Industry value added and employment of migrant workers

Industry value added and employment of migrant workers Elena Gentili March, 2018 Draft version Abstract This paper investigates how the competitive structure of an industry influences the employment of migrant workers. The theoretical model combines an industrial monopolistic competitive structure with labor market frictions and workers who are heterogeneous in their reservation wages. If on average migrant workers have lower reservation wages than non-migrant workers and there are no labor shortages, more competitive industries tend to disproportionally hire more migrant workers to decrease their marginal costs. The empirical analysis is conducted at industry level with US manufacturing industry data. Industry value added is instrumented through a structural change in trade policy that differentially impacted industry competitiveness: the entry of China in the WTO. Results confirm the theoretical predictions, with more competitive industries employing more migrant workers and paying lower wages. Keywords: Migration, Monopolistic competition, Industry value added. JEL codes: F16, J23, J60, L13. Corresponding author: Elena Gentili, Institute of Economics, via Buffi 13, 6904 Lugano, Switzerland (elena.gentili@usi.ch). Institute of Economics (IdEP), Università della Svizzera Italiana (USI), Switzerland.

1 Introduction This paper analyzes the relationship between the competitive structure of an industry and the employment of migrant workers. Particularly, while a large fraction of the recent migration literature focuses on the impact of immigration on several economic outcomes (Dustmann et al., 2017; Ottaviano and Peri, 2012; Borjas and Katz, 2007; Borjas, 2003; Card, 2001 among the others), the aim of this paper is to understand what is the role of the industrial structure in shaping the demand for migrant workers. The literature about the role of labor demand in determining migration inflows can be traced back to Piore (1979). However, while descriptively examining how labor demand can influence migration inflows, this literature does not attempt to understand the relationship between the market conditions faced by the firms and the demand for migrant workers. Conversely, this paper attempts to quantify this relationship providing both a theoretical model and some empirical evidence about the impact of industry competition on the employment of migrant workers. The theoretical model combines the model of monopolistic competition developed by Melitz and Ottaviano (2008) with classical search models as discussed in Eckstein and Van den Berg (2007). Particularly, I consider an economy where many firms produce different varieties of the same good, workers are heterogeneous in their reservation wages and there are search frictions in the labor market. The optimal number of firms in the market, as well as firm profits and markups are endogenously determined by the distribution of reservation wages. The aim of the model is to show that in the presence of workers that are homogeneous in their productivity level and heterogeneous in their reservation wages, firms that work in lower markup industries have a stronger incentive to depress their marginal costs and, in turn, the wages of workers. Assuming that immigrant workers have lower reservation wages, stronger pressures for lower marginal costs induce firms to disproportionally hire more immigrant workers. To empirically test the theoretical model s predictions, I exploit US data. First, I regress the share of migrant workers on a measure of industry competition, i.e. industry value added. However, since the composition and the extent of migration inflows may influence industry competition, OLS estimates may be biased by reverse causality. As a result, to isolate the impact of industry value added on the employment of migrant workers, I implement an instrumental variable approach. In particular, I instrument industry value added with a structural break in trade policy that changed the competitive pressures in some industries but not in others: the official entry of China in the WTO agreements. Prior to entrance in the WTO, China was already granted a Normal Trade Relationship (NTR) status by the US. This NTR status entailed the application of more favorable tariffs with respect to non-ntr countries but had to be renewed every year through Presidential 1

extensions. Entering in the WTO agreements, the Chinese NTR status became permanent, ending the uncertainty about the application of NTR tariffs and boosting international trade. Thus, the entry of China in the WTO disproportionally increased the competition faced by more protected industries but, assuming that all the effects on employment are mediated by this change in competitive pressures, did not directly affect migration inflows. Pierce and Schott (2016) document the effects of the entry of China in the WTO on employment implementing a diff-in-diff strategy. In this paper, I first replicate their exercise on the share of foreign workers. Then, I use their diff-in-diff term to instrument industry value added. Even though increasing demand for cheap labor inputs is just a way that firms have to overcome competitive pressures (firms can also offshore or outsource, see for example Ottaviano et al. (2013) and Grossman and Rossi-Hansberg (2012) for offshoring and Antras and Helpman (2004) and Grossman and Helpman (2002) for outsourcing), I find that a 1% increase in industry value added (i.e., a 1% decrease in competitiveness) increases the employment of foreign workers with respect to native workers of about 1%. This result holds for both low and highly educated migrants and seems to be due to a decrease in wages paid to workers in more competitive industries. In addition, the reduced form equations show the expected signs, with more protected industries facing a stronger increase in the employment of foreign workers. All the results hold controlling for the technological content of industries. This paper adds to several strands of the literature. First, it adds to the literature on migration, providing evidence of a possible channel through which pull factors determine migration flows. Second, it adds to the literature on industrial organization, providing evidence about the effects that different degrees of competition may have on employment decisions. Third, it indirectly adds to the trade literature and how changes in trade policy can impact the labor market structure of the countries involved. The remaining of the paper is structured as follows. The following section introduces the theoretical model. Then, the third and the fourth sections describe the data and explain in detail the empirical strategy adopted. Finally, Section 5 presents the results and Section 6 concludes. 2 Theoretical model The main goal of this section is to uncover the relationship between industry value added and variations in the average wages paid by the firms. In particular, I am going to show that firms in industries with lower value added are upward constrained in setting wages. The mechanism works as follows. The model encompasses a labor market with search frictions, workers who are heterogeneous in their value of home production and a monopolistic competitive structure of the 2

final good markets. Final good markets represent different industries and each final good market comprises different varieties of the final good. Given the monopolistic competitive setting, the maximum price each firm can charge to stay on the market depends on some structural characteristics, such as barriers to entry, bounds of the distribution of accepted wages, and degree of differentiation of good varieties within each market. These structural characteristics are the same for all the firms within the market. Within each market, firms pay different wages because of workers heterogeneity in reservation wages. Moreover, because of search frictions, firms can pay workers below their marginal productivity and workers may accept this wage offers to avoid the cost of future job search. Across different final good markets, firms with higher threshold prices can afford paying higher wages to their workers. Thus, if immigrant workers are a subset of the population with lower values of home production, industries with lower value added will disproportionately hire more immigrant workers. 2.1 Final good market Preferences and output demand As in Melitz and Ottaviano (2008), consider an industry with many firms producing differentiated varieties i Ω of the same good. For simplicity, I focus on a single industry, even though the following theoretical framework may be extended to encompass all the industries in the economy. Each individual participates in the labor market and provides either 1 unit of labor, if employed, or 0 unit of labor, if unemployed. For simplicity, I assume that workers are homogeneous in their productivity and each worker produces only one unit of output. Employed workers earn a wage ω c, while unemployed workers earn their value of home production v c. Accepted wages are distributed according to F (w c ), with w c [ w, w] while home production values across unemployed workers are distributed according to G(v c ) with v c [ v, v]. Notice that the distributions of accepted wages and home production values may overlap. Consumers share the same utility function over preferences for varieties: U c = q0 c + α qi c di 1 i Ω 2 γ (qi c ) 2 di 1 ( 2 i Ω 2 η qi di) c (1) i Ω where q c 0 is the numeraire good, and q i is the consumption of other varieties. The parameters α and η are demand shifters with respect to the numeraire good, while the parameter γ indexes the degree of differentiation between varieties. Normalizing the maximum working time to 1, the value of consumption is equal to ω c (1 q0 c) for the employed and to vc (1 q0 c ) for the unemployed. If each consumer demands a positive quantity of the numeraire good, the inverse demand for each 3

variety i is given by: p i = α γq c i ηq c (2) Focusing on the subset of varieties i with positive consumption, the individual demands for variety i are: q c i = α ηn + γ 1 γw c p i + ηn ηn + γ 1 p (3) γwc where N are the consumed varieties and p = (1/N) i Ω p i di is the average price. For unemployed consumers the wage w c is replaced by the value of home production v c. The aggregate demand for variety i in the economy can be derived integrating the individual demands over F (w c ) and G(v c ): q i = α ηn + γ φ ψ γ p i + ηn ψ p (4) ηn + γ γ where φ = w c F (w) df (wc )+ v c G(v) dg(vc ) and ψ = w c F (w) (1/wc )df (w c )+ v c G(v) (1/vc )dg(v c ). The threshold price p M above which the demand for variety i goes to 0 can be found setting the previous equation equal to 0 and solving for p i : p M = γα ψ(ηn + γ) φ + ηn p ηn + γ Thus, the threshold price is a function of the overall number of firms in the market, the average price, and the distribution of wealth across consumers. (5) Production Consider a production structure in which labor is the only production factor and there is only one type of labor required in the production process. To enter the market, firms face an entry cost f E for research and development. Thus, only firms able to cover f E decide to enter the market. After entering the market, firms hire workers for production. If every worker can only produce one unit of output, the marginal cost of each unit produced is the marginal wage of the last worker w. Each firm considers the number of firms and the average price in the market as given and maximizes its profits according to the residual good demand in Equation (4). Equating marginal revenues to marginal costs, the optimal price and the optimal quantity must satisfy: q(w) = ψ [p(w) w] (6) γ Notice that since every worker produces only one unit of output, the number of workers in the firm l(w) coincides with the output produced, that is l(w) = q(w). Now, as in Melitz and Ottaviano (2008), it is possible to write the optimal price, quantity, profits and markups as functions of average wages and threshold prices: p(w) = 1 2 (p M + w) (7) 4

q(w) = ψ 2γ (p M w) (8) π(w) = ψ 4γ (p M w) 2 (9) µ(w) = 1 2 (p M w) (10) Thus, an increase in the threshold price p M has a positive effect on all the four variables under consideration. On the other hand, an increase in the wage paid by the firm has a positive effect on price and a negative effect on quantity produced. Rearranging Equation (10), it is straightforward to observe that ceteris paribus higher threshold prices imply higher wages. Specifically, while higher wages translate into lower markups within the same industry (i.e. holding p M fixed), firms in industries with higher threshold prices are able to set higher wages for a given markup. Equilibrium From the maximum price condition in (5), replacing p = (p M + w)/2, the maximum number of firms on the market is: N = 2γ η α p M p M w where w = p M 0 wdf (w)/f (p M ) is the average wage in the economy. 1 Under free entry, only the firms with profits greater than the fixed costs f E enter the market and the zero profit condition for the marginal firm is: pm 0 π(w)df (w) = ψ 4γ pm 0 (11) (p M w) 2 df (w) = f E (12) As in Melitz and Ottaviano (2008), to proceed with the analysis, I assume a specific distribution for the accepted wage distribution F (w). For simplicity, I assume a uniform distribution. As long as the maximum value of the accepted wage distribution w is larger than the threshold price, the threshold price is: [ 12γ( w E p M = w)f ψ Thus, firms in industries with structurally higher barriers to entry and more differentiated goods experience larger markups and are able to set higher wages than industries with smaller barriers to entry. 1 Notice that for w > α, N is an increasing function of the threshold price p M. ] 1 3 (13) 5

2.2 Labor market Supply side In the labor market there are search frictions. In every period, unemployed workers receive wage offers at arrival rate λ, and existing jobs are destroyed at rate δ. Wage offers are distributed according to a generic distribution H(w), with w [ w, w]. If the worker accepts the wage offer, then she works with the same firm until the job is destroyed and she goes back to unemployment. If the worker rejects the wage offer, she keeps on searching for wage offers on the market. To model workers search behavior, I refer to the baseline model of search frictions by Eckstein and Van den Berg (2007) with no search on the job. Each worker compares the present value of being employed, V e, with the present value of being unemployed, V u, and stops searching when the first value outweighs the second. Thus, the worker is indifferent between accepting or refusing the wage offer when: { } ρv u = v c + λ w max[0, V e (w) V u ]dh(w) δ w where ρ is a discount factor, ρv u is the present value of unemployment, and the RHS is the present value of search. The worker stops searching when she finds an offer w which is larger than her reservation wage w. If she rejects the offer she obtains the value of home production, v c. Since the expected value of being employed is V e (w) = 0 e (ρ+δ)t wdt = w/(ρ + δ), the value of search for the worker can be stated as: { } w = v c + λ w [w w ]dh(w) (15) ρ + δ w Demand side Firms post their wage offers and hire workers for production. The optimal employment level for firms is equal to the optimal quantity derived in the previous subsection, i.e. l(w) = ψ 2γ [p(w) w]. Moreover, since the search frictions in the labor market make the job search costly for workers, firms can post wages below the value of the marginal productivity of workers, i.e. w < p(w). Thus, firms find it optimal to hire workers until the optimal quantity of production is attained. (14) Equilibrium In equilibrium the flow of workers to employment should be equal to the flow of workers to unemployment. Indexing with u the number of workers in unemployment and with m the number of workers in the economy, this is equivalent to require that: λu = δ(m u) (16) 6

Moreover, in equilibrium, the stock of employed workers at wage w is F (w)(m u). The flow into this stock is given by the share of wage offers received by unemployed workers times the probability that these offers will be above the workers reservation wages, i.e. λuh(w)p (w > w ). To the same extent, the flow of workers out of employment is given by the fraction of jobs destructed in every period, i.e. δf (w)(m u). Equating these two flows and exploiting the condition in the previous equation, the wage offer distribution is the ratio between the accepted wage distribution and the complementary of the values of home production distibution 2 : H(w) = F (w) 1 G(w ) (17) Thus, the distribution of accepted wages is different from the distribution of wage offers, because of workers heterogeneity in reservation wages. However, the distribution of wage offers can be recovered specifying a distribution for accepted wages and a distribution for reservation wages. Finally, notice that in equilibrium the stock of workers employed in the economy (m u)f (p M ) should be equal to the aggregate quantity produced in the economy. Equating the stock of employed workers to the total quantity produced, it is possible to derive the number of employed workers in equilibrium (m u) as a function of the threshold price p M and the number of firms in the market N: m u = ψ 4γ p 2 M p M w N (18) As expected, an increase in the threshold price increases the number of employed workers in the economy. Discussion This setting predicts that firms in industries with larger barriers to entry and more differentiated varieties can charge higher markups after entering the final good market. This translates in the possibility of offering higher wages to workers and in the hiring of a larger workforce in equilibrium. If migrant workers have lower reservation wages, they will be disproportionately hired in industries with lower markups. Notice that the implications of this model are theoretically equivalent to a model in which individuals differ in productivity levels but have same reservation wages. If native workers are more productive than immigrant workers because of native-specific human capital, they will be hired more in larger markup firms, which are willing to pay higher wages. However, within this framework I prefer to stress differences in reservation wages than in productivity levels. Indeed, migrant workers may have different reference points in terms of wage levels. As highlighted by Falk 2 Notice that the optimal reservation wage is just a linear transformation of values of home productions. Thus, reservation wages are distributed in the same way. 7

et al. (2006), reference points may be fundamental in determining what is regarded as a fair wage. This may induce migrant workers to accept lower wages with respect to native workers. This model encompasses a closed economy. However, the identification strategy in the empirical part rests on a structural change in international trade policy, i.e. the entry of China in the WTO. Melitz and Ottaviano (2008) provide an exhaustive discussion of the impact of trade in their monopolistic competition framework, i.e. assuming homogeneous labor force. Indeed, the implications of their model assuming workers heterogeneity in reservation wages are quite similar. If preferences and technology are the same, the arrival of a trading partner increases competition in the final good market and decreases the maximum threshold price. Thus, only firms paying lower wages are able to stay on the market. This decreases employment and induces firms to decrease their wage offers, driving workers with higher reservation wages out of employment. Finally, notice that the model presumes an excess supply of workers in the labor market (i.e. there is unemployment). In fact, this analysis may not be true in the presence of labor shortages. If there are labor shortages, the hiring of migrant workers is driven by the need of attaining the optimal production level rather than by low industry markups. 3 Data To test the implications of the theoretical model above, I need to combine different data sources. First of all, to compute industry value added, I need detailed budget industry data. To this end, I focus on the United States and exploit the data from the NBER-CES Manufacturing Industry Database by Becker et al. (2013). Then, for the instrumental variable approach, I exploit a structural change in trade policy that impacted the employment of migrant workers only through the change in value added, i.e. the entrance of China in the WTO (more details in Section 4.2). To this end, I exploit tariff data from the NBER Trade Database by Feenstra et al. (2002). Finally, I need information on the nationality and/or migrant status of workers in the manufacturing sector. I obtain this information from the American Community Survey (ACS) and the 1990 Census. Then, I merge individual level data to the NBER-CES Manufacturing Industry Database through the industry of work. The final dataset includes year 1990 and years between 2000 and 2011. The following subsections discuss these databases in more detail. 3.1 NBER-CES Manufacturing Industry Database The NBER-CES Manufacturing Industry Database by Becker et al. (2013) is a panel dataset at industry level. It is constructed combining information from the Economic Census and the Annual 8

Survey of Manufactures and spans between 1958 and 2011. Industry codes are harmonized and coded according to the 6-digit North American Industry Classification System (NAICS). For each industry and year, it reports data on number of workers employed, number of workers in production, value of shipments, value added, cost of materials, cost of energy and other fuels, wages, investments in capital, inventories and real capital stock. Moreover, the authors provide interesting additional information about price deflators of investments, value of shipments and cost of materials. In the NBER-CES Manufacturing Industry Database, value added is defined as the difference between the value of shipments and the cost of materials, supplies, containers, fuel, purchased electricity and paid work and is adjusted for inventories and value added from merchandising operations. I further adjust the value added according the Consumer Price Index issued by the Federal Reserve. 3 From this database, I also extract two other control variables: the real capital stock normalized by the value of shipments and the share of production workers out of total employment. Panel A of Table 1 shows the descriptive statistics for these variables. 3.2 NBER Trade Database Data on US tariffs are drawn from the NBER Trade Database by Feenstra et al. (2002). They report ad valorem equivalent tariff rates on traded goods between 1989 and 2001. Traded goods are coded according to the 8-digit Harmonized System (HS). Since I am interested in the difference between non-ntr tariffs and NTR tariffs, I first compute this tariff gap for every product. Then, to link the information on tariffs to the industries where the traded goods are produced, I use the concordance table provided by the authors (see Pierce and Schott, 2012, for more details). Finally, I collapse tariff gaps by industry and merge them to the NBER-CES Manufacturing Industry Database. Figure 1 shows the trends in employment, share of workers in production, capital intensity and value added for more protected and less protected industries. The solid line depicts industries where the tariff gap is above the median, while the dashed line depicts industries where the tariff gap is below the median. It is possible to observe that more protected industries faced stronger employment losses after 2001, stronger reductions in the share of workers in production and stronger increases in capital intensity. These results are in line with previous literature (Pierce and Schott, 2016; Acemoglu et al., 2016; Autor et al., 2014; Ebenstein et al., 2011), suggesting that the decrease in employment after the entrance of China in the WTO was paired by an increase in offshoring of US manufacturing firms and a shift towards more capital-intensive products. It is also interesting to notice that more protected industries are the ones with lower average value added, and they faced a sharper decrease in value added after 2001. 3 Available at: https://fred.stlouisfed.org/series/cpaltt01usa661s 9

3.3 ACS and 1990 Census data The American Community Survey (ACS) contains data about the nationality of workers, the industry in which they work, their occupations and their annual earned income. Industries in the 1990 Census are reported according to the Census industry classification system. Thus, I convert 1990 Census industry codes into NAICS codes through the concordance table provided by the Census Bureau. From 2000 on, instead, industries are coded according to NAICS codes at different levels of precision. 4 To overcome this problem, I convert these NAICS codes into 4-digit NAICS codes, retaining the first 4 digits of 5- and 6-digit NAICS codes, and randomly assigning 1-, 2-, and 3-digit NAICS codes to 4-digit NAICS subcategories. After harmonizing the definition of industries across ACS and 1990 Census data, I restrict the sample to employed workers aged 18 or more with remunerated labor the week before who are not self-employed. Then, I only retain individuals working in a manufacturing industry and drop individuals with missing NAICS. After this process, the overall sample size amounts to roughly 2.3 million of observations. Panel B of Table 1 shows some descriptive statistics comparing this sample to the sample of workers without restricting to the manufacturing sector. On average, people working in the manufacturing sector are more likely to be men, married, less educated and to be employed as operators and laborers. Foreign workers are defined according to the variable Citizen. Following the literature (see for example Ottaviano and Peri, 2012), naturalized citizens and non-citizens are classified as foreign workers, while individuals born abroad of American parents are classified as natives. The variable Migrate1 reports the migration decisions of workers with respect to the previous year, asking whether respondents are staying in the same house as the year before, moved within the state, moved across states or moved from abroad. For 1990, this information is available only with respect to the previous 5 years (variable Migrate5 ). Wages contained in the variable Incwage are adjusted for inflation and transformed into logarithms. I also construct two dummy variables for high and low educated workers. Highly educated workers are defined as workers with some college education or more. Low educated workers are defined as workers with high school diploma or less. At this stage of the discussion, it is interesting to investigate whether foreign and native workers actually differ in their reservation wages. Of course, the information about workers reservation wages is very difficult to find and I proxy reservation wages with actual wages. Particularly, I focus on three categories of workers: foreign born workers, foreign born workers that migrated from abroad, across states or within state the previous year, and native workers that migrated from abroad, across states or within state the previous year. In performing these regressions, I control 4 NAICS codes vary between 1 and 6 digits according to the precision of industry definition. 10

for basic demographic and socio-economic variables such as age, age squared, gender, marital status, education, race, type of occupation, Siegel prestige score. Table 2 shows the results. All the three categories seem to earn lower wages then their native and/or non-migrant counterparts, suggesting that indeed migrant workers may be willing to accept lower wages than non-migrant workers. Finally, all the individual level information is collapsed at industry level according to 4-digit NAICS codes. From the collapsed database, I compute the two main dependent variables: the share of foreign workers out of native workers and the share of foreign workers that recently migrated from abroad, across states or within state out of recently migrated natives. While the first variable accounts for the stock of foreign workers out of native workers, the second variable accounts for the stock of foreign migrants with respect to native migrants. Summary statistics about these two variables are reported in Panel A of Table 1. 4 Empirical strategy To investigate the impact of industry value added on the employment of migrant workers, I first compute simple OLS correlations between industry value added and the share of foreign workers employed. However, reverse causality may bias the OLS estimates, since the concentration of migrant workers in a particular area or industry may also affect the production decisions of firms (see for example Lewis (2011)) and ultimately value added. For this reason, I also adopt an instrumental variable strategy, which relies on a structural change in trade policy. The intuition behind is that grating PNTR to China increased the competitive pressures faced by firms in the manufacturing sector, decreasing industry value added. At the same time, change in trade with China should not have directly affected the employment of foreign born workers. 5 This is equivalent to assume that the impact of the entry of China in the WTO on the employment of foreign workers is entirely mediated by the change in industry value added. Since this assumption may be too strong, in Section 5 I also report reduced form results. 4.1 Baseline specification The baseline OLS specification is: Y it = α + βlog(v A it ) + ΓX it + κ t + ε it (19) where V A is the value added of industry i at time t. Y it refers to the dependent variables, X it is a set of controls, and κ t are time fixed effects. In the baseline regressions, the set of controls includes 5 Migration flows from China to the US are quite small and Chinese migrants only account for roughly 1% of the US population. 11

the logarithm of capital intensity, computed as the real capital stock normalized by the deflated value of shipments, and some fixed effects for industry branch. Controlling for capital intensity is extremely important in order to compare industries with similar productive structures. The share of workers employed in production may also be an indicator of the production structure. However, I prefer not to include this variable in the main estimates, since the size of the workforce may be correlated to foreign to native ratios. Nevertheless, the inclusion of this variable in the estimates does not alter the main results (see Table A.2 in Appendix). 4.2 Identification strategy Since reverse causality may bias the OLS estimates, I derive an instrumental variable approach to identify the impact of industry value added on the employment of migrant workers. This strategy relies on the work by Pierce and Schott (2016), who investigate the impact of the entry of China in the WTO on the employment in the US manufacturing sector. Since 1980, China has been granted the Normal Trade Relationship (NTR) status by the US, which entailed the application of more favorable tariffs. However, this status had to be renewed every year through Presidential extensions. With the entrance in the WTO, China was granted a Permanent Normal Trade Relationship (PNTR) status, ending the uncertainty about the implementation of the NTR tariffs. This allowed US importers to stipulate longer term contracts with Chinese exporters and boosted trade of manufacturing goods between the two countries. As noted by Pierce and Schott (2016), there are four possible mechanisms through which the entry of China in the WTO had an impact on US manufacturing employment: (i) inducing US firms to acquire either inputs or final goods from Chinese producers; (ii) inducing Chinese firms to expand in the US; (iii) inducing US firms to shift production towards more skill-intensive and capital-intensive products; (iv) inducing US firms to shift production or part of the production offshore. Mechanisms (i) and (ii) can be attributed to the increase in the market size and in the competitive pressures encompassed in the theoretical model. Mechanism (iii) and (iv) are not directly discussed in the theoretical model but, if anything, should have diverted migration flows away from the US. In their paper, Pierce and Schott (2016) exploit the difference between non-ntr tariffs and NTR tariffs, and interact this difference with the time of entry of China in the WTO. The intuition behind is that more protected industries, i.e. industries with a larger difference between the potential non- NTR tariffs and NTR tariffs, faced a lager increase in competition from China with respect to less protected industries. As shown by Pierce and Schott (2016), the entry of China in the WTO had a strong detrimental effect on employment in the US manufacturing sector. Also, it coincided 12

with an upsurge in manufacture offshoring and with a shift of US manufactures towards more capital-intensive products. In this paper, I first replicate the Pierce and Schott (2016) exercise to elicit the impact of the entry of China in the WTO on the share of foreign workers employed. Then, I instrument industry value added with this interaction term. Assuming that the entry of China in the WTO did not directly affect migration flows, I obtain non-biased estimates of the impact of industry value added on the employment of foreign workers. 5 Results Reduced form results are shown in Table 3. In addition to the two dependent variables of interest, i.e. the share of foreign to native workers and the share of foreign to native workers among people that migrated the year before, the table shows the results for other two dependent variables, i.e. the logarithm of wages and industry value added. Notice that the impact of the tariff gap interacted with the entry of China in the WTO on industry value added represents the first stage of the IV regressions presented below. Since 4-digit industry fixed effects account for the large part of the variation in the data, all the estimates are reported controlling for either industries defined at 2-digit NAICS level (odd columns) or industries defined at 3-digit NAICS level (even columns). Results are generally consistent across the two different specifications. The shares of foreign to native workers show positive coefficients, suggesting that more protected industries attract more foreign workers than less protected industries. Moreover, these effects are quite large. Relying on the more conservative estimates in Columns (2) and (4), a one standard deviation increase in tariff gap corresponds to an increase of 17% in the mean of the foreign to native ratio and of 12% in the mean of the foreign to native recent migrants ratio. Also, more protected industries seem to pay lower wages and have lower value added. Table 4, instead, shows the impact of industry value added on foreign to native ratios and workers wages. OLS results are reported in the first and the second columns, while IV results are reported in the third and the fourth columns. For both the foreign to native ratios, OLS coefficients are quite small in magnitude and change sign after controlling for 3-digit industry codes. However, as aforementioned, OLS coefficients may be biased by reverse causality. Indeed, if migrant workers tend to cluster in some particular industries providing cheaper labor (think for example to foreign migrant networks), firms in those industries may not have incentives to shift towards more productive technologies or higher value added products. Conversely, IV coefficients are always negative and larger in magnitude, suggesting that firms with larger value added tend to hire less foreign workers. Even though the inclusion of a more 13

demanding set of fixed effects in Column (4) reduces the relevance of the instrument, the IV coefficients in Columns (3) and (4) are qualitatively similar. Particularly, a 1% increase in industry value added corresponds to a 1% increase in the mean of foreign to native ratio and a 0.8% increase in the mean of foreign to native recent migrants ratio. In addition, industry value added is positively correlated with wages, suggesting that firms with larger value added also pay higher wages to their workers. Before concluding, I present some robustness checks. Table A.1 in Appendix reports the main results by education. Indeed, results are consistent for both highly educated and low educated workers, and the size of the effect is quite similar for the two groups. Table A.2 shows the results for the reduced form including the share of workers in production among the controls. Again, coefficients are consistent with the previous results. Finally, adopting a different type of merge between the ACS and 1990 Census and the NBER-CES Manufacturing Industry Database does not alter the main results. In this second version of the merge, I average the values in the NBER- CES Manufacturing Industry Database according to lower digit NAICS codes and I merge them to the industries reported in the ACS without further adjusting ACS codes. Reduced form results are similar to the previous ones and are reported in Table A.3. 6 Conclusion This paper investigates whether the competitive structure of an industry influences the employment of migrant workers. To this end, I develop a theoretical model and an instrumental variable approach. Theoretically, I model this problem combining a monopolistic competitive market for the final good with a labor market with search frictions and workers that are heterogeneous in their reservation wages. Particularly, there is a threshold maximum price for the final good above which the firm is forced to exit the market. For this reason, firms attempt to decrease their marginal costs and to hire lower reservation wage workers. If migrant workers have lower reservation wages with respect to non-migrant workers, more competitive firms hire more migrant workers. The empirical analysis confirms these results. More competitive industries seem to employ more migrant workers and to pay lower wages. These results have strong policy implications. This analysis shows that migratory behaviors are intrinsically related to competitive markets. As long as firm markups are small, firms attempt to decrease their marginal costs hiring cheap labor. In this case, policies aiming at limiting migration setting stricter border rules may not be effective. Indeed, as long as international migrants are aware of the possibility of finding a job in the destination country, stricter border enforcement would only foster illegal immigration. Rather, an effective migration policy should also take into 14

account the characteristics of the labor demand of national firms. For instance, the inflow of low educated workers may be better hampered in the long run through better investments in industries with higher value added. The empirical analysis is limited by the quality of the data used. Particularly, matched employee-employer databases would be more effective in investigating this research question. Moreover, it would be interesting to extend the analysis adding a geographical level and exploiting the geographical concentration of industries across states. Further research will follow these directions. 15

References Acemoglu, D., Autor, D., Dorn, D., Hanson, G. H., Price, B., 2016. Import competition and the great US employment sag of the 2000s. Journal of Labor Economics 34 (S1), S141 S198. Antras, P., Helpman, E., 2004. Global sourcing. Journal of Political Economy 112 (3), 552 580. Autor, D. H., Dorn, D., Hanson, G. H., Song, J., 2014. Trade adjustment: Worker-level evidence. The Quarterly Journal of Economics 129 (4), 1799 1860. Becker, R., Gray, W., Marvakov, J., 2013. NBER-CES Manufacturing Industry Database: Technical Notes. NBER Working Paper 5809. Borjas, G. J., 2003. The labor demand curve is downward sloping: Reexamining the impact of immigration on the labor market. Quarterly Journal of Economics 118 (4), 1335 1374. Borjas, G. J., Katz, L. F., 2007. The evolution of the Mexican-born workforce in the United States. In: Mexican immigration to the United States. University of Chicago Press, pp. 13 56. Card, D., 2001. Immigrant inflow, native outflow, and the local labor market impact of higher immigration. Journal of Labor Economics 19 (1), 22 64. Dustmann, C., Schönberg, U., Stuhler, J., 2017. Labor supply shocks, native wages, and the adjustment of local employment. The Quarterly Journal of Economics 132 (1), 435 483. Ebenstein, A., McMillan, M., Zhao, Y., Zhang, C., 2011. Understanding the Role of China in the Decline of US Manufacturing. Manuscript, Hebrew University of Jerusalem. Eckstein, Z., Van den Berg, G. J., 2007. Empirical labor search: A survey. Journal of Econometrics 136 (2), 531 564. Falk, A., Fehr, E., Zehnder, C., 2006. Fairness perceptions and reservation wages the behavioral effects of minimum wage laws. The Quarterly Journal of Economics 121 (4), 1347 1381. Feenstra, R. C., Romalis, J., Schott, P. K., 2002. Us imports, exports, and tariff data, 1989-2001. NBER Working Paper 9387. Grossman, G. M., Helpman, E., 2002. Integration versus outsourcing in industry equilibrium. The Quarterly Journal of Economics 117 (1), 85 120. Grossman, G. M., Rossi-Hansberg, E., 2012. Task trade between similar countries. Econometrica 80 (2), 593 629. 16

Lewis, E., 2011. Immigration, skill mix, and capital skill complementarity. The Quarterly Journal of Economics 126 (2), 1029 1069. Melitz, M. J., Ottaviano, G. I., 2008. Market size, trade, and productivity. The review of economic studies 75 (1), 295 316. Ottaviano, G. I., Peri, G., 2012. Rethinking the effect of immigration on wages. Journal of the European Economic Association 10 (1), 152 197. Ottaviano, G. I., Peri, G., Wright, G. C., 2013. Immigration, offshoring, and american jobs. The American Economic Review 103 (5), 1925 1959. Pierce, J. R., Schott, P. K., 2012. A concordance between ten-digit us harmonized system codes and sic/naics product classes and industries. Journal of Economic and Social Measurement 37 (1, 2), 61 96. Pierce, J. R., Schott, P. K., 2016. The surprisingly swift decline of us manufacturing employment. The American Economic Review 106 (7), 1632 1662. Piore, M. J., 1979. Birds of passage: migrant labor and industrial societies. Cambridge England Cambridge Univ. Press 1979. 17

Figure 1: Trends in industry level characteristics by tariff gaps Notes - All the variables are converted into logarithms and plotted. The solid line depicts industries facing a tariff gap above the median, while the dashed line depicts industries facing a tariff gap below the median. The share of workers in production is the ratio between the number of workers in production and employment. Capital intensity is the share between the real capital stock and the deflated value of shipments. In the NBER-CES Manufacturing Industry Database, value added is defined as the difference between the value of shipments and the cost of materials, supplies, containers, fuel, purchased electricity and paid work and is adjusted for inventories and value added from merchandising operations. I further deflate this value according to the Consumer Price Index published by the Federal Reserve. Sources: NBER-CES Manufacturing Industry Database - years 1990-2011. 18

Table 1: Descriptive statistics PANEL A: INDUSTRY LEVEL CHARACTERISTICS Mean Standard Deviation Value added (in log) 8.22 0.98 Average payroll (in log) 6.97 0.85 Capital intensity (in log) -0.79 0.46 Share of workers in production (in log) -0.35 0.19 Foreign/Native Ratio 0.22 0.15 Foreign/Native Ratio - Recent Migrants 0.27 0.18 PANEL B: INDIVIDUAL CHARACTERISTICS Manufacturing sector Whole sample % Men 0.68 0.51 % White 0.82 0.81 % Black 0.08 0.09 % Married 0.67 0.61 % College education 0.44 0.57 Age 41.79 41.09 % Foreign born 0.13 0.12 % Recent foreign migrants 0.04 0.03 % Recent native migrants 0.22 0.23 Siegel occupational prestige score 38.24 41.07 Managerial occupations 0.21 0.30 Technical, sale and administrative support 0.19 0.31 Service occupations 0.02 0.13 Precision production, craft and repair 0.18 0.10 Operators, fabricators, and laborers 0.40 0.13 Observations 2,293,754 15,449,656 Notes - PANEL A: Industries are defined according to 4-digit NAICS codes. This sample only includes NAICS codes with at least one corresponding individual in the ACS/1990 Census sample. Average payrolls and value added are deflated according to the Consumer Price Index published by the Federal Reserve. In the NBER-CES Manufacturing Industry Database, value added is defined as the difference between the value of shipments and the cost of materials, supplies, containers, fuel, purchased electricity and paid work and is adjusted for inventories and value added from merchandising operations. Capital intensity is the share between the real capital stock and the deflated value of shipments. The share of workers in production is the ratio between the number of workers in production and employment. PANEL B: The whole sample contains employed workers older than 18, with remunerated work the week before, not self-employed and with non-missing industry of work (NAICS). The manufacturing sector sample only contains the individuals of the whole sample working in the manufacturing sector (NAICS beginning with 3). Sources: NBER-CES Manufacturing Industry Database; ACS and 1990 Census - years 1990, 2000-2011. 19

Table 2: Wages of migrant workers Foreign Recent foreign Recent native born migrants migrants Column (1) (2) (3) Wages (log) -0.101*** -0.132*** -0.012*** (0.00) (0.00) (0.00) Observations 2,274,176 2,274,176 2,274,176 R-squared 0.39 0.39 0.39 Demographic controls Yes Yes Yes Socio-economic controls Yes Yes Yes Year FE Yes Yes Yes Notes - The dependent variable of interest is the logarithm of annual wages adjusted for CPI. Recent foreign migrants are foreign born workers that migrated from abroad, across states or within state the year before. Recent native migrants are defined in the same way. Demographic controls include age, age squared, gender, marital status, and race. Socio-economic controls include education, type of occupation, and Siegel prestige score of the occupation. Significance levels: *** p < 0.01, ** p < 0.05, * p < 0.1. Standard errors (in parenthesis) are robust to heteroskedasticity. Sources: ACS and 1990 Census - years 1990, 2000-2011. 20

Table 3: Impact of the entry of China in the WTO on foreign/native shares, wages and value added - reduced form Foreign/Native Foreign/Native Ratio Wages Value added Ratio Recent migrants (in log) (in log) Column (1) (2) (3) (4) (5) (6) (7) (8) Post*Tariff gap 0.385 0.231 0.351 0.196-1.185-0.390-1.650-0.588 (0.06) (0.04) (0.06) (0.06) (0.27) (0.39) (0.31) (0.37) Observations 677 677 677 677 677 677 677 677 R-squared 0.43 0.73 0.32 0.53 0.23 0.58 0.18 0.69 Mean Post*Tariff gap 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 Std. Dev. Post*Tariff gap 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 Capital intensity (log) Yes Yes Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes 2-digit NAICS FE Yes No Yes No Yes No Yes No 3-digit NAICS FE No Yes No Yes No Yes No Yes Notes - The foreign/native ratio is the share of foreign workers with respect to native workers. The foreign/native ratio for recent migrants is share of foreign workers that migrated from abroad, across states or within state the year before out of native workers that migrated from abroad, across states or within state the year before. Wages comes from the NBER-CES Manufacturing Industry Database and are adjusted for CPI. In the NBER-CES Manufacturing Industry Database, value added is defined as the difference between the value of shipments and the cost of materials, supplies, containers, fuel, purchased electricity and paid work and is adjusted for inventories and value added from merchandising operations. I further deflate it for CPI. Capital intensity is computed as the share between the real capital stock and the deflated value of shipments. The Post*Tariff gap variable is the interaction between the non-ntr/ntr tariff gap and the entry of China in the WTO. Significance levels: *** p < 0.01, ** p < 0.05, * p < 0.1. Standard errors (in parenthesis) are robust to heteroskedasticity. 21

Table 4: Impact of industry value added on the nationality of workers and wages OLS Column (1) (2) (3) (4) Foreign/Native Ratio Value added (log) -0.034 0.025-0.233-0.393 (0.01) (0.01) (0.05) (0.26) Observations 677 677 677 677 Kleibergen-Paap F 29 2 Mean Dependent Var. 0.22 0.22 0.22 0.22 IV Foreign/Native Ratio - Recent Migrants Value added (log) -0.027 0.025-0.213-0.333 (0.01) (0.01) (0.05) (0.24) Observations 677 677 677 677 Kleibergen-Paap F 29 2 Mean Dependent Var. 0.27 0.27 0.27 0.27 Workers wages (in log) Value added (log) 0.731 0.929 0.718 0.664 (0.02) (0.02) (0.08) (0.28) Observations 677 677 677 677 Kleibergen-Paap F 29 2 Mean Dependent Var. 6.97 6.97 6.97 6.97 Capital intensity (log) Yes Yes Yes Yes Year FE Yes Yes Yes Yes 2-digit NAICS FE Yes No Yes No 3-digit NAICS FE No Yes No Yes Notes - The foreign/native ratio is the share of foreign workers with respect to native workers. The foreign/native ratio for recent migrants is share of foreign workers that migrated from abroad, across states or within state the year before out of native workers that migrated either from abroad, across states or within state the year before. Wages comes from the NBER-CES Manufacturing Industry Database and are adjusted for CPI. In the NBER-CES Manufacturing Industry Database, value added is defined as the difference between the value of shipments and the cost of materials, supplies, containers, fuel, purchased electricity and paid work and is adjusted for inventories and value added from merchandising operations. I further deflate it for CPI. Capital intensity is computed as the share between the real capital stock and the deflated value of shipments. The IV is the interaction between the non-ntr/ntr tariff gap and the entry of China in the WTO. Significance levels: *** p < 0.01, ** p < 0.05, * p < 0.1. Standard errors (in parenthesis) are robust to heteroskedasticity. 22