local residents’ employment rates and wage levels.

We hypothesized that FDI originating from a nontraditional source country will have positive employment and negative wage effects, but to a different degree than will FDI originating from traditional source countries. Research shows that early movers tend to have better market positions due to their FSAs based on accumulated knowledge, sophisticated technology, and reputation capital. Lacking such resources, late movers are under the pressure of price competitiveness. In the process of international expansion, their competitiveness based on low production costs is likely to become unavailable, thus requiring them to adopt unique strategies that differ from those implemented by established multinationals (see Guillen and Garcia-Canal 2009). Along this line, we tested our hypotheses based on the assumption that foreign investors in the same industry entering the same host country may show different investment behaviors depending on their country of origin and such strategic disparities can moderate their impacts on the host labor market.

Results showed a positive relationship between KDI and local employment rates, which supports our hypothesis that FDI originating from nontraditional source countries lead to higher employment growth for residents. However, when compared to both NDI and ODI, KDI is associated with a lower increase in wages. In other words, individuals’ chances of employment increase to a greater degree, but individuals’ wages increase to a lesser degree as a result of KDI.

Our findings, supplemented with interview data, suggest that corporate strategies as well as location selection of the firms may have created different outcomes. For example, Korean firms, as late movers in the global automobile industry, had no choice but to compete based on price competitiveness against established competitors with higher levels of FSAs. Thus, the Korean assembler not only chose a lower-cost-oriented county within Alabama to be their production site, but also implemented cost reduction strategies compared to German and Japanese producers. From the analyses, we conclude that the institutional changes can attract FDI of any sources but labor market outcomes for local populations that result from these investments can be distinctive depending on the MNEs’ country of origin and their corporate strategies.

This, in turn, indicates that FDI from nontraditional sources has different implications for the local labor market than FDI from traditional sources does. While KDI has resulted in decreased economic rewards, one should not jump to a conclusion that reverse FDI negatively affects labor market outcomes. Before settlement of Hyundai and its’ suppliers, Montgomery and the adjacent counties had less-advantaged labor market conditions than counties in the northern parts of Alabama where other foreign automakers were located; the labor force in these counties were not only relatively less educated but also less likely to be employed and had lower income. Also, as Alabama is a right-to-work state where employment itself is prioritized over higher wages, there has not been much pressure on net income improvement despite emphasis on an increase in the absolute number of jobs (see Zhao 1998). In other words, reverse FDI may lead to a trade-off between employment and wages due to the MNEs’ efforts to reduce costs. Still, it can contribute to regional development when the location of choice is low growth, low income areas that place more emphasis on job creation than income improvement.

This study makes several contributions to existing literature on labor market inequalities. First, it corroborates that foreign capital is an important predictor that explains labor market changes in the global contexts. It also shows that developed countries are subject to the impacts of foreign investments, especially those originating from nontraditional sources. Compared to the effects of FDI, which flows from developed countries to less developed countries, the effects of reverse FDI have been given insufficient academic attention. However, this study highlights that foreign capital can affect socioeconomic conditions in host countries that had been considered to be more developed than the firms’ country of origin. Specifically, the MNEs may implement various cost-reduction strategies due to their liability of origin, which may lead to a trade-off between job creation and wage increase. Still, the host government can take advantage of reverse FDI inflows in relatively disadvantaged regions within the country.

This also has policy implications for the MNEs engaging in reverse FDI. The four states that competed to win KDI (that is, Alabama, Kentucky, Mississippi, and Ohio) are among the less developed regions in the U.S., although the country itself is economically more advanced than Korea. Also, when searching for a potential production site, Hyundai had important geographical considerations: surplus of cheap, readily available labor and a union-free environment. That is, Hyundai chose a location where its latecomer disadvantages could be easily offset by area-specific contexts. This suggests that late movers may successfully engage in up-market FDI by selecting locations where they can be more favorably received.

Second, this study highlights the importance of collaborative governance for regional development. Researchers argue that late movers often experience difficulties acquiring legitimacy in the host country due to their latecomer disadvantages (see Stopford and Strange 1992). However, this study demonstrates that foreign investments, regardless of their country of origin, can be strongly welcomed when the firms earn host governments’ confidence in their capabilities to contribute to local development. Also, while researchers have focused primarily on FDI attraction policies at the national level to evaluate whether the host country benefits from foreign investments, our findings show that the local governments (for example, state- or county-level) are also deeply involved, playing critical roles in coordinating the foreign investment processes on one hand and intervening in labor markets on the other. In particular, they not only put a lot of effort into attracting FDI, but also ensure local populations benefit from an increased labor demand by implementing other institutional supports such as employee training programs and an anti-illegal immigration act. That is, the host governments, irrespective of their levels, attempt to build cooperative relations with foreign investors for economic development.

Empirically, this study has wider implications outside the U.S. given that Korean automotive firms (i.e., Hyundai-Kia Group) have pursued a very similar investment strategy in Europe. Specifically, they established one assembly plant in Czechia and another in Slovakia, which, on the one hand, allowed the firms to share (mostly Korean) suppliers by taking advantage of geographic proximity between two countries (Courtenay 2006; Korea JoongAng Daily 2006). On the other hand, the firms were able to depress wages by monopolizing the labor, as these locations are distant from traditional regions of the automotive industry (e.g., Germany, France, and Italy) (European Automobile Manufacturers Association 2020). It is also noteworthy that both countries have relatively unfavorable labor market conditions such as high unemployment rates (Eurostat 2020a) and low minimum wage (Eurostat 2020b) compared to other European Union (EU) member states. The position of these countries within the EU, which is similar to that of Alabama and Georgia within the US, suggests that Korean automakers made a strategic location choice to reduce production costs as well as to avoid tariffs that can incur if their products are imported from non-EU member countries.

Despite these theoretical and empirical contributions, it is difficult to assess from this study whether the cases of Korean MNEs in Alabama have wider implications in other U.S. states or in different sectors. Given that FDI inducement policies and socioeconomic contexts vary by state, the applicability of the findings from this study to other states remains in question. For example, many state governments in the South have taken proactive approaches to incentivize automobile assemblers and parts suppliers to compensate for their relative dearth of well-established infrastructure and skilled labor compared to abundance of those resources in the traditional auto states in the Midwest (Hamilton 2004; Lucci 2016; Smith 2003). Korean firms turned such a pro-business environment to their best advantage, which might not have been possible in more labor-friendly states within the automotive industry (e.g., Michigan and Illinois). In the future, it would be interesting to conduct research that covers all the U.S. states to examine the ways in which multinationals from nontraditional sources exercise leverage on the host labor market and to compare under what socioeconomic and institutional contexts their roles are best displayed.

Footnotes

Appendix

Table A1.

Descriptive Statistics of PUMA-Level Control Variables, 2005–2011.

	Total	No FDI	Korean FDI	Other FDI
% College graduates	22.00	26.30	16.30	19.70
Population size (mean)	169,338	176,479	155,957	168,826
% Foreign-born (mean)	3.45	3.68	2.91	3.47
% Black (mean)	20.14	20.50	28.52	15.79
% Employed in manufacturing industry (mean)	21.13	18.90	23.21	22.29
% Employed in agricultural/mining industry (mean)	3.09	2.00	5.13	3.16
% Pre-2005 FDI penetration	40.00	35.40	36.00	67.10
N	210	83	41	86

Acknowledgments

I am very grateful to the following colleagues who have offered helpful comments on earlier drafts: Mauro Guillen, Emilio Parrado, Janice Madden, Ann Harrison, and two anonymous reviewers.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Eunbi Kim

Notes

References

Aaron

1999. The Contribution of FDI to Poverty Alleviation. Washington, DC: Foreign Direct Investment Advisory Service.

Alabama Department of Commerce. 2013. Alabama’s Hyundai Suppliers and Vendors. Accessed September 29, 2020. https://www.madeinalabama.com/assets/2013/01/Hyundaimaplist3-15-13.pdf.

Bagchi-Sen

1991. “Employment in Foreign Owned Manufacturing Firms in the United States: The Impact of Modes of Entry.” Tijdschrift voor Economische en Sociale Geografie 82:282–94. doi: 10.1111/j.1467-9663.1991.tb00807.x.

Bagchi-Sen

1995. “Foreign Direct Investment in U.S. Manufacturing Industries: Source-specific Variations.” Human Geography 77 (1): 17–29. doi: 10.1080/04353684.1995.11879678.

Bank of Korea. 1997–2007. “Exchange Rate System.” Accessed August 8, 2020. https://www.bok.or.kr/eng/main/contents.do?menuNo=400186.

Baran

1956. The Political Economy of Growth. New York: Monthly Review Press.

Barth

Bryson

Davis

J. C.

Freeman

. 2016. “It’s Where You Work: Increases in the Dispersion of Earnings across Establishments and Individuals in the United States.” Journal of Labor Economics 34 (2): 67–97. doi: 10.1086/684045.

Bartlett

C. A.

Ghoshal

. 2000. “Going Global: Lessons from Late Movers.” Harvard Business Review 78 (2): 132–42.

Blomström

Kokko

. 2003. “The Economics of Foreign Direct Investment Incentives.” NBER Working Paper 9489, NBER, Cambridge, MA.

10.

Bloom

D. E.

Canning

Fink

. 2010. “Implications of Population Ageing for Economic Growth.” Oxford Review of Economic Policy 26 (4): 583–612. doi: 10.1093/oxrep/grq038.

11.

Bloom

D. E.

Freeman

R. B.

. 1986. “The Effects of Rapid Population Growth on Labor Supply and Employment in Developing Countries.” Population and Development Review 12 (3): 381–414. doi: 10.2307/1973216.

12.

Borjas

G. J.

1987. “Immigrants, Minorities, and Labor Market Competition.” ILR Review 40 (3): 382–92. doi: 10.1177/001979398704000305.

13.

Bornschier

Chase-Dunn

Rubinson

. 1978. “Cross-national Evidence of the Effects of Foreign Investment and Aid on Economic Growth and Inequality.” American Journal of Sociology 84:651–83. doi: 10.1086/226831.

14.

Brown

Ingram

Wadsworth

. 2004. “The Price Is Right? Pay Settlements and Nominal Wage Rigidity in Britain.” British Journal of Industrial Relations 42 (3): 507–25. doi: 10.1111/j.1467-8543.2004.00327.x.

15.

Brown

Raines

. 2000. “The Changing Nature of Foreign Investment Policy in Europe: From Promotion to Management.” In Regions, Globalization and the Knowledge-Based Economy, edited by Dunning

J. H.

, 106–19. Oxford, UK: Oxford University Press.

16.

Card

De la Rica

. 2006. “The Effect of Firm Level Contracts on the Structure of Wages: Evidence from Matched Employer-employee Data.” Industrial and Labor Relations Review 59 (4): 573–92.

17.

Clerides

S. K.

Lach

Tybout

J. R.

. 1998. “Is Learning by Exporting Important? Micro-dynamic Evidence from Colombia, Mexico, and Morocco.” Quarterly Journal of Economics 113:903–47. doi:10.1162/003355398555784.

18.

Cole

R. E.

Deskins

D. R.

. 1988. “Racial Factors in Site Location and Employment Patterns of Japanese Auto Firms in America.” California Management Review 31 (1): 9–22. doi: 10.2307/41166535.

19.

Courtenay

Vince

. 2006. Hyundai Czech Plant Gets OK; Kia Slovakia in Trial Production. WardsAuto. Accessed August 8, 2020. https://www.wardsauto.com/news-analysis/hyundai-czech-plant-gets-ok-kia-slovakia-trial-production.

20.

Crescenzi

Iammarino

. 2017. “Global Investments and Regional Development Trajectories: The Missing Links.” Regional Studies 51 (1): 97–115. doi: 10.1080/00343404.2016.1262016.

21.

Cuervo-Cazurra

Genc

. 2008. “Transforming Disadvantages into Advantages: Developing Country MNEs in the Least Developed Countries.” Journal of International Business Studies 39 (6): 957–79. doi: 10.1057/palgrave.jibs.8400390.

22.

De Beule

Elia

Piscitello

. 2013. “Entry and Access to Competencies Abroad: Emerging Market Firms versus Advanced Market Firms.” Journal of International Management 20 (2): 137–52. doi: 10.1016/j.intman.2013.05.002.

23.

De Gier

2010. “Paradise Lost Revisited: GM and the UAW in Historical Perspective.” Visiting Fellow Working Papers . Cornell University ILR School, Ithaca, NY.

24.

Dickens

W. T.

Katz

L. F.

. 1986. “Interindustry Wage Differences and Industry Characteristics.” NBER Working Paper No. 2014 , NBER, Cambridge, MA.

25.

Eisenstein

P. A.

1991. “Lost Hyundai Maps Road Back.” Chicago Tribune. Accessed August 10, 2020. https://www.chicagotribune.com/news/ct-xpm-1991-10-20-9104040535-story.html.

26.

Evans

1979. Dependent Development: The Alliance of Multinational, State and Local Capital. Princeton, NJ: Princeton University Press.

27.

Evans

Timberlake

. 1980. “Dependence, Inequality, and the Growth of the Tertiary: A Comparative Analysis of Less Developed Countries.” American Sociological Review 45 (4): 531–52. doi: 10.2307/2095006.

28.

European Automobile Manufacturers Association. 2020. Interactive Map: Fiscal Income from Motor Vehicles in the EU-15, by Country. Accessed August 8, 2020. https://www.acea.be/statistics/article/interactive-map-fiscal-income-from-motor-vehicles-in-the-eu-15-by-country.

29.

Eurostat. 2020a. Unemployment Statistics. Accessed August 8, 2020. https://ec.europa.eu/eurostat/statistics-explained/index.php/Unemployment_statistics.

30.

Eurostat. 2020b. Minimum Wage Statistics. Accessed August 8, 2020. https://ec.europa.eu/eurostat/statistics-explained/index.php/Minimum_wage_statistics.

31.

Firebaugh

1992. “Growth Effects of Foreign and Domestic Investment.” American Journal of Sociology 98 (1): 105–30. doi: 10.1086/229970.

32.

Firebaugh

Goesling

. 2004. “Accounting for the Recent Decline in Global Income Inequality.” American Journal of Sociology 110 (2): 283–312. doi: 10.1086/421541.

33.

Firn

J. R.

1975. “External Control and Regional Development: The Case of Scotland.” Environment and Planning A 7:393–414. doi: 10.1068/a070393.

34.

Gerlach

Stephan

. 2006. “Pay Policies of Firms and Collective Wage Contracts: An Uneasy Partnership?” Industrial Relations 45 (1): 47–63. doi: 10.1111/j.1468-232X.2006.00413.x.

35.

Groshen

E. L.

1991. “Sources of Intra-industry Wage Dispersion: How Much Do Employers Matter?” Quarterly Journal of Economics 106 (3): 869–84. doi: 10.2307/2937931.

36.

Guillen

M. F.

Garcia-Canal

. 2009. “The American Model of the Multinational Firm and the “New” Multinationals from Emerging Economies.” Academy of Management Perspectives 23 (2): 23–35. doi: 10.5465/amp.2009.39985538.

37.

Hallwood

1986. “The Offshore Oil Supply Industry in Aberdeen: The Affiliates, Their Characteristics and Importance.” North Sea Study Occasional Paper No. 23 , Department of Political Economy, University of Aberdeen, Aberdeen, UK.

38.

Hamilton

F. L.

2004. “The Alabama Industrial Incentives Offered to Hyundai.” Economic Development Journal 3 (3): 40–48.

39.

Harrington

J. W.

Burns

Cheung

. 1986. “Market-oriented Foreign Investment and Regional Development: Canadian Companies in Western New York.” Economic Geography 62:155–66.

40.

Harris

A. H.

Lloyd

M. G.

Newlands

D. A.

. 1988. The Impact of Oil on the Aberdeen Economy. Aldershot, UK: Avebury.

41.

Harrison

McMillan

. 2011. “Offshoring Jobs? Multinationals and U.S. Manufacturing Employment.” Review of Economics and Statistics 93 (3): 857–75. doi: 10.1162/REST_a_00085.

42.

Harrison

Scorse

. 2010. “Multinationals and Anti-sweatshop Activism.” American Economic Review 100 (1): 247–73. doi: 10.1257/aer.100.1.247.

43.

Holmes

Rusonik

. 1991. “The Break-up of an International Labour Union: Uneven Development in the North American Auto Industry and the Schism in the UAW.” Environment and Planning A: Economy and Space 23 (1): 9–35. doi: 10.1068/a230009.

44.

Iyanda

1999. “The Impact of Multinational Enterprises on Employment, Training and Regional Development in Namibia and Zimbabwe: A Preliminary Assessment Multinational Enterprises Programme.” Multinational Enterprises Programme Working Paper 84, International Labour Office, Geneva, Switzerland.

45.

Johanson

Vahlne

. 1977. “The Internationalization Process of the Firms: A Model of Knowledge Development and Increasing Foreign Market Commitments.” Journal of International Business Studies 8 (1): 23–32. doi: 10.1057/palgrave.jibs.8490676.

46.

Johnson

J. R.

1993. “NAFTA and Trade in Automotive Goods.” In Assessing NAFTA: A Trilateral Analysis, edited by Globerman

Walker

, 87–129. Vancouver, BC: The Fraser Institute.

47.

Jones

2017. “Agglomeration Economies and the Location of Foreign Direct Investment: A Meta-analysis.” Journal of Regional Science 57 (5): 731–57. doi: 10.1111/jors.12335.

48.

Katz

H. C.

Meltz

N. M.

. 1991. “Profit Sharing and Auto Workers’ Earnings: The United States vs Canada.” Industrial Relations 46 (3): 515–30. doi: 10.7202/050693ar.

49.

Kerr

S. P.

Kerr

Özden

Ç.

Parsons

. 2017. “High-skilled Migration and Agglomeration.” Annual Review of Economics 9:201–34. doi: 10.1146/annurev-economics-063016-103705.

50.

Korea JoongAng Daily. 2006. “Kia’s Factory in Slovakia to Supply Hyundai Motors.” Accessed August 8, 2020. https://koreajoongangdaily.joins.com/2006/12/08/economy/Kias-factory-in-Slovakia-to-supply-Hyundai-motors/2855441.html.

51.

Little

J. S.

1986. “The Effects of Foreign Direct Investment on U.S. Employment during Recession and Structural Change.” New England Economic Review (Nov):40–8.

52.

Llull

2018. “The Effect of Immigration on Wages: Exploiting Exogenous Variation at the National Level.” Journal of Human Resources 53:608–62. doi: 10.3368/jhr.53.3.0315-7032R2.

53.

Lopez-de-Silanes

Markusen

J. R.

Rutherford

T. F.

. 1994. “The Auto Industry and the North American Free-trade Agreement.” In Modelling Trade Policy: AGE Assessments of North American Free Trade, edited by Shields

Francois

, 223–55. London, UK: Cambridge University Press.

54.

Lucci

2016. Michigan vs. Illinois: Manufacturing and Out-migration. Chicago: Illinois Policy.

55.

Luo

Tung

R. L.

. 2007. “International Expansion of Emerging Market Enterprises: A Springboard Perspective.” Journal of International Business Studies 38:481–98. doi: 10.1057/palgrave.jibs.8400275.

56.

Mair

Florida

Kenney

. 1988. “The New Geography of Automobile Production: Japanese Transplants in North America.” Economic Geography 64 (4): 352–73. doi: 10.2307/144233.

57.

Markusen

J. R.

Venables

A. J.

. 2000. “The Theory of Endowment, Intra-industry and Multi-national Trade.” Journal of International Economics 52 (2): 209–34. doi: 10.1016/S0022-1996(99)00055-0.

58.

Mathews

J. A.

2002. Dragon Multinationals: A New Model of Global Growth. New York: Oxford University Press.

59.

Meyer

K. E.

Mudambi

Narula

. 2011. “Multinational Enterprises and Local Contexts: The Opportunities and Challenges of Multiple Embeddedness.” Journal of Management Studies 48 (2): 235–52. doi: 10.1111/j.1467-6486.2010.00968.x.

60.

Monastiriotis

Jordaan

J. A.

. 2010. “Does FDI Promote Regional Development? Evidence from Local and Regional Productivity Spillovers in Greece.” Eastern Journal of European Studies 1 (2): 139–64.

61.

Morgan

2008. “The True Price of Auto Labor Costs.” CBS News. Accessed August 9, 2020. https://www.cbsnews.com/news/the-true-price-of-auto-labor-costs/.

62.

Ndofor

H. A.

Priem

R. L.

. 2011. “Immigrant Entrepreneurs, the Ethnic Enclave Strategy, and Venture Performance.” Journal of Management 37 (3): 790–818. doi: 10.1177/0149206309345020.

63.

Peri

2016. “Immigrants, Productivity, and Labor Markets.” Journal of Economic Perspectives 30 (4): 3–30. doi: 10.1257/jep.30.4.3.

64.

Phelps

N. A.

Raines

. 2003. The New Competition for Inward Investment: Companies, Institutions, and Territorial Development. Cheltenham, UK: Edward Elgar.

65.

Ramamurti

2009. “Why Study Emerging-market Multinationals?” In Emerging Multinationals in Emerging Markets, edited by Ramamurti

Singh

J. V.

, 3–22. New York: Cambridge University Press.

66.

Reid

1990. “Spatial Patterns of Japanese Investment in the US Automobile Industry.” Industrial Relations Journal 21 (1): 49–59. doi: 10.1111/j.1468-2338.1990.tb00086.x.

67.

Rubin

B. A.

1996. Shifts in the Social Contract: Understanding Change in American Society. Thousand Oaks, CA: Sage.

68.

Ruggles

Alexander

Genadek

Goeken

Schroeder

Sobek

. 2010. Integrated Public Use Microdata Series: Version 5.0. Minneapolis, MN: University of Minnesota.

69.

Sinek

1994. “Hyundai Shutters 2 Canada Plants: 840 Workers at Bromont Let Go.” Automotive News. Accessed August 10, 2020. https://www.autonews.com/article/19940328/ANA/403280711/hyundai-shutters-2-canada-plants-840-workers-at-bromont-let-go.

70.

Smith

D. C.

2003. “Hyundai’s U.S. Hopes Ride on Hope Hull.” WardsAuto. Accessed August 8, 2020. https://www.wardsauto.com/news-analysis/hyundais-us-hopes-ride-hope-hull.

71.

Sonn

J. W.

Lee

. 2012. “Revisiting the Branch Plant Syndrome: Review of Literature on Foreign Direct Investment and Regional Development in Western Advanced Economies.” International Journal of Urban Sciences 16 (3): 243–59. doi: 10.1080/12265934.2012.733589.

72.

Stopford

Strange

. 1992. Rival States, Rival Firms: Competition for World Market Shares. Cambridge, UK: Cambridge University Press.

73.

Suarez

Lanzolla

. 2005. “The Half-truth of First-mover Advantage.” Harvard Business Review 83 (4): 121–27, 134.

74.

Thorpe

2005. “Quebec Mistakes Still Haunt Hyundai: Bromont Was Recipe for How Not to Build Cars in North America.” Automotive News. Accessed August 10, 2020. https://www.autonews.com/article/20050613/SUB/506130710/quebec-mistakes-still-haunt-hyundai.

75.

Tuman

J. P.

Erlingsson

. 2019. “Foreign Direct Investment Flows to Mexican States: A Study of the Automobile Industry, 2004–2014.” Growth and Change 50 (3): 1164–84. doi: 10.1111/grow.12304.

76.

Ulgado

F. M.

1996. “Location Characteristics of Manufacturing Investments in the U.S.: A Comparison of American and Foreign-based Firms.” Management International Review 36 (1): 7–26. doi: 10.1257/jep.28.1.3.

77.

Waldkirch

Nunnenkamp

Bremont

J. E. A.

. 2009. “Employment Effects of FDI in Mexico’s Non-Maquiladora Manufacturing.” Journal of Development Studies 45 (7): 1165–83. doi: 10.1080/00220380902952340.

78.

Wallerstein

1974. The Modern World-system. London, UK: Academic Press.

79.

Watts

1981. The Branch Plant Economy: A Study of External Control. London, UK: Longman.

80.

Wren

Taylor

. 1999. “Industrial Restructuring and Regional Policy.” Oxford Economic Papers 51:487–516. doi: 10.1093/oep/51.3.487.

81.

Yeaple

S. R.

2003. “The Complex Integration Strategies of Multinationals and Cross Country Dependencies in the Structure of FDI.” Journal of International Economics 60 (2): 293–314. doi: 10.1016/S0022-1996(02)00051-X.

82.

Yeaple

S. R.

2009. “Firm Heterogeneity and the Structure of U.S. Multinational Activity.” Journal of International Economics 78 (2): 206–15. doi: 10.1016/j.jinteco.2009.03.002.

83.

Yildiz

H. E.

2014. “Not All Differences Are the Same: Dual Roles of Status and Cultural Distance in Sociocultural Integration in Cross-border M&As.” Journal of International Management 20 (1): 25–37. doi: 10.1016/j.intman.2013.03.014.

84.

Zhang

Zhao

S. X.

. 2007. “Geographical Changes in Foreign Direct Investment and Impacts on Regional Economic Integration in China: The Case Study of the Pan-PRD Regionalization.” The China Review 7 (2): 191–219.

85.

Zhang

Felmingham

. 2002. “The Role of FDI, Exports and Spillover Effects in the Regional Development of China.” Journal of Development Studies 38 (4): 157–78. doi: 10.1080/00220380412331322451.

86.

Zhao

1998. “The Impact of Foreign Direct Investment on Wages and Employment.” Oxford Economic Papers 50 (2): 284–301. doi: 10.1093/oxfordjournals.oep.a028646.