Gender Inequality in U.S. Manufacturing : Evidence from the Import Competition

Size: px
Start display at page:

Download "Gender Inequality in U.S. Manufacturing : Evidence from the Import Competition"

Transcription

1 Gender Inequality in U.S. Manufacturing : Evidence from the Import Competition Chan Yu February 9, 2019 Most Recent Draft Here Abstract In this paper, I analyze the effect of import competition from China on gender inequality in the US manufacturing sector. China s import competition between provides a unique opportunity to explore the effect of gender inequality by a trade liberalization that hits harder on female-intensive sectors. I conduct a detailed analysis of how Chinese import competition affect employment and wage outcomes differently between men and women. I find that female workers experienced a more substantial decline in the manufacturing employment and wage in relative to their male counterparts. This gender difference by the impacts of Chinese import competition is not fully explained by gender difference in exposure to trade shock across industries. By comparing male and female manufacturing employment across education and age groups, I find the gender inequality is felt most strongly by low skilled and old women. Meanwhile, there is an increase of nonmanufacturing employment for low-skilled women which offsets their job losses in the manufacturing sector. 1 Keywords: Trade,Gender,Inequality, Employment, Wages, Manufacturing JEL Codes: F16, J23, J31, J71, L60, R12 1 Chan Yu: Graduate Student (Fourth Year), Department of Economics, Department of Economics- The University of Texas at Austin, 2225 Speedway, BRB 1.116, C3100, Austin, Texas ( chanyu@utexas.edu); Phone:

2 1 Introduction There is a growing number of studies about gender difference in the labor market adjustment with respect to trade liberalization. However, no uniform pattern is found in terms of the relationship between trade liberalization and gender inequality. On the one hand, Beckers taste-based model implies that increasing product market competition induced by trade liberalization will reduce the gender wage gap. This is because gender discrimination against women becomes weaker in the labor market 2. Empirical evidence in developed countries also supports this theory by showing a narrowed gender employment and wage gap 3 in sectors exposed with higher import competition (Black and Brainerd (2004); Benguria and Ederington (2017)). On the other hand, trade liberalization may increase gender inequality in labor outcomes if males and females are distributed unevenly across trade sectors. Because if trade liberalization asymmetrically affects female-intensive sectors, the capital reallocation from female-intensive sectors to male-intensive sector reduces more job opportunities for women who rely on capital (Sauré and Zoabi, 2009). Cross-country analysis has shown that trade liberalization reduces gender inequality only when the country has more comparative advantage in the female-intensive sectors (Do et al., 2014). More recent literature in gender pay gap also emphasizes that the industry difference between men and women may explain a large portion of the gender wage gap in post 1990 period (Blau and Kahn, 2017). To better understand the mechanism that causes the gender inequality in labor market outcomes, it is important to incorporate industry difference across genders. 2 Becker s theory implies that prejudiced employer will not hire female workers unless female workers accept a lower wage relative to male workers to compensate for disutility associated with employing women. Non-prejudiced employers can earn higher profits than the prejudiced employer by hiring cheaper female workers. When competition in the market intensifies, the prejudiced employers will lose their business and be driven out of the market. Therefore, there is a decrease in the gender pay gap when more prejudiced employers exit the market. 3 All gender gap in this paper refers to the employment or wage difference between men and women. 1

3 China s rise in manufacturing since 1990 provides a unique opportunity to study how declining female intensive sectors affects gender difference in labor outcomes. First, China has the highest growth rate in low skilled labor-intensive sectors such as apparel, miscellaneous and leather that many female workers dominate. Figure A.1 displays the relationship between import penetration and the ratio of female to male worker across manufacturing industries in pre (1990) and post period (2007) 4. Among sectors (apparel, miscellaneous) with the highest import penetration rates 5, the total number of women is three times as large as it for men. Unsurprisingly, the decline in the gender ratio is also larger in these sectors after trade shock occurred. Second, the size of Chinese trade shock is much larger than any other trade shock (Goldberg and Pavcnik (2016), Autor, Dorn, and Hanson (2013)). Between 1990 and 2007, imports from China in manufacturing industry increased from 2% to 12% and China became the largest trading partner with the US, which accounted for about 16% the total trade flow in the US. Unlike other trade policies which correlate with domestic economic activities, China s trade growth is mainly driven by its internal (domestic policy and institution changes) and external factors which are unanticipated. Following, I construct the import exposure measure at the local commuting zone level by exploiting two sources of variation: variation in initial industry composition across regions and variation in the US imports from China across industries. One concern of using the observed import growth in the US is that it might be driven by any unobserved productivity shock from the United States. Suppose the decreasing productivity leading to import growth reduces the employment and wage growth in the manufacturing, then using the observed import growth in the OLS model will underestimate the effect of Chinese import growth on men and women s labor outcomes. Therefore, I use the import growth in eight other devel- 4 The number of industries in Figure A.1 uses broadly defined Census industry categorization. However, when I construct the import measure, I use SIC four-digit industry code which leads to 396 manufacturing industries. 5 The import penetration rate is calculated by dividing US imports from China with total US shipment, see Autor, Dorn, and Hanson (2013). 2

4 oped countries 6 excluding United States to instrument for the observed imports change in the United States. While previous studies suggest that women in the manufacturing sector should benefit from trade liberalization by facing weakened discrimination in the labor market, I find a larger decrease in their manufacturing employment and wage rate by Chinese import competition. Autor, Dorn, and Hanson (2017) examine the overall employment by young workers and find men have worse overall labor outcomes than women do. Their study considers that overall men have a higher concentration rate in the manufacturing sector than women do. Rather than being a substitute, my work extends their work by shedding new lights on the gender difference within manufacturing. Overall men received much higher import exposure than women did. However, within the manufacturing sector, women on average received $ 4000 more import exposure than men because women work in sectors that are hit harder by China s rise. After controlling for the source of variation in the gender-specific import exposure due to initial industrial employment distribution, I find that gender-specific employment becomes even stronger which suggests that the initial industry distribution difference by gender is not the primary explanation for the differential labor adjustments across genders. In other words, female workers are more responsive to negative demand shock than male workers at least in the manufacturing sector. The average effect of import competition on gender-specific labor outcomes can be masked by the workforce composition, particularly the education and age difference between men and women. I disaggregate the labor force by age and education, further examining the heterogeneity effect across demographic groups in employment and wage. If women who select to work in manufacturing are mainly low-skilled workers without a college degree, then the average effect in gender inequality should be small. Previous aging literature points out that 6 The eight developed countries are those that have comparable trade data between : Australia, Denmark, Finland, Germany, Japan, New Zealand, Spain, and Switzerland. 3

5 labor demand may be steeper for old men and old women since their productivity is lower than younger workers (Carlsson and Eriksson, 2017). I find the most significant effect is among workers with less than high school education. Some basic facts can support the heterogeneous result: high school dropout workers in manufacturing mainly work in occupations such as machine operators or assemblers which uses low skilled routine and manual tasks (Autor, 2010). These occupations are more likely to be cut down when the manufacturing sector declines. To characterize which age group of workers experience the greatest declines in employment and wage, I further make an analysis of employment and wage change within gender, age and education groups. I find that demographic group suffering the most decline in manufacturing employment is old women (40-64) with middle or low skill (high school graduate or below ). I test the gender difference in employment and wage effects of young workers and old workers, finding that young men do not show a statistical difference to young women in manufacturing employment and wage. However, the gap becomes significant among old men and old women who are beyond 40. I conduct a falsification exercise on old workers in pre-period between , but did not see any pattern in the gender gap in manufacturing employment or wage in the pre-period. The insignificant results suggest that the rising gender inequality by low skilled and old aged women results from China s rise rather than other underlying trends in the manufacturing sector or local labor market condition. Given the robustness of this result, I make additional analyses to better understand why low-skilled old women in manufacturing are more responsive to import competition. Together with the strong adverse manufacturing employment effect by Chinese import competition, I find a positive effect of nonmanufacturing employment by old aged high school graduate women that previous literature did not find among low skilled workers 7. This increasing 7 Autor, Dorn, and Hanson (2013) use Census data and finds only negative effect in the nonmanufacturing sector among noncollege workers uses longtitudinal data to analyze the effect of trade exposure on 4

6 nonmanufacturing employment by women results in a decline in the manufacturing concentration rate by women. All these evidence tells a striking story that old aged female workers may migrate to the nonmanufacturing sector to recover their job losses in the manufacturing sector. In other words, the negative effect in manufacturing is offset by the positive effect in nonmanufacturing for old aged women. It turns out the overall decline in employment is almost the same for men and women. The nonmanufacturing sector which has the most increase of these women workers are professional and related service such as libraries, nursing hospitals. However, these women who migrating to the nonmanufacturing sector still works at least paid occupations including clerical and retail assistants. For men, I only find a tiny positive effect of their nonmanufacturing sector in transportation. One possibility for this gender difference in job mobility across sectors is that old men can find new job opportunities in the manufacturing sector and they retain their jobs within the same sector. Growing service industry in local labor markets attract old women who have fewer job opportunities within the manufacturing sector. These results point out that the manufacturnig decline may lead to women migrating to service sector where is traditionally perceived as female s sector and provide more job opportunities for women. 2 Contribution and Previous Literature My paper contributes to the growing literature of trade liberalization and gender inequality by finding evidence on how men and women mitigate their costs of trade adjustment differently. Previous studies argue that gender inequality decreases by the globalization from Becker s discrimination model (Black and Brainerd, 2004). By examining the change of manufacturing employment by men and women as well as the residual wage gap, I find little change on the unexplained wage gap which rules out the possibility of discrimination channel. worker s adjustment. They find low-wage workers churn primarily among the manufacturing sector and only high-wage workers are more likely to move out of manufacturing. 5

7 Instead, I find low skilled female workers are borne with higher costs of trade adjustment than low skilled male workers, and this cost is felt the most strongly by old women. In the long run, women can switch their jobs to the nonmanufacturing sector to offset their losses in the manufacturing sector. By focusing on the trade liberalization, this paper also extends a vast existing literature on identifying the differences across demographic groups during recessions. My study uses a different shock from trade liberalization than business cycles. The primary explanation for the demographic differences in the impacts of the cycle in previous literature is that workers are sorted into industries with different cyclicality. For instance, low skilled men are more likely to be employed in the manufacturing sector which also has greater exposure to recession cycles. Hoynes, Miller, and Schaller (2012) find that the impacts of the Great Recession is felt most strongly by men with low education. In their work, differential industry-occupation exposure to cycles explains a substantial portion of impacts across gender groups. I highlight a different channel by which Chinese import competition generates more negative outcomes for female workers in the manufacturing sector. By constructing a gender-specific import exposure to control for variation source across genders, I test the role of the industry exposure to China s rise in causing the gender difference in employment and wages. My results suggest that industry distribution difference across genders plays a modest role in affecting gender inequality. This paper conjectures that men and women may have different sensitivity to the declining local labor demand change. There is scant empirical evidence on the effect of globalization on job switch across sectors by low skilled workers. Autor, Dorn, Hanson, and Song (2014) use worker history administrative data and find that low-skilled workers face higher barriers of moving to the nonmanufacturing sector. Their findings show that low-skilled workers retained in the manufacturing sector and received repeatedly import exposure. By breaking into gender groups, 6

8 I find that both high skilled young men and women can partially offset their employment losses in the manufacturing sector by switching to the nonmanufacturing sector. Among low skilled workers, the decline in manufacturing employment is not accompanied with an increase in nonmanufacturing employment. In addition, I also explore how local labor demand affects older workers, who are more responsive and feel more difficult to recover their job losses (Maestas et al., 2013). It links to previous studies about Chinese trade shock and local labor adjustments by Autor, Dorn, and Hanson (2013) and Autor, Dorn, and Hanson (2017). who assessed the marriage market of young adult workers (age between 16-39). Instead of looking at the overall labor outcomes for young adults, I extend their work by highlighting the labor adjustment difference across age and skill groups. I find the increasing gender inequality in the manufacturing sector mainly comes from the old worker group. I find a significant increase in nonmanufacturing employment for middle-skilled young men and old women recovers around forty percent of the manufacturing employment decline. Finally, it also provides empirical evidence to the growing literature on studying recent falling convergence rate of gender wage gap. Recent studies in gender wage gap hypothesize that industry and occupation difference between men and women may explain a larger portion of the wage gap since the 1990s than previous decade (Blau and Kahn, 2017). However, there is little empirical evidence of identifying how industry or occupation change affects the post-1990s gender wage gap. The challenge in this literature is how to find an exogenous industry-specific shock. To address this challenge, I use the import shock from China which generates differential import exposure within the manufacturing sector. 7

9 3 China s Trade Shock The last decade witnessed a tremendous growth in commercial trade between U.S. and China. According to UN Comtrade, the total trade between the two countries increased from $5 billion in 1980 to $634 billion in During the 1990s, China gradually dismantled its inefficient centralized trade planning and adopted a series of trade policies. These reforms in China made traded goods easier to sell to the international market for Chinese producers of export goods. Some statistics document that exporting products subject to licensing and quotas was reduced from around 67% in 1991 to only 8 percent in Moreover, China s entry into the World Trade Organization (WTO) in 2001 liberalized China s trade regime and granted more equal trading rights to Chinese manufacturers. All these policy changes promoted China productivity and accelerated its export growth to developed countries. In 1992, China initiated a new reform named Socialist Market Economy which terminated its central planning economy era. The new reform transformed the Chinese economy into a market-oriented economy and greatly improved the efficiency of domestic production by State Owned Enterprises. On the other hand, Chinas successfully entry to WTO in 2001 is not anticipated by the public which effectively prevents firms and workers in the US to adjust their behaviors before the trade growth in China. The Tiananmen Square event in 1989 cast doubt on Chinas market transition and many WTO members were skeptical about Chinas future since then (Autor et al., 2013). Additionally, for China, the accession of WTO is a very complicated process involved with a lot of WTO countries which also adds uncertainty. China not only needs to negotiate the bilateral concession agreements with each WTO member, but also the protocol of accession with the whole WTO. 8

10 4 Data 4.1 Data Sources The wage and employment data for men and women come from the Census Integrated Public Use Micro Samples (IPUMS) data for the Census year 1970, 1980, 1990, 2000 and Dataset for 2007 year is constructed using a three-year ACS sample which covers The geographically continuous unit in Census IPUMS dataset is at county group or Public Used Microdata Area (PUMA) level. To convert Census geographic units into commuting zones, I use a projection method developed by Dorn (2009), which links each PUMA or county group j to every commuting zone k by computing the probability of a resident from PUMA or county group j lives in commuting zone k. Using the crosswalk between PUMA or county group with commuting zone provided by, I replace each observation in Census data with 722 observations that are identical to initial observations. The person weight is then equal to the initial Census original person weight multiplied with the David Dorns imputed probabilities. I restrict the sample to working-age population who were working in the year preceding survey. I drop individuals belonging to institutional group quarters such as prisons and psychiatric institutions. All calculations are weighted by the sampling weight (Autor et al., 2013). Trade data comes from the UN Comrade Database. The bilateral trade data is recorded at six-digit Harmonized System product level. I use the crosswalk in Pierce and Schott (2016) to map the HS products to four-digit SIC industries. There are some SIC industries that are not assigned with any HS products, I aggregate these 4-digit SIC industries so that each of the 397 manufacturing industries matches to at least one trade code (Autor et al., 2013). Though the public dataset from Census uses detailed occupational classifications (occ1990), 9

11 one limitation of using Census definition is that occupational structure changes over time. For instance, the economics instructors in occ1990 category is no longer reported after the 2000 Census. This unbalanced structure is problematic in studying the change in the distribution of employment across occupation as occupations are inconsistently categorized. To correct this bias, I use the occupation crosswalk created by Dorn (2009) and convert Census occupation code (occ1990) into occ1990dd so that all occupations exist for all years and are consistent categorized over time. As a result, I have 332 occupations in my sample. 4.2 Measure of Import Exposure To construct the main import exposure measure (gender combined), I first construct the industry specialization at local labor market level which is measured by the share of national industry employment in that CZ. I use the County Business Patterns (CBP) data in 1980, 1990 and The CBP data series provide detailed information on employment, firm size and payroll for each establishment by county and industry level (NAICS). I aggregate the county level total employment of each industry to the commuting zone Autor, Dorn, and Hanson (2013). The main measure of local labor markets exposure to import growth is the observed change in US import from China in industry j between two periods and then weighted by the initial periods specialization in import-intensive industries in a local labor market: IP W it = j L ijt Import us jt (1) L it L jt where j indexes for tradable industry j. i is the local labor market index, L ijt is the region is employment in industry j and L jt is the U.S. employment in industry j for period t. Thereby the local import exposure is calculated by the value change of US imports in industry j, 10

12 Import us jt scaled by the local labor force L it. The measure of import exposure per worker at region i then is obtained through interacting the industry specialization at initial period L ijt L jt of a decade with the import growth across industries. In later section, when I analyze gender separate import exposure measure, I use Census data set to construct the import exposure measure for men and women separately. I did not use County Business Pattern Data because it does not break the total employment into demographic groups. To do so, I construct male and female specific import exposure measure using Census IPUMS data to measure manufacturing industrial employment share by gender as below: IP W M it IP W F it = j = j L M ijt L M it,mfg L F ijt L F it,mfg Import us jt L jt (2) Import us jt L jt (3) L M ijt/l M it,mfg and LF ijt/l F it,mfg are share of men or women working in industry j within manufacturing at initial time t of a decade. Since my data source is different than the baseline results which use CBP data, I examine the potential bias resulting from using an alternative data source by comparing the results and the correlation of the IP W it from two datasets (Appendix figure). Table 1 shows the summary statistics for main variables from As one can see in Table 1, for the main import exposure in equation (1), the mean import exposure is around 3810 $ per worker between However, the gender separate import exposure suggests different import exposure magnitude across genders. Male worker overall (all sectors) receives higher import exposure than female worker as more male workers work in the manufacturing sector. However, if we just look into the average import exposure per worker within the manufacturing sector (row (4)-(5)), women are exposed with much higher 11

13 import exposure because they concentrate in the most trade affected sectors. To illustrate the importance of within manufacturing in gender-specific exposure, I draw the distribution graph in import exposure using within manufacturing and within all sectors and show them in Figure 1 and Figure 2. Figure 1 visually compares the size of the gender-specific import exposure measure within all sectors and within the manufacturing sector. The top graph shows the main import exposure measure from equation (2)-(3). It considers that not only men and women have different distribution of employment across manufacturing sectors, but also that men and women have different manufacturing employment concentration. The bottom figure uses only the employment within the manufacturing sector. Obviously, distribution of import exposure within manufacturing for female workers shifts more outward, suggesting a higher import exposure borne by women in the manufacturing sector. 5 Empirical Methodology 5.1 Identification There is abundant literature on studying the impact of trade policy on local labor outcomes (Goldberg and Pavcnik (2016), Autor, Dorn, and Hanson (2013)). One issue for the identification strategy is that trade policy is usually endogenously determined. For instance, the export growth in Mexico and Central America is driven by the product demand change of their trading partner - US which correlates with the labor demand change (). Studies using tariff changes also face a similar issue that the tariff imposed on specific industry is usually driven by the market condition change in a country (Dai et al., 2018). Other non-tariff barriers (quotas, anti-dumping duties) may involve with strong pretrend effect if those barriers exist even prior to the imposition of barriers. China s trade growth has the advantage to avoid these identification issues. The dra- 12

14 matic export growth in China in the1990s and 2000s is driven by the structural change of the Chinese economy itself, rather than any industry demand or productivity change from highly developed countries. Before 1978, Chinas domestic production was not adjusted by the market demand but under the control of its government, which generated a lot of inefficiency and distortion. However, a series of new reforms led by the new chairman-deng Xiao Ping, aiming to develop socialism with Chinese characteristics, transformed Chinese economy from highly centralized economy to market-oriented type and promoted the growth of Chinas productivity since then. Following Autor, Dorn, and Hanson (2013), the identification strategy in this paper relies on two sources of variation: first, import growth varies across industries within the manufacturing sector; second, the local industry mix varies across regions (commuting zones). This approach allows for geographic variation in the local labor market condition in the sectoral reallocation. US labor market has become less mobile and fluid since the 1980s (Molloy et al., 2011). If Chinese import competition hits harder on some local labor markets than the others, this variation in the impacts across labor markets may not be offset by the flowing of workers across markets as labor mobility is not complete (Dix-Carneiro and Kovak (2017); Müller, Stegmaier, and Yi (2015)). Thus, this labor mobility cost may result in different impacts of Chinese import competition across regions in the equilibrium (Topel, 1986). I identify the effect of import competition on the gender specific wage and employment at local labor market by exploiting initial differences of industry specialization in local labor markets as well as different import growth across industries from The baseline model is as below: L M it = β M IP W us it + X M it + γ t + e it (4) L F it = β F IP W us it + X F it + γ t + e it (5) 13

15 where L M it and L F it are the decadal change in the male and female manufacturing employment to population ratio or log wage rate in commuting zone i. IP W us captures the change in the US imports from China exposed to the local labor market i defined in the previous section. γ t is the time fixed effect to control for aggregate trend. X it controls for observable variables that correlate with the labor outcomes by men and women. I also add nine census division dummies to control for dynamic regional economic growth. In the X it vector, I control for the share of men and women with a college education, the share of population that is foreign-born, the share of female employment, two variables defined by Autor, Dorn, and Hanson (2013) to absorb manufacturing trends in routine tasks and offshoring activities. All these variables are measured at the start-period of each decade to reduce their contemporaneous correlations with error terms. When estimating the model for the interval between , I stack the ten-year equivalent first differences for two periods, and , and include the time dummy for each decade. The stacked model is just a standard linear model in the differences between dependent and independent variables. The estimate from the stacked model should be consistent with the one from three periods fixed effect model, but with fewer restrictive assumption. To produce a consistent result, the fixed effect model requires the idiosyncratic error is serially uncorrelated, however, the stacked first difference model only requires the idiosyncratic error follows a random walk (Wooldridge, 2015). In many cases, the exogeneity assumption for the stacked first difference model is less restrictive compared to the fixed effect model in the sense that the former only needs the idiosyncratic error is not contemporaneously correlated with the past import growth or the future import growth, while the latter requires no serial correlation between Chinese import growth and idiosyncratic error in any time period which is usually not the case. To verify this assumption for the stacked first difference holds or not, I will test the relationship between past gender wage/employment gap and future import growth. If past gender wage/employment gap is weakly correlated 14

16 with the future import growth or not, it will be more convincing in using the stacked first difference model here. 5.2 Instrument Variable One threat to the simple OLS model is that unobservable productivity shock increasing the US import growth may decrease the employment and wage by men or women. The observed import change can be decomposed into two parts: the internal growth in China or falling trade costs from the supply side and the unobserved domestic product demand or supply shock (or productivity shock) from the US. For instance, if the domestic supply of a final good is reduced for some reason, then the import demand for that good will go up. This domestic product demand shock may threat my identification of import competition if the unobserved domestic product demand shock shifts the labor demand for male and female workers in the manufacturing sector.then estimates by OLS model do not reflect how import competition affects male and female labor demand across occupations or industries but are biased by the product demand change in the US. For instance, if there is a negative productivity shock in female-intensive industries, then women will be more negatively affected than men. To isolate the supply driven components in import growth from the unobserved U.S. industry shock that might shift the import demand, I use Chinese import growth to other highly developed countries, Import jt, to instrument for US import growth. The industrial specialization is calculated using the share of employment in specific industry at the initial time in the prior decade. The instrument variable and first stage is as below: IP W oth it = j L ijt 1 Import oth jt (6) L it 1 L jt 1 15

17 IP Wit us = β IP Wit oth + X it + γ t + ɛ it (7) where γ t is the time fixed effect to control for aggregate trend and I put the initial share of employment in manufacturing to control for any aggregate growth in the manufacturing sector. The identification assumption is that the common component of import growth between the US and other highly developed countries only comes from the rising comparative advantage or falling trade costs in China. This assumption is plausible in the sense that the main component in import change is usually unique in the countrys own setting (policy reform or institution), which is not related to other countrys change. However, for any reason, if the common component in import growth in developed countries (including US) is a function of any industry shock that may shift the import demand, then my IV is also biased in the same direction as OLS model. To test whether this identification assumption is valid or not, I use a gravity-based approach in the section following Autor, Dorn, and Hanson (2013) to provide an alternative measure to capture the change of import exposure at local labor market level in later section 8.3. Another assumption for the stacked model is zero contemporaneous correlation. I examine it by looking at whether the future change in import exposure predicts the gender pay gap or not (import demand shock, productivity shock). This model does not consider a role of the US exports to China. Basic trade theory tells us that the productivity growth or falling trade costs in China may increase the US export supply to China and positively affect the labor demand in the US. However, the size of imports from China greatly exceeds exports to China so that the impact from export side should not be as significant as the import change 8. Moreover, the major exports from the 8 Based on some facts, the trade balance for goods and service in US has shown a deficit since the 1980s. 16

18 US to China are usually services which are difficult to be measured precisely. Unlike goods which can be compiled on a customs basis, services are compiled on a balance of payments basis. For simplicity and accuracy, I ignore the export channel at this moment, but I will consider the channel from export side in US in the robustness exercise by using net import growth rather than gross import growth in later section. 6 Results 6.1 Effects on Gender-Specific Labor Outcomes Before moving to the estimated results, I explore the graphical analysis of the impact of Chinese imports growth on raw changes in gender-specific labor outcomes. I divide the 722 commuting zones into ten decile groups and rank them based on the magnitude of average import exposure. As one can see in Figure 3, low-skilled women experienced much larger declines in manufacturing employment than low-skilled men. The wage effect is relatively modest which is consistent with previous standard literature of wage rigidity in the developed countries. Though less significant in magnitude, it is still noticeable in Figure 4 that the hourly wage for women grew less in regions (commuting zone) with higher import exposure over and this pattern is not found on male manufacturing wage. Estimates in Table 2 are consistent with previous graph analyses that the adverse effect by import competition is more pronounced among low-skilled women 9. I include census division dummies, time fixed effects and full controls described in Section 5 in the main specification. Results from OLS and 2SLS model are presented in Table 2, showing that low skilled women experienced a more significant and greater decline in manufacturing employment to 9 In this paper, I define high school dropouts are low skilled workers. Middle-skilled are workers with high school graduate degree. 17

19 Chinese imports competition than their male counterparts. The first four columns report the estimated employment effect predicted by OLS model, and the last four columns show the ones via 2SLS. By comparing OLS and 2SLS estimates, I find manufacturing employment by low skilled women is significant in 2SLS model while insignificant in the OLS model. The difference in two models suggest a negative demand shock in female-intensive sectors which may decrease the labor demand of women and downard bias the OLS estimates on female labor outcomes. With 1000$ increase in import competition, there is a percentage point decrease in manufacturing employment by high school dropout women. In contrast, import competition has little effect on high school dropout men. By allowing the gender heterogeneity in response to China import competition, I found a positive effect on the nonmanufacturing sector which is not shown in previous studies (Autor (2018), Autor, Dorn, and Hanson (2013)). Consistent with who find a negative effect of nonmanufacturing employment among noncollege workers, Column (3)-(4) suggest Chinese import competition decreases nonmanufacturing employment for both low skilled men and women in my results. There are around and percentage points decrease in nonmanufacturing employment for low skilled men and women, with 1000$ increase in the import exposure per worker. This indirect effect can be either through changing the labor supply by workers in the manufacturing sector, or through local demand spillover via linked production activities between manufacturing and nonmanufacturing sector (Acemoglu et al., 2016). Interestingly, results for male workers with college above degree suggest that nonmanufacturing employment increases with Chinese import competition. Though the coefficient is only marginally significant, it is economically significant and suggests some evidence of sectoral reallocation effect:high skilled male workers may move to other sectors to recover their job losses in the manufacturing sector. For the high skilled women, I did not find any significant change in their nonmanufacturing employments, but the positive coefficients are statistically indistinguishable from males coefficients. This finding is different than previ- 18

20 ous work by Autor, Dorn, and Hanson (2013) who did not find any nonmanufaturing effect among high-skilled workers 10. It is possible that the average effect in nonmanufacturing employment is masked by the gender heterogeneity in the labor force participation. If high skilled women are less attached to the labor market, then the declining manufacturing sector will drive more women out of the labor market rather than switching to other sectors (Appendix A.1). Next, I show estimated wage effect in Table 3. If Chinese import competition decreases the labor demand more for low-skilled women than men, one may expect a larger wage reduction for low-skilled women than men. Shown in Panel B and C of Table 3, 1000$ increase in the import exposure per worker leads to and percentage points reduction in wage rate for female high school graduate and dropout respectively. To further probe if this wage effect results from a labor supply adjustment along the intensive margin, I examine the effect of import competition on weekly working hours and display the results in column (5)-(6). I find that insignificant change of working hours for neither men nor women, so it is less likely that the wage effect is driven by labor supply change along the intensive margin. I cannot rule out the possibility of labor supply along the extensive margin as the labor nonparticipation rate increases with the Chinese import competition as well (see AppendixA.1). This negative wage effect could be a larger shift of labor demand which creates a downward pressure on wage rate. It could also arise from selection problem that women with higher productivity left the manufacturing sector and leads to a decrease in female manufacturing wage. 6.2 Specification using Gender Separate Import Exposure Previous baseline model relies on the assumption that Chinese import competition generates homogeneous treatment across gender groups. Now I consider a gender-specific model to re- 10 In their study, the estimated nonmanufacturing employment is around percentage points among workers with no college education. 19

21 lax the assumption and allow for a different source of variation in the import exposure from China. This new model can isolate the variation in female-specific import exposure from the variation in male-specific import exposure due to the initial worker employment distribution within the manufacturing sector. The gender-specific import exposure model shares a common feature with Autor, Dorn, and Hanson (2017) but differs in looking at the industry specialization variation within the manufacturing sector. explores the effect of marriage rate for the young men by Chinese import competition and create the industry specialization using industrial employment by overall employment across all sectors. In their setting, they use the share of workers concentrating in a specific industry in relative to all workers rather than just manufacturing workers. It is reasonable for them to do so because Chinese shock may not affect the marriage outcomes only to manufacturing workers but may spread the effect to other sectors via a local affiliation of men and women in the same marriage market. In contrast, this paper explores how men and women differ in their labor adjustments to trade shock. I did not consider the spillover effect to workers outside the manufacturing sector at this moment to simplify my analysis. Male workers are much more likely to concentrate in the manufacturing sector while female workers are more likely to work outside the manufacturing sector (see Appendix Figure A.2). Even though the total import exposure is larger for women than men in manufacturing since women work in most trade affected industries, the import exposure for overall men is larger than it for overall women in the same labor market (Figure 2). Here, my genderspecific import exposure only accounts for import exposure within the manufacturing sector (Figure 2, bottom). The modified sex-separate model is slightly different than previous baseline model in the 20

22 way that I replace the denominator with the manufacturing employment by gender: L M it = β M IP W M it + X M it + γ t + e it (8) L F it = β F IP W F it + X F it + γ t + e it (9) where L M it and L F it are the male and female employment or wage change of a decade period t in commuting zone i. Instead of using gender combined measure for industrial specialization in import-intensive industries, I separate the share of commuting zone i in US by male employment and female employment in industry j and construct male and female industrial specialization respectively. The data source for gender-specific industrial specialization comes from Census IPUMS data for 1980, 1990 and 2000 rather than the County Business Pattern (CBP) data originally used in the baseline model. One problem is that the Census survey sample does not cover all counties in the US. It may generate measurement error for the gender-specific import exposure. To address this concern, I show the scatter plot of two import exposure measures using different data sources: Census and CBP data set. The strong correlation between the import exposure measured by two data sources supports the reliability of my following analysis (Appendix Figure A.3). Figure 5 Panel A shows the first stage of the gender-specific import exposure measure in the first stage and shows strong predictive power the genderspecific instrument variables. My main results are robust to changing the baseline model to the gender-specific model. The conclusion remains the same that female workers experienced much larger declines in their manufacturing employment and wage in relative to male workers. Figure 5 Panel B and C show the reduced form results for male and female employment as well as weekly wage changes. After controlling the differential import exposure by men and women, I find 21

23 a stronger female manufacturing employment effect than the baseline result. Results from Table 4 basically suggest that differential import exposure that men and women are borne with is not the main explanation for the gender difference their labor outcomes. As one can see in coefficients by Panel C and D in Table 4 are statistically distinguishable. 1000$ rise of import exposure per worker reduces the female manufacturing employment by percentage points but only percentage points for men. A simple decomposition exercise suggest that more than 90 percent of the difference between men and women in the effect of manufacturing employment by Chinese import competition comes from different sensitivity by men and women 11. To examine if Census survey data causes any impreciseness or measurement error, I estimate the baseline model using Census IPUMS data and show the results in Panel B. Results in Panel A of Table 4 are the baseline estimates by CBP data which are the same as the ones in Section 6. Though Census estimates are slightly larger than CBP estimates, the difference between two data sources is tiny. 7 Explaining Gender Heterogeneity 7.1 Age Effect Local labor demand shift could have differential impacts on the outcomes of different age groups. Though old workers are less likely to lose their jobs than their younger counterparts as the former group has longer job attachment with firms, the long-run unemployment effect may be more prominent among older workers as it is more difficult for the older workers get back to new work. If manufacturing firms are reluctant to rehire old workers, then the unemployment duration spell hits more intensively on the older workers. In addition, previous literature on aging workforce also finds evidence that older women are underrepresented in 11 The change in male and female employment difference can be decomposed using the average male and female import exposure shown in the summary statistics. 22

24 the labor market (Vandenberghe, 2011). In this case, I examine the employment effect for old workers and find that old workers beyond 40 also suffered from China s rise by reducing their employment. When comparing the manufacturing employment effect across gender-age groups, I find a larger decrease in old women. This age heterogeneity is fully responsible for the rising gender inequality in my main finding. Table 5 shows the manufacturing employment effect across gender-age-education groups. The response to Chinese import competition by young men and women is similar and statistically indistinguishable across all education groups (Column (1)-(2)) 12. However, I find a significant different response by old men and old women. Displayed in column (3)-(4), for old women with high school and below education, the manufacturing employment decreases by around and percentage points. In contrast, the decline in manufacturing employment is much smaller among old men. I only find the significant decline in the manufacturing employment by high school graduate men (0.260 percentage points). The estimated results in Table 5 tell a striking story that the high-skilled old women 13 experienced a smaller manufacturing employment decline than the middle and low skilled old women. However, the negative manufacturing employment by middle and low-skilled women also offset part of their losses outside the manufacturing sector. Displayed in Column (8), $ 1000 increase in import exposure leads to a percentage point increase in the nonmanufacturing employment by middle-skilled old women. Comparing the magnitude of impacts across sectors, I find that the increasing nonmanufacturing employment offsets approximately half of the decreasing manufacturing employment among middle skilled old women. While the manufacturing employment for high skilled old women decreases by percentage points, I did not find a corresponding increase of employment in the nonmanu- 12 This finding is similarly as the study by Autor, Dorn, and Hanson (2017). 13 High skilled workers refer to those with college and above education. Middle-skilled are workers graduated with high school degree and low skilled workers are workers with less than high school graduation. 23

25 facturing sector. I examine the change in labor market nonparticipation rate to see if high skilled old women change their labor supply along the extensive margin in respond to the import competition. I find an increase in their labor market nonparticipation rate by percentage points (Appendix Table A.1) which is statistically equal to their manufacturing employment change. Instead of migrating to the nonmanufacturing sector, it seems that high skilled women are more likely to exit the labor market. Since most old women beyond 40 are married, it might be the case that marriage provides a buffer for these high skilled women and reduces their incentive to stay in the labor market. Now, I turn to the middle- and low-skilled old women who experienced the largest decline in the manufacturing employment. For low-skilled old women, though I did not find any significant change in their nonmanufacturing employment in Table 5. This might be masked by the changing demand in the nonmanufacturing sector towards low-skilled workers who worked in nonmanufacturing service occupations that are closely linked to manufacturing activities. I also examine the age specific employment and wage outcomes using gender-separate model to see if the results are explained by variation in gender-specific import exposure. As Appendix A.2 shows, the conclusion does not change after controlling the source of variation in the import exposure by gender. Old women experience a much larger manufacturing decline in both employment and wage. With $1000 increase of female import exposure, the weekly wage rate for old women decreases by 1 percent in the manufacturing sector. 7.2 Discrimination Effect Previous literature on globalization and gender wage gap suggests that increased competition from international trade led to a narrowing gender wage gap during the 1980s Black and 24

26 Brainerd (2004). The manufacturing product market iin the US is confronted with increasing competition due to more imports from China to the US market. According to Becker s theory of taste based discrimination, we expect to see discriminatory firms in the manufacturing sector lost profits by hiring more expensive male workers than female workers. In the long run, discriminatory manufacturing firms will run out of business and exit the local markets and only nondiscrimnatory firms remain in the market. However, the manufacturing employment declines more for female workers than male workers, which works in the opposite direction to Becker s taste based model. I further investigate the discrimination effect by decomposing the average gender wage gap into explained component and unexplained component. By doing so, it provides better understanding by which component in the average gender wage gap that Chinese import competition works through. Following previous studies about the gender wage gap, I decompose the wage used an Oaxaca-Blinder decomposition (Blau and Kahn, 2017). I first conduct a simple OLS model at individual level by regressing log weekly wage for each individual worker on the observable explanatory variables (education, age, marital status, race, industry and occupation) for each commuting zone for each year. This generates a matrix of coefficients. L M,cz it = β cz,t M XM,cz it + e cz it (10) L F,cz it = β cz,t F XF,cz it + e cz it (11) M = male, F = female, cz = 1, 2, 3,..., 722, t = 1990, 2000, 2007 where i indexes for worker i, cz is the commuting zone. X M,cz it and X F,cz it which include education, age, marital status, race, industry and occupation. are control variables The average wage difference between men and women at commuting zone level is then 25

27 measured by: L M cz,t L F cz,t = β cz,t M ( X M cz,t X F cz,t) + X cz,t(β F cz,t M βcz,t F ) (12) The first term on the far right hand side is explained part of wage gap. The explained portion is the estimated gender differences due to observable characteristics in education, age, race, marrital status, industry and occupation. The second term on the right hand side is the unexplained wage gap between men and women that are not explained by observable differences in workers characteristics. Back to the baseline model, I test the impacts of Chinese import competition on explained and unexplained portion of the wage gap using the 2SLS stacked difference model: Explained cz,t = β exp IP W cz,t + X cz,t + γ t + e cz,t (13) Unexplained cz,t = β unexp IP W cz,t + X cz,t + γ t + e cz,t (14) Table A.3 report estimated results on explained and unexplained part of the gender wage gap by Chinese import competition. The negative coefficients on the unexplained wage gap suggest Chinese import competition reduces the discrimination which generates a downard pressure on the gender wage gap, however, the effect on explained wage gap is much statistically more significant which works opposite to the discrimination effect. This results in an increasing wage gap between men and women overall by Chinese import competition. 7.3 Job Switches between Sectors One concern here is the age composition by old men and old women might be different within manufacturing sector. If old women who left the manufacturing sector overall had a younger age and are more wanted by firms, then the positive female nonmanufacturing employment effect I find is due to worker composition difference by age. I provide further 26

28 empirical evidence on the age distribution by male and female workers who are beyond 40 in the manufacturing sector, to see if the gender difference in the sector employment is merely driven by age composition effect. Figure 6 compares the 1990 s age distribution for men and women beyond 40. I did not find any observable difference in the age distribution for men and women in the manufacturing sector. Thus, it rules out the possibility that age composition is the primary explanation. To take a closer look at which age group of workers move across sectors under the impact of Chinese import competition, I use a finer defined age group and estimate the employment effect for high school graduates in Figure 7. The coefficient plot basically show the estimated results of the series of regression estimates by each single age group. Each point on the graph represents estimate from a separate regression. For instance, the first point on Male s graph in Panel A is interpreted as $1000 increase in the import exposure per worker leads to a 0.5 percentage point decrease in the manufacturing employment for male workers between The major age group of old women who experienced the largest manufacturing employment decline falls into women with age between $ import exposure per worker generates around percentage points decline in the manufacturing employment. Panel B in Figure 7 displays the coefficient plots of estimated nonmanufacturing employment effect by Chinese import competition. There is a statistically increase in the nonmanufacturing employment for old women between and a marginal effect for old women between and These positive results in nonmanufacturing employment change further supports the evidence of a higher labor mobility of old women to the nonmanufacturing sector. Unlike old women, I did not find any significant positive effect in the nonmanufacturing employment for old men. It is a puzzle why these old men did not move to the manufacturing 27

29 sector. If job opportunities are limited for old workers during economic downturn, then old women should be segregated by the market as well 14. However, I only find a significant increase in nonmanufacturing employment effect for old aged women rather than men. One possible reason is that nonmanufacturing employment effect is masked by the changing labor force participation of male workers. Column (3) in Appendix Table A.1 shows that $1000 import exposure per worker leads to an increase of labor force nonparticipation rate by percentage points, which is almost equal to the size of the decrease in the manufacturing employment by middle skilled old men. It could be old men who are displaced can obtain the alternative income sources from public assistance programs, which are more likely to be discouraged from entering the labor markets 15. However, I cannot test this hypothesis directly due to limitation of accessing to individual take-up of these transfer programs. An alternative possibility is from the added worker theory by Stephens (2002). Married women may change their labor supply to account for the husband s lost income during the downturn. Then more women should stay in the market when their husbands experience employment losses. 7.4 Geographic Variation in Job Tenure The difference in manufacturing employment between old men and women may relate with geographic variation in the hire and separation rates across local labor markets. Women are known as having less firm specific attachment than men as they experience more life cycle events such as marriage, childbirth and family care which interrupt their career. If women in the manufacturing sector have lower job tenure years than men, could the difference in 14 Hutchens (1988) shows that older displaced workers face a much lower probability of finding a job in different sector or occupation than younger workers. Vandenberghe (2011) also shows that older women are less likely to be employed than younger women or men at any ages because their productivity is lower. 15 These include Social Security Disability insurance, Temporary Assistance for Needy Families or Supplemental Nutrition Assistance Program (SNAP). 28

30 the job attachment between men and women explains for why female workers in the manufacturing sector suffer more from Chinese import competition? I use job tenure and occupational mobility data from CPS for the year 1986, 2000 and 2006 which are the years that have job tenure information available, and closest to the periods I use in my sample 16. I mainly use the information on the length of time worked at current job in years. Figure 8 shows the 1990 s tenure distribution for men and women who worked in the manufacturing sector. The density plot suggests little difference in the job tenure years by young men and young women in the manufacturing sector. However, the job tenure length is much higher for old men in comparison to old women in the manufacturing sector. On average, the job tenure year is around 13.1 years for old men while only 8.8 years for old women. Next, I examine whether the observed job tenure difference by men and women across local labor markets can explain for why old women in the manufacturing sector are more likely to be laid off. In the Appendix A.4, I plot the state variation in the job tenure gap between men and women in the manufacturing sector and the nonmanufacturing sector in It shows that men in the manufacturing sector have a much longer job tenure length than women. In the nonmanufacturing sector, the average job tenure year is similar for both men and women. To control for the geographic variation in gender difference in the manufacturing job attachment, I equally split my full sample into two samples based on the average gender gap in the manufacturing job tenure year at state level. This exercise examines whether the gender tenure difference is the factor that drives the gender inequality in the manufacturing labor outcomes or not. 16 CPS job tenure supplements conducted survey every two years since I use 1986 s data to proxy the job tenure in CPS does not provide geographic identifier by Census defined PUMA region. 29

31 Table 6 show results by estimating the baseline model using subsamples divided by the average job tenure length difference by gender. For instance, the high sample includes commuting zones from states with job tenure difference between men and women more than 4.13 years in The low sample includes commuting zones from states with the tenure gap less than 4.13 years. I find that the gender difference in the manufacturing employment effect is more pronounced and significant in the high sample. In column (3)-(4), $1000 import exposure generates a more significant decline for female manufacturing employment by around percentage points. In the low sample, this gender difference is modest and statistically indistinguishable. This initial variation in job tenure seems to explain some of the gender gap in the labor responses to Chinese import competition. 7.5 Growing Service Sector and Women A third potential interpretation for why I did not find a nonmanufacturing employment increase for old men is related with growing service sector in the local labor market. International trade stimulated the international sourcing and improved the performance of the service sector by OECD countries, which provide new job opportunities for women more than men. It is well known that the service sector benefits women more than men as service industry has less requirement of physical strength and women have higher average productivity in the service sector than in the manufacturing sector. Then the sectorial shift from the manufacturing sector to the nonmanufacturing service sector may benefit old women more than men by retaining the displaced old women workers in the labor market. If the sectorial shift is the mechanism behind the gender differences in the manufacturing and nonmanufacturing sector, we expect to see a significant change in the sector concentration ratio by Chinese import growth. To test this hypothesis, I estimate the main specification but replacing the main outcomes with the change of log manufacturing con- 30

32 centration. The manufacturing concentration is measured by the share of men or women in manufacturing versus the total male or female employment in a local labor market. By doing so, the coefficients imply how much the concentration of employment has been changed by Chinese import competition within the local labor market. Table 7 presents the estimates of import competition on the log change of manufacturing employment concentration. The coefficients in column (3) suggest a weak effect of old men s manufacturing concentration change but indicates a significant reduction of old women s concentration in the manufacturing sector by around 4 percent with $1000 increase in import exposure per worker. This positive effect suggests some evidence that Chinese import competition drives more old aged women than men to reallocate in the nonmanufacturing sector. Next, I consider what industry and occupation that old aged women may flow into when they leave the nonmanufacturing sector. Table 8 and Table 9 present the results of the nonmanufacturing employment change by industry and occupation among high school graduates. The dependent variable is the share of employment to working-age population by industry or occupation. Specifically, I focus on workers with high school graduate education to reduce the selection issue arose from education. Moreover, as I discuss in previous section, high school graduate workers are the major workforce in the manufacturing sector which are more representative. Table 8 shows that the majority of old aged women reallocate their labor force in the service-related industries. Specifically, the coefficient for professional and related services shows some evidence that there is approximately percentage point increase in the employment for old women. However, I did not find a similar result for old men. Overall, the effect in the nonmnaufacturing employment for old men is weak. This finding is in line with previous studies that find a positive relationship between service sector growth and female employment growth. The empirical evidence by cross-country analysis suggests a 0.82 correlation between relative female employment change and aggregate service employment change (Rendall, 2014). 31

33 Taking a closer look at the nonmanufacturing employment by occupation (Table 9), I find that old women also prefer to work in service and clerical occupations. Recall that most of low skilled women concentrate in machine operators and assemblers in the manufacturing sector, it is surprising that these women did not stay in the same occupations when they move outside manufacturing. Instead, low skilled old women are mostly rehired by service occupations and clerical administrative support occupations. The comparison of the employment effect across different demographic groups implies a different labor adjustment mechanism by men and women in the manufacturing sector, especially for old workers. Though old women suffered more in the manufacturing sector, they are more likely to move to service related sectors where women have advantage which partially offsets their labor costs due to China s rise. 8 Robustness and Alternative Specification In the remaining section, I examine whether my main results are robust to different robustness exercises. One concern is that the China s import shock is a time-varying shock which increases over time, Goldsmith-Pinkham, Sorkin, and Swift (2018) suggest a pretrend test to rule out the effect by Chinese import growth in the past. I conduct a preperiod falsification test to better understand how Chinese import growth affects the gender-specific labor outcomes in the preperiod. Then I examine the validity of the assumption for my instrument variable by following a gravity approach developed by (Autor et al., 2013). 32

34 8.1 Preperiod I first conduct a falsification test which uses the change in employment and wage by men and women between as the dependent outcomes. By doing so, I can rule out the possibility of reversal causality of gender inequality in the manufacturing sector. In other words, the increasing gender inequality is not a cause for the increasing import competition but a consequence. Shown in Table 10, I regressed the employment and wage outcomes by male and female workers on the future import exposure which is the average import exposure over 1990s and 2000s. Column (1)-(4) of Table 10 show the correlation between the change in the manufacturing employment and nonmanufacturing employment with the future import exposure in preperiod. It suggests weak evidence of gender differences in the employment change between men and women before Though there is a statistical significant correlation between manufacturing employment and future import exposure per worker among high school graduates and college women, the sign of coefficients is opposite to what I found in the post 1990s. This means that trends in the female manufacturing employment in the preperiod are very different to what happened in the post 1990s. Also, it is surprising that the labor condition in the nonmanufacturing sector did not benefit old women as much as it did during post 1990s. The negative correlation between future import growth and nonmanufacturing employment change in column (4) of Table 10 also indicates an increasing labor demand in the manufacturing sector in relative to the nonanufacturing sector for old aged workers. For young workers, I show that the falsification test in Appendix Table A.4. Overall, it suggests little pretrend effect for college above and high school graduate workers. However, the coefficient for high school dropout women manufacturing employment is statistically negative between One may concern the declining employment share among young female low skilled workers in manufacturing sector in the preperiod predicts the falling demand for all women in the manufacturing sector, I control the initial share of female workers in 33

35 the manufacturing sector in the main specification and my estimates do not change (see Appendix TableA.4). 8.2 Alternative Measure So far the main import competition measure uses the change of import growth between China and United States, however, it may be the case that China s rise also enhances the sales of products from the US to the other foreign markets (). To incorporate the indirect effect of China s impacts on the international market, I include the export changes from the US to the other countries. The measure of import exposure should include import growth and the export changes from US to other countries, then the previous measure is modified as below: IP W ex it = j L ijt Import us jt + X o,us,jt o X ojt Import oc jt (15) L it L jt where IP W ex it is the alternative measure incorporating the exposure at international markets to China s growth for local labor market i at time t. The only difference than previous main measure is that model (10) adds another variation of imports between other countries (excluding China) and United States. Instead of adding the imports between US and other countries in the equation (10), I use a weighted average of import growth between China and other countries, Import oc jtl jt, where the weight is equal to the share of total spending in the US products X o,us,jt versus the total amount of spending X ojt in each industry j. One may notice that equation (10) does not use the US export level but use the import growth in other countries from China. Compared with the export growth in the US which might correlate with unobserved supply shock in the US product market, using the import growth between China and other countries mitigates this concern by avoiding the import changes in the US entering the equation. 34

36 Table 11 shows the results of manufacturing and nonmanufacturing employment using alternative import measures. The coefficients are pretty similar to the previous results displayed in Table 5. In column (3) and (4) of Table 11, the coefficients between male and female old aged workers are statistically significant and consistent with my main conclusion that old aged women experienced much larger decline in the manufacturing employment but also increasing the nonmanufacturing employment in relative to old aged male workers. 8.3 Gravity Approach The identification assumption for the instrumental variable approach is that the common component of import growth in US and other developed countries only comes from China s growing productivity or falling trade cost. I verify this assumption by using a gravity model developed by. The basic idea behind is to control any industry fixed effect and country fixed effect which determines China s export supply capability in relative to the US 18. Then residuals from modified model reflect only the changes in China s comparative advantages or falling trading costs rather than any productivity shocks associated with industry or importing country. The gravity model is estimated via: ln(x cjkt ) ln(x ujkt ) = α j + α k + ɛ jkt (16) where X cjkt and X ujkt represent the export level by China and United States to other country k in industry j at time t. α j controls for fixed effect of industry j and it controls for 18 It is the relative export supply capability that determines the equilibrium wage and employment. On one hand, the increasing export supply capability by China decreases wage and employment, on the other hand the increasing export capability by US increases wage and employment. 35

37 the initial comparative advantage by China. α k is the fixed effect of importer country k and controls for the time invariant trend in trading costs involved between importer country k and China or importer country k and United States. If the relative trading cost decreases, then we expect to see an increase of import from China in country k because it is relatively cheaper to import products from China. ɛ jkt is the residual term of the gravity model. Now the gravity based measure can be obtained using the residual term and initial level of import in US: IP W git = j L ijt 1 L it 1 ɛ jt Import jt 1 L jt 1 (17) where ɛ jt is the average of ɛ jkt across all importing countries. Impobyrt jt 1 is the import level in US in industry j from China at initial period. Table 12 shows the simple regression using gravity based import exposure measure. Though the magnitude of coefficients on manufacturing employment change and nonmanufacturing employment change becomes smaller after controlling fixed effects of industry or importing country, the implication of my main finding still holds. Column (4) basically finds a larger decline in old aged female workers manufacturing employment. With $1000 per worker increase in net import exposure to Chinese trade resulting from rising comparative advantage and falling trade costs, it reduces the manufacturing employment by percentage points for women and only percentage points for male workers with high school graduate education. 36

38 9 Conclusion The post-1990s witnessed a slower convergence in the gender wage gap in the developed countries but also experienced a more liberalized trade environment between the US and other developing countries. In this paper, I explore how the gender gap in the manufacturing sector is affected by the impact of Chinese import competition since the 1990s. Standard literature in the gender gap and globalization suggest that the gender wage gap tends to decrease with the increasing import competition in tradable sectors, which supports Becker s taste discrimination model. However, I provide new evidence on the impacts on gender inequality by product market competition. Different than standard finding, I find that Chinese import competition increases the gender inequality in the US local labor market. I find little change on the unexplained gap, which rules out the possibility of discrimination channel that previous literature has focused on. By further examining the manufacturing employment and wage effect by age and education group, I find the gender inequality in the manufacturing sector comes from the declining labor demand towards old women than old men. This gender difference in the labor outcomes among old men and women is not fully explained by the initial employment distribution across sectors but suggests some new evidence that old women may be more sensitive to a negative decline in the manufacturing sector than men. By examining the observable characteristics by workers, I find the average job tenure length is significantly larger for old men than old women. However, this gender gap in job tenure length is only found among old workers. The longer job attachment to manufacturing firms is shown to be important in explaining the observed gender difference in the manufacturing employment effect by China s rise. To explore further how Chinese import competition generates different manufacturing employment consequences for old men and old women, I examine the nonmanufacturing employment and find old women recovered half 37

39 of their employment decline in the manufacturing sector by shifting to the nonmanufacturing sector. The most growing sectors in the nonmanufacturing sector for women are the professional-related sector. This growing service sector in the local labor market provides buffers for women who suffered from import competition. The literature on gender inequality and trade liberalization continues to grow. Increasing women s economic empowerment and achieving gender equality is the key to sustainable development. The gender difference in men and women labor response to increasing Chinese import competition is important for people to pay more attention to improve employment opportunities for women who are in the manufacturing sector. In addition, studying the labor adjustments to trade liberalization from the gender dimension also provides new implications for understanding who bears the trade cost and how trade generates distributional effects across gender groups. 38

40 References Acemoglu, Daron et al. (2016). Import competition and the great US employment sag of the 2000s. In: Journal of Labor Economics 34.S1, S141 S198. Autor, David et al. (2010). The polarization of job opportunities in the US labor market: Implications for employment and earnings. In: Center for American Progress and The Hamilton Project 6. Autor, David, David Dorn, Gordon Hanson, et al. (2017). When work disappears: manufacturing decline and the falling marriage-market value of men. National Bureau of Economic Research. Autor, David H et al. (2014). Trade adjustment: Worker-level evidence. In: The Quarterly Journal of Economics 129.4, pp Autor David, H (2018). Trade and labor markets: Lessons from Chinas rise. In: IZA World of Labor. Autor David, H, David Dorn, and Gordon H Hanson (2013). The China syndrome: Local labor market effects of import competition in the United States. In: American Economic Review 103.6, pp Benguria, Felipe and Josh Ederington (2017). Decomposing the Effect of Trade on the Gender Wage Gap. In: Black, Sandra E and Elizabeth Brainerd (2004). Importing equality? The impact of globalization on gender discrimination. In: ILR Review 57.4, pp Blau, Francine D and Lawrence M Kahn (2017). The gender wage gap: Extent, trends, and explanations. In: Journal of Economic Literature 55.3, pp Carlsson, Magnus and Stefan Eriksson (2017). The effect of age and gender on labor demand: evidence from a field experiment. Dai, Mi, Wei Huang, and Yifan Zhang (2018). How Do Households Adjust to Trade Liberalization? Evidence from China s WTO Accession. In: 39

41 Dix-Carneiro, Rafael and Brian K Kovak (2017). Trade liberalization and regional dynamics. In: American Economic Review , pp Do, Quy-Toan, Andrei A Levchenko, and Claudio Raddatz (2014). Comparative advantage, international trade, and fertility. The World Bank. Dorn, David (2009). Essays on inequality, spatial interaction, and the demand for skills. PhD thesis. Verlag nicht ermittelbar. Goldberg, Pinelopi K and Nina Pavcnik (2016). The effects of trade policy. In: Handbook of commercial policy. Vol. 1. Elsevier, pp Goldsmith-Pinkham, Paul, Isaac Sorkin, and Henry Swift (2018). Bartik Instruments: What, When, Why, and How. Tech. rep. National Bureau of Economic Research. Hoynes, Hilary, Douglas L Miller, and Jessamyn Schaller (2012). Who suffers during recessions? In: Journal of Economic perspectives 26.3, pp Hutchens, Robert M (1988). Do job opportunities decline with age? In: ILR Review 42.1, pp Maestas, Nicole, Kathleen Mullen, and David Powell (2013). The Effect of Local Labor Demand Conditions on the Labor Supply Outcomes of Older Americans. In: Molloy, Raven, Christopher L Smith, and Abigail Wozniak (2011). Internal migration in the United States. In: Journal of Economic perspectives 25.3, pp Müller, Steffen, Jens Stegmaier, and Moises Yi (2015). Industry Mix, Local Labor Markets, and the Incidence of Trade Shocks. In: Pierce, Justin R and Peter K Schott (2016). The surprisingly swift decline of US manufacturing employment. In: American Economic Review 106.7, pp Rendall, Michelle (2014). The service sector and female market work. In: Sauré, Philip U and Hosny Zoabi (2009). Effects of trade on female labor force participation. In: Stephens Jr, Melvin (2002). Worker displacement and the added worker effect. In: Journal of Labor Economics 20.3, pp

42 Topel, Robert H (1986). Local labor markets. In: Journal of Political economy 94.3, Part 2, S111 S143. Vandenberghe, Vincent (2011). Boosting the employment rate of older men and women. In: De Economist 159.2, p Wooldridge, Jeffrey M (2015). Introductory econometrics: A modern approach. Nelson Education. 41

43 Table 1: Summary Statistics Male Female (1) (2) (3) (4) (5) (6) Imports from China to US per worker (1000$) (4.35) (4.35) All Sector: Imports from China to US per male worker (1000$) 5.04 (2.73) Imports from China to US per female worker (1000$) 3.16 (2.15) Within Manufacturing: Imports from China to US per male worker (1000$) (7.33) Imports from China to US per female worker (1000$) (7.30) Percentage of working-age population (employed in manufacturing %) (7.14) (6.11) (2.86) (4.86) (2.93) (2.89) Percentage of working-age population (employed in nonmanufacturing %) (7.43) (6.88) (2.73) (5.25) (3.38) (3.08) Percentage of high school dropout (employed in manufacturing) (6.99) (6.2) (4.01) (7.34) (5.24) (4.73) Percentage of woring-age population ((age > 39) employed in manufacturing) (8.36) (7.48) (3.66) (5.41) (4.23) (3.19) Labor force non-participation rate (%) (1.62) (1.32) (0.89) (2.07) (1.34) (1.3) Average log weekly wage (manufacturing) (0.18) (0.19) (0.09) (0.18) (0.21) (0.1) Average log weekly wage (non-manufacturing) (0.18) (0.16) (0.06) (0.2) (0.18) (0.06) Percentage of population with college degree (at initial period of the decade) (6.77) (7.6) (5.74) (6.95) Obs Note: Import data comes from the UN Comtrade data base. All employment and wage outcomes are measured through Census in 1990, 2000 and American Community Survey First row shows the import exposure measure in the baseline model. Gender-specific import exposure measure is calculated via equation (2)-(3). Row (2)-(3) is the gender-specific import exposure using all sector employment as denominator. Row (5)-(6) shows the gender-specific import exosure using manufacturing employment as denominator. All statistics are weighted using initial share of national population at CZ level in each decade. 42

44 Table 2: Chinese Import Growth and Change of Employment to Population Ratio by Gender and Skill Group: OLS and 2SLS Estimates Dependent variable: change in percentage of working-age pop (%pts) OLS 2SLS Manufacturing Nonmanufacturing Manufacturing Nonmanufacturing Male Female Male Female Male Female Male Female (1) (2) (3) (4) (5) (6) (7) (8) Panel A. College and Above: imports from China *** *** *** 0.252* to US/worker (0.052) (0.022) (0.056) (0.047) (0.128) (0A.099) (0.133) (0.110) 43 Panel B. High School Graduates: imports from China ** ** 0.100* *** *** to US/worker (0.041) (0.034) (0.054) (0.045) (0.109) (0.091) (0.109) (0.105) Panel C. High School Dropouts: imports from China *** *** *** to US/worker (0.088) (0.073) (0.129) (0.105) (0.106) (0.115) (0.178) (0.231) Note: N=1444 (two periods 722 commuting zones). This figure shows the estimated results of OLS model and 2SLS model. Panel A-C shows manufacturing employment effect by different demographic groups. Each panel uses the change of manufacturing employment as a share of working-age population that fell into the given demographic group as dependent variable. Estimate in each cell is obtained via regressing the dependent variables on the import exposure per worker. All regressions include full controls for initial share of manufacturing employment, share of population with college education, share of foreign-born population, female employment rate, share of employment in routine occupations, offshorability, eight census division dummies and time fixed effect. Models are weighted using initial share of national population at CZ level in each decade. Robust standard errors in parentheses are clustered to state level. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level.

45 Table 3: Chinese Imports and Change in Manufacturing Wage Effect, : 2SLS Estimates (log pts) Dependent variable: Change in Wages and Hours (log pts) Weekly Wage Hourly Wage Weekly Hours Male Female Male Female Male Female (1) (2) (3) (4) (5) (6) Panel A. College and Above: imports from China to US/worker (0.265) (0.521) (0.209) (0.423) (0.110) (0.125) Panel B. High School Graduates: imports from China *** * to US/worker (0.267) (0.176) (0.279) (0.152) (0.058) (0.083) Panel C. High School Dropouts: imports from China * * to US/worker (0.566) (0.503) (0.471) (0.427) (0.131) (0.267) Note: N=1444 (two periods 722 commuting zones). This figure shows the estimated manufacturing wage effect to import competition by estimating the 2SLS model. Each panel uses the change of log manufacturing wage rate that fell into the given demographic group as dependent variable. Wage sample includes only workers who are fully employed. Hourly wages are computed using annual wage divided by the product of weeks worked and usual weekly hours. Top-coded yearly wages are mutiliplied by a factor of 1.5 and hourly wages are set not to exceed this value divided by 50 weeks times 35 hours. Hourly wage rate below the first percentile of the national hourly wage distribution is set to the value of first percentile. All wage rates are inflated to the year 2007 using the Personal Consumption Expenditure Index. All models are estimated via stacked first difference model and include the initial period commuting zone weekly wage in the manufacturing sector, other basic demographic controls, a time dummy, eight census division dummies. Robust standard errors in parentheses are clustered to state level. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 44

46 Table 4: Chinese Imports and Change in Employment and Wage Effect by Gender-Separate Model: Dependent variable: Change in Employment (%pts) and Weekly Wage (log pts) Manufacturing Employment (%) Manufacturing Weekly Wage Male Female Male Female (1) (2) (3) (4) Panel A. Import Exposure (CBP) imports from China *** **** *** to US/worker (0.089) (0.084) (0.003) (0.003) Panel B.Import Exposure (Census) imports from China *** *** *** to US/worker (0.089) (0.090) (0.003) (0.003) Panel C. Male Import Exposure gender specific imports * from China to US/male worker (0.023) (0.101) Panel D. Female Import Exposure gender specific imports *** from China to US/ female worker (0.017) (0.127) Male predicted imports from 1.247*** China to US/ male worker (0.100) First Stage Female predicted imports from 1.148*** China to US/ female worker (0.099) Demographic Controls Yes Yes Census Division Yes Yes Obs R square Note:N=1444. This table compares the estimated manufacturing employment and wage effect by Chinese import competition between gender-combined and gender-separate import exposure measure. Panel A and B show results predicted by baseline model which uses gender-combined import exposure measure (equation (4)-(5)). Panel A uses the County Business Pattern data and Panel B uses Census IPUMS data to construct the sex-combined import exposure measure. The difference in estimated employment and wage effect in Panel A and B is tiny. Panel C and D estimate male and female outcomes via a gender-separate model in equation (8)-(9), the data source is from Census IPUMS data set. First stage coefficients are obtained by regressing observed gender-specific import exposure on the predicted gender-specific import exposure respectively. All models include full controls as baseline model. Robust standard errors in parentheses are clustered to state level. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 45

47 Table 5: Chinese Import Growth and Employment Effect, Gender, Skill and Age Dependent variable: change in percentage of working-age pop (%pts) Manufacturing Nonmanufacturing Age Age Age Age Male Female Male Female Male Female Male Female (1) (2) (3) (4) (5) (6) (7) (8) Panel A. College and above: imports from China *** *** * 0.502** 0.403*** to US/worker (0.202) (0.153) (0.125) (0.082) (0.212) (0.149) (0.131) (0.100) Panel B. High School Graduates: imports from China *** *** ** *** 0.268* * to US/worker (0.145) (0.108) (0.127) (0.102) (0.161) (0.136) (0.095) (0.124) 46 Panel C. High School Dropouts: imports from China * *** *** *** *** to US/worker (0.101) (0.111) (0.202) (0.142) (0.223) (0.243) (0.233) (0.310) Note: N=1444 (two periods 722 commuting zones). This table shows estimated employment effect in the manufacturing and the nonmanufacturing sector. Column (1)-(2) and column (5)-(6) show estimated employment effect for young workers between Column (3)-(4) and column (7)-(8) show estimated employment effect for old workers beyond 40. Estimates in each panel are obtained via regressing the dependent variables of that group on the import exposure per worker. Each panel shows the estimated effect by education group. All regressions include full controls and eight census division dummies. Models are weighted using initial share of national population at CZ level in each decade. Robust standard errors in parentheses are clustered to state level. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level.

48 Table 6: Manufacturing Employment Effect, Gender and Age by High and Low Job Tenure Gap Regions Dependent variable: change in percentage of working-age pop (%pts) High Low Age Age Age Age Male Female Male Female Male Female Male Female (1) (2) (3) (4) (5) (6) (7) (8) imports from China *** *** ** *** *** *** *** from China to US/worker (0.139) (0.158) (0.173) (0.164) (0.094) (0.061) (0.073) (0.064) 47 Observations Adjusted R Note: Data on job tenure comes from CPS job tenure supplement in the year I use the 1986 to represent the job tenure information in 1990 as CPS job tenure supplement is not conducted every single year. Since CPS job tenure does not contain geographic identifier of Census data defines, I use average job tenure by men and women at state level. The median level of men and women difference in the job tenure years at state level is around 4.13 in The high and low sample is split equally based on whether the commuting zone is from a state with average gender gap in the job tenure length above 4.13 or not in the initial period. Column (1)-(4) use commuting zones with average gender difference in job tenure length more than the median level and regress the dependent variables on the import exposure measure. Column (5)-(8) use only sample of commuting zones with less gender difference in the average job tenure years.all regressions include full controls and eight census division dummies. Models are weighted using initial share of national population at CZ level in each decade. Robust standard errors in parentheses are clustered to state level. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level.

49 Table 7: Chinese Import Growth and Change of Manufacturing Concentration: 2SLS Estimates Dependent variable: change in log manufacturing employment share (log pts) Age Age Male Female Male Female (1) (2) (3) (4) Panel A. College and above imports from China *** * to US/worker (0.983) (1.494) (0.906) (1.789) Panel B. High School Graduates imports from China *** *** * *** to US/worker (0.988) (1.178) (0.523) (0.942) Panel C. High School Dropouts imports from China * to US/worker (0.955) (1.604) (0.107) (0.982) Note:N=1444. Dependent variable in this table uses the change of log manufacturing employment to population ratio. All models are estimated via stacked first difference model and include the initial period commuting zone manufacturing employment share in manufacturing sector, other basic demographic controls, a decade dummy, eight census division dummies. Robust standard errors in parentheses are clustered to state. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 48

50 Table 8: Chinese Import Growth and Change of Nonmanufacturing Employment by Industry, 2SLS Estimates: Old Age with High School Education Dependent variable: change in manufacturing employment by working-age population (%pts) Industry Male Female Transportation 0.067* (0.040) (0.033) Communications (0.015) (0.022) Utilities and sanitary services (0.023) (0.013) Wholesale Trade (0.035) (0.024) Retail Trade * (0.055) (0.065) Finance, Insurance, and Real Estate (0.029) (0.044) Business and Repair Services (0.035) (0.048) Personal Services (0.032) (0.026) Entertainment and Recreation Services (0.028) (0.033) Professional and Related Services * (0.059) (0.104) Public Administration (0.067) (0.047) Note: This table shows the nonmanufacturing employment for the old workers with high school education. Each cell shows estimated effect of Chinese import competition on the employment to population ratio within different sector. The industry categories are constructed according to Census definition. All models are estimated via stacked first difference models and include the initial period commuting zone hourly wage in manufacturing sector, other basic demographic controls, a decade dummy, eight census division dummies Robust standard errors in parentheses are clustered to state. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 49

51 Table 9: Chinese Import Growth and Change of Nonmanufacturing Employment by Occupation,2SLS Estimates: Old Age with High School Education Dependent variable: change in manufacturing employment by working-age population (%pts) Occupation Male Female Clerical and retail * (0.042) (0.067) Service * (0.051) (0.108) Product ** (0.017) (0.012) Transportation and mechanic (0.072) (0.047) Operator * (0.022) (0.012) Manager and professional (0.113) (0.069) Note: This table shows the nonmanufacturing employment for the old workers with high school education. Dependent variables are the nonmanufacturing employment to population ratio within different occupations. The occupation categories are constructed according to Census definition. All models are estimated via stacked first difference models and include the initial period commuting zone hourly wage in manufacturing sector, other basic demographic controls, a decade dummy, eight census division dummies Robust standard errors in parentheses are clustered to state. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 50

52 Table 10: Falsification Exercise: Chinese Import Growth and Employment Status by Old Workers between Dependent variable: change in employment share (% pts) Manufacturing Nonmanufacturing Male Female Male Female (1) (2) (3) (4) Panel A. College and above imports from China *** *** to US/worker (0.139) (0.111) (0.140) (0.100) Panel B. High School Graduates imports from China 0.378** 0.236*** ** *** to US/worker (0.159) (0.071) (0.163) (0.097) Panel C. High School Dropouts imports from China to US/worker (0.182) (0.158) (0.124) (0.148) Note:N=1444. This table uses the change of employment between It only includes old workers between in this sample and falsification exercise for young workers are shown in the appendix Table A4. All models are estimated via stacked first difference models and include a decade dummy, eight census division dummies. Robust standard errors in parentheses are clustered on state. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 51

53 Table 11: Chinese Import Growth and Change of Employment by Gender, Skill and Age Group using Alternative Import Exposure: 2SLS Estimates Dependent variable: change in percentage of working-age pop (%pts) Manufacturing Nonmanufacturing Age Age Age Age Male Female Male Female Male Female Male Female (1) (2) (3) (4) (5) (6) (7) (8) Panel A. College and above all imports from China *** *** * 0.165*** 0.127*** to US/worker (0.160) (0.123) (0.104) (0.073) (0.084) (0.072) (0.063) (0.066) Panel B. High School Graduates all imports from China *** *** ** *** 0.171* * to US/worker (0.116) (0.079) (0.096) (0.081) (0.085) (0.066) (0.050) (0.068) 52 Panel C. High School Dropouts all imports from China * *** ** * *** to US/worker (0.081) (0.091) (0.161) (0.120) (0.140) (0.114) (0.121) (0.156) Note: N=1444 (two periods 722 commuting zones). This table uses an alternative import exposure defined in equation (10) which incorporates the US exposure to China s rise at international markets. All regressions include full controls and eight census division dummies. Robust standard errors in parentheses are clustered to state level. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level.

54 Table 12: Gravity Approach: Chinese Import Growth and Change of Employment by Gender, Skill and Age Group: OLS Estimates Dependent variable: change in percentage of working-age pop (%pts) Manufacturing Nonmanufacturing Age Age Age Age Male Female Male Female Male Female Male Female (1) (2) (3) (4) (5) (6) (7) (8) Panel A. College and above Residual imports from China *** *** * 0.165*** 0.127*** to US/worker (0.060) (0.058) (0.055) (0.049) (0.084) (0.072) (0.063) (0.066) Panel B. High School Graduates Residual imports from China *** *** * *** 0.171* * to US/worker (0.052) (0.039) (0.044) (0.050) (0.085) (0.066) (0.050) (0.068) 53 Panel C. High School Dropouts Residual imports from China *** *** *** ** * *** to US/worker (0.071) (0.058) (101) (0.082)(0.140) (0.114) (0.121) (0.156) Note: N=1444 (two periods 722 commuting zones). Residual imports from China is the import exposure measure from the gravity model (see equation (12)). Models are weighted using the initial share of national population at CZ level in each decade. Estimates in each panel are obtained via regressing the dependent variables of that group on the import exposure per worker. Column (1)-(2) and column (5)-(6) use young workers between age 18-39, column (3)-(4) and (7)-(8) use old workers between All regressions include full controls and eight census division dummies. Robust standard errors in parentheses are clustered to state level. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level.

55 Figure 1: Geographic Variation in Import Exposure (kusd) Note: N=722. Data source comes from UN Comtrade data. Each figure shows the geographical variation in the import exposure per worker (1000 USD). Top plots use the share of male or female workers in total male or female employment across all sectors as geographic variation in industrial specialization. Bottom plots use the share of male or female workers in manufacturing male or female employment as industrial specialization. 54

56 Figure 2: Gender Separate Import Exposure (kusd) Note: N=722. Top plot shows the distribution in the gender-specific import exposure by calculating the industry specific male or female employment as a percentage of total male or female workers across all sectors. Bottom plot uses industry-specific male or female employment as a percentage of male or female workers within the manufacturing sector. 55

57 Figure 3: Change of Manufacturing Employment by Skill Group (%) Note: N=722. This figure shows the relationship of change in the employment to population and change in import growth between across two broadly defined education groups. X axis is the average import exposure per worker (kusd) by commuting zone groups by ten decile groups. From 1 to 10, the average import exposure per worker increases from 509 to dollars. Y axis shows the average change in the manufacturing employment as a percentage of population in each decile group ranked by the average import exposure per worker (kusd) between

58 Figure 4: Change of Manufacturing Employment by Skill Group (%) Note: N=722. This figure shows the relationship of change in the log hourly wage rate and change in import growth between across two broadly defined education groups. Worker sample only includes individuals who are fully-employed. X axis is the average import exposure per worker (kusd) by commuting zone groups by ten decile groups. From 1 to 10, the average import exposure per worker increases from 509 to dollars. Y axis shows the average change in the manufacturing hourly wage rate in each decile group ranked by the average import exposure per worker (kusd) between

59 Figure 5: Change of Manufacturing Employment by Skill Group (%) Panel A.First Stage Panel B.Reduced Form, Manufacturing Employment %pts Panel C. Reduced Form, Manufacturing Wage (log pts) Note: N=722. Regression models in this figure use change between and include initial manufacturing employment or wage and census dummies. All regressions are weighted by start of period CZ share of national population. 58

60 Figure 6: Age Distribution for Male and Female Workers in Manufacturing (%) Note: Data source comes from Census This figure plots the distribution in age for male and female workers employed in the manufacturing sector. 59

61 Figure 7: Coefficient Plot of Manufacturing Employment by Age Group High School Graduates (%) Panel A Panel B Panel C Note: Each plot displays the coefficients of the employment effect using subsample of workers with age falling into that category. The coefficient plot shows the estimated results of the series of regression estimates by each single age group. Each point on the graph represents the estimate from a single regression. For instance, the first point on Male s graph in Panel A is interpreted as $1000 increase in the import exposure per worker leads to a 0.5 percentage point decrease in the manufacturing employment for male workers between All models are estimated via stacked first difference models and include a decade dummy, eight census division dummies. Robust standard errors in parentheses are clustered on state. Models are 60

62 weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 61

63 Figure 8: Distribution in Job Tenure by Manufacturing Workers in 1990, CPS data (%) Note: Data source comes from CPS job tenure supplement in Each graph shows the density plot for male and female workers who are fully employed in the manufacturing sector. The job tenure length is obtained from the survey question about the number of years the respondent has worked in his/her current job. The left graph shows the job tenure distribution for young workers between and right graph shows the distribution for old workers between

64 Appendix A. Table A.1: Chinese Import Growth and Change of Labor Market Nonparticipation by Gender, Skill and age Group: 2SLS Estimates Dependent variable: change in percentage of working-age pop (%pts) Age Age Male Female Male Female (1) (2) (3) (4) Panel A. College and above imports from China 0.165* 0.155* *** to US/worker (0.089) (0.085) (0.057) (0.073) Panel B. High School Graduates imports from China 0.277*** 0.359*** 0.173** 0.296** to US/worker (0.101) (0.128) (0.073) (0.132) Panel C. High School Dropouts imports from China 0.686*** 0.525*** 0.651** 0.645** to US/worker (0.223) (0.171) (0.210) (0.277) Note: N=1444 (two periods 722 commuting zones). This table shows estimated labor force nonparticipation change for different group of workers. Column (1)-(2) and column (5)-(6) show estimated the labor force nonparticipation effect for young workers between Column (3)-(4) and column (7)-(8) show estimated the labor force nonparticipation effect for old workers beyond 40. Estimates in each panel are obtained via regressing the dependent variables of that group on the import exposure per worker. Each panel shows the estimated effect by education group. All regressions include full controls and eight census division dummies. Models are weighted using initial share of national population at CZ level in each decade. Robust standard errors in parentheses are clustered to state level. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 63

65 Table A.2: Chinese Import Growth and Change of Labor Market Outcomes by Age Group using Gender-Separate Import Measure: 2SLS Estimates Dependent variable: change in employment (%pts) and wage (log pts) Employment Age Age Male Female Male Female (1) (2) (3) (4) Panel A. Manufacturing Employment: Gender-specific imports *** *** *** from China to US/worker (0.021) (0.018) (0.036) (0.023) Panel B. Nonmanufacturing Employment: Gender-specific imports * from China to US/worker (0.032) (0.031) (0.033) (0.032) Panel C. Laborforce Nonparticipation: Gender-specific imports *** from China to US/worker (0.029) (0.025) (0.022) (0.034) Weekly Wage Age Age Male Female Male Female (1) (2) (3) (4) Panel D. Manufacturing: Gender-specific imports *** from China to US/worker (0.333) (0.461) (0.253) (0.378) Panel E. Nonmanufacturing: Gender-specific imports *** ** *** *** from China to US/worker (0.191) (0.151) (0.156) (0.149) Panel F. All Sectors: Gender-specific imports *** *** ** *** from China to US/worker (0.213) (0.153) (0.133) (0.169) Note:N=1444. Gender-specific import exposure is constructed by Census IPUMS data. This table shows results by estimating the gender-separate model in equation (8)-(9). Panel A-C show the employment effect and Panel D-F show the wage effect. Column (1)-(2) use only young workers between and column (3)-(4) use old workers between All models are estimated via stacked first difference models and include the initial period employment/wage in the manufacturing sector, other basic demographic controls, a decade dummy, eight census division dummies. Robust standard errors in parentheses are clustered on state. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 64

66 Table A.3: Decomposition Wage Gap in the Manufacturing Dependent variable: change in wage component (log pts) Manufacturing Weekly wage Hourly wage (1) (2) Panel A. Total Wage Gap imports from China 0.539*** 0.441*** to US/worker (0.192) (0.162) Panel B. Explained Wage Gap imports from China 0.773*** 0.554* to US/worker (0.432) (0.318) Panel C. Unexplained Wage Gap imports from China to US/worker (0.459) (0.331) Note:N=1444. This table regresses the change of explained and unexplained component of the gender gap in the manufacturing sector on the import exposure change between Dependent variables in panel A arethe manufacturing weekly and hourly wage gap between men and women. In Panel B, dependent variables are the explained component by regressing the wage gap on age, race, marital status, education, industry and occupation for each single commuting zone. The estimates are obtained by regressing the change in the explained wage gap on the import exposure. Panel C shows the estimated effect on residual wage by regressing the change in residual wages on the import exposure. All models are estimated via stacked first difference models and include a decade dummy, eight census division dummies. Robust standard errors in parentheses are clustered on state. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 65

67 Table A.4: Falsification Exercise: Chinese Import Growth and Employment Status by Young Workers Dependent variable: change in employment share (% pts) Manufacturing Nonmanufacturing Male Female Male Female (1) (2) (3) (4) Panel A. College and above imports from China 0.192* 0.432*** ** *** to US/worker (0.105) (0.148) (0.114) (0.137) Panel B. High School Graduates imports from China * ** to US/worker (0.178) (0.104) (0.141) (0.087) Panel C. High School Dropouts imports from China *** *** to US/worker (0.121) (0.158) (0.113) (0.165) Note:N=1444. This table uses the change of employment by young workers between All models are estimated via stacked first difference models and include the initial period commuting zone hourly wage in manufacturing sector, other basic demographic controls, a decade dummy, eight census division dummies. Robust standard errors in parentheses are clustered on state. Models are weighted using initial period of commuting zone share of national population in each decade. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. 66

68 Figure A.1: Import Penetration and Female Representation(%) Note: N=722. Import penetration rate is calculated by the ratio of US import change from China to total shipments plus net imports. Female representation is defined as the share of female workers to total workers within a sector. The first y axis is the total female employment to male employment in the manufacturing sector. 67

69 Figure A.2: Distribution in Manufacturing Employment Concentration by Gender 1990 (%) Note: N=722. The manufacturing employment concentration is calculated by the ratio of manufacturing employment to nonmanufacturing employment. 68

70 Figure A.3: Correlation of Census and CBP measured IPW (%) Note: N=722. This figure shows the import exposure meaure from two sources of data. X axis is the main import exposure per worker calculated using County Business Pattern data on employment at industryregion level. Y axis shows the import exposure per worker using another data source: Census survey data. The difference between CBP data and Census data comes from the number of categories in industry and size of sample. CBP data is more comprehensive and covers the whole US but Census is a sample survey. In addition, industry code used by CBP data is based on SIC four digit code but Census industry code is on a more aggregate level. One limitation of CBP data is that it does not break employment into demographic groups. 69

71 Figure A.4: State Variation in Job Tenure Gap by Gender in 1990 Note: N=48. The data source comes from CPS job tenure supplement in The gender gap in job tenure (years) is then calculated by averaging the number of job tenure years for male and female workers to the state level. 70

The China Syndrome. Local Labor Market Effects of Import Competition in the United States. David H. Autor, David Dorn, and Gordon H.

The China Syndrome. Local Labor Market Effects of Import Competition in the United States. David H. Autor, David Dorn, and Gordon H. The China Syndrome Local Labor Market Effects of Import Competition in the United States David H. Autor, David Dorn, and Gordon H. Hanson AER, 2013 presented by Federico Curci April 9, 2014 Autor, Dorn,

More information

Labor Market Adjustments to Trade with China: The Case of Brazil

Labor Market Adjustments to Trade with China: The Case of Brazil Labor Market Adjustments to Trade with China: The Case of Brazil Peter Brummund Laura Connolly University of Alabama July 26, 2018 Abstract Many countries continue to integrate into the world economy,

More information

The impact of Chinese import competition on the local structure of employment and wages in France

The impact of Chinese import competition on the local structure of employment and wages in France No. 57 February 218 The impact of Chinese import competition on the local structure of employment and wages in France Clément Malgouyres External Trade and Structural Policies Research Division This Rue

More information

Explaining the Unexplained: Residual Wage Inequality, Manufacturing Decline, and Low-Skilled Immigration. Unfinished Draft Not for Circulation

Explaining the Unexplained: Residual Wage Inequality, Manufacturing Decline, and Low-Skilled Immigration. Unfinished Draft Not for Circulation Explaining the Unexplained: Residual Wage Inequality, Manufacturing Decline, and Low-Skilled Immigration Unfinished Draft Not for Circulation October 2014 Eric D. Gould Department of Economics The Hebrew

More information

Research Report. How Does Trade Liberalization Affect Racial and Gender Identity in Employment? Evidence from PostApartheid South Africa

Research Report. How Does Trade Liberalization Affect Racial and Gender Identity in Employment? Evidence from PostApartheid South Africa International Affairs Program Research Report How Does Trade Liberalization Affect Racial and Gender Identity in Employment? Evidence from PostApartheid South Africa Report Prepared by Bilge Erten Assistant

More information

Explaining the Unexplained: Residual Wage Inequality, Manufacturing Decline, and Low-Skilled Immigration

Explaining the Unexplained: Residual Wage Inequality, Manufacturing Decline, and Low-Skilled Immigration DISCUSSION PAPER SERIES IZA DP No. 9107 Explaining the Unexplained: Residual Wage Inequality, Manufacturing Decline, and Low-Skilled Immigration Eric D. Gould June 2015 Forschungsinstitut zur Zukunft der

More information

Labor Market Dropouts and Trends in the Wages of Black and White Men

Labor Market Dropouts and Trends in the Wages of Black and White Men Industrial & Labor Relations Review Volume 56 Number 4 Article 5 2003 Labor Market Dropouts and Trends in the Wages of Black and White Men Chinhui Juhn University of Houston Recommended Citation Juhn,

More information

The Electoral Consequences of Rising Trade Exposure ] Rev. March 2, 2017 (First version November 16, 2016)

The Electoral Consequences of Rising Trade Exposure ] Rev. March 2, 2017 (First version November 16, 2016) ANoteontheEffectofRisingTradeExposureonthe2016PresidentialElection [Appendix to Autor, Dorn, Hanson, and Majlesi Importing Political Polarization? The Electoral Consequences of Rising Trade Exposure ]

More information

High Technology Agglomeration and Gender Inequalities

High Technology Agglomeration and Gender Inequalities High Technology Agglomeration and Gender Inequalities By Elsie Echeverri-Carroll and Sofia G Ayala * The high-tech boom of the last two decades overlapped with increasing wage inequalities between men

More information

The Electoral Effects of Chinese Import Competition in the United States

The Electoral Effects of Chinese Import Competition in the United States The Electoral Effects of Chinese Import Competition in the United States Su-Hyun Lee Nanyang Technological University, Singapore 2017 Annual Meeting of the IPES University of Texas, Austin November 18,

More information

The WTO Trade Effect and Political Uncertainty: Evidence from Chinese Exports

The WTO Trade Effect and Political Uncertainty: Evidence from Chinese Exports Abstract: The WTO Trade Effect and Political Uncertainty: Evidence from Chinese Exports Yingting Yi* KU Leuven (Preliminary and incomplete; comments are welcome) This paper investigates whether WTO promotes

More information

Skill Classification Does Matter: Estimating the Relationship Between Trade Flows and Wage Inequality

Skill Classification Does Matter: Estimating the Relationship Between Trade Flows and Wage Inequality Skill Classification Does Matter: Estimating the Relationship Between Trade Flows and Wage Inequality By Kristin Forbes* M.I.T.-Sloan School of Management and NBER First version: April 1998 This version:

More information

Family Ties, Labor Mobility and Interregional Wage Differentials*

Family Ties, Labor Mobility and Interregional Wage Differentials* Family Ties, Labor Mobility and Interregional Wage Differentials* TODD L. CHERRY, Ph.D.** Department of Economics and Finance University of Wyoming Laramie WY 82071-3985 PETE T. TSOURNOS, Ph.D. Pacific

More information

Gender preference and age at arrival among Asian immigrant women to the US

Gender preference and age at arrival among Asian immigrant women to the US Gender preference and age at arrival among Asian immigrant women to the US Ben Ost a and Eva Dziadula b a Department of Economics, University of Illinois at Chicago, 601 South Morgan UH718 M/C144 Chicago,

More information

Volume 35, Issue 1. An examination of the effect of immigration on income inequality: A Gini index approach

Volume 35, Issue 1. An examination of the effect of immigration on income inequality: A Gini index approach Volume 35, Issue 1 An examination of the effect of immigration on income inequality: A Gini index approach Brian Hibbs Indiana University South Bend Gihoon Hong Indiana University South Bend Abstract This

More information

Immigrant Legalization

Immigrant Legalization Technical Appendices Immigrant Legalization Assessing the Labor Market Effects Laura Hill Magnus Lofstrom Joseph Hayes Contents Appendix A. Data from the 2003 New Immigrant Survey Appendix B. Measuring

More information

Immigration and Internal Mobility in Canada Appendices A and B. Appendix A: Two-step Instrumentation strategy: Procedure and detailed results

Immigration and Internal Mobility in Canada Appendices A and B. Appendix A: Two-step Instrumentation strategy: Procedure and detailed results Immigration and Internal Mobility in Canada Appendices A and B by Michel Beine and Serge Coulombe This version: February 2016 Appendix A: Two-step Instrumentation strategy: Procedure and detailed results

More information

5A. Wage Structures in the Electronics Industry. Benjamin A. Campbell and Vincent M. Valvano

5A. Wage Structures in the Electronics Industry. Benjamin A. Campbell and Vincent M. Valvano 5A.1 Introduction 5A. Wage Structures in the Electronics Industry Benjamin A. Campbell and Vincent M. Valvano Over the past 2 years, wage inequality in the U.S. economy has increased rapidly. In this chapter,

More information

Wage Trends among Disadvantaged Minorities

Wage Trends among Disadvantaged Minorities National Poverty Center Working Paper Series #05-12 August 2005 Wage Trends among Disadvantaged Minorities George J. Borjas Harvard University This paper is available online at the National Poverty Center

More information

THE GENDER WAGE GAP AND SEX SEGREGATION IN FINLAND* OSSI KORKEAMÄKI TOMI KYYRÄ

THE GENDER WAGE GAP AND SEX SEGREGATION IN FINLAND* OSSI KORKEAMÄKI TOMI KYYRÄ THE GENDER WAGE GAP AND SEX SEGREGATION IN FINLAND* OSSI KORKEAMÄKI Government Institute for Economic Research (VATT), P.O. Box 269, FI-00101 Helsinki, Finland; e-mail: ossi.korkeamaki@vatt.fi and TOMI

More information

The Impact of Unionization on the Wage of Hispanic Workers. Cinzia Rienzo and Carlos Vargas-Silva * This Version, May 2015.

The Impact of Unionization on the Wage of Hispanic Workers. Cinzia Rienzo and Carlos Vargas-Silva * This Version, May 2015. The Impact of Unionization on the Wage of Hispanic Workers Cinzia Rienzo and Carlos Vargas-Silva * This Version, May 2015 Abstract This paper explores the role of unionization on the wages of Hispanic

More information

Unequal Recovery, Labor Market Polarization, Race, and 2016 U.S. Presidential Election. Maoyong Fan and Anita Alves Pena 1

Unequal Recovery, Labor Market Polarization, Race, and 2016 U.S. Presidential Election. Maoyong Fan and Anita Alves Pena 1 Unequal Recovery, Labor Market Polarization, Race, and 2016 U.S. Presidential Election Maoyong Fan and Anita Alves Pena 1 Abstract: Growing income inequality and labor market polarization and increasing

More information

IMMIGRATION AND LABOR PRODUCTIVITY. Giovanni Peri UC Davis Jan 22-23, 2015

IMMIGRATION AND LABOR PRODUCTIVITY. Giovanni Peri UC Davis Jan 22-23, 2015 1 IMMIGRATION AND LABOR PRODUCTIVITY Giovanni Peri UC Davis Jan 22-23, 2015 Looking for a starting point we can agree on 2 Complex issue, because of many effects and confounding factors. Let s start from

More information

LECTURE 10 Labor Markets. April 1, 2015

LECTURE 10 Labor Markets. April 1, 2015 Economics 210A Spring 2015 Christina Romer David Romer LECTURE 10 Labor Markets April 1, 2015 I. OVERVIEW Issues and Papers Broadly the functioning of labor markets and the determinants and effects of

More information

Household Inequality and Remittances in Rural Thailand: A Lifecycle Perspective

Household Inequality and Remittances in Rural Thailand: A Lifecycle Perspective Household Inequality and Remittances in Rural Thailand: A Lifecycle Perspective Richard Disney*, Andy McKay + & C. Rashaad Shabab + *Institute of Fiscal Studies, University of Sussex and University College,

More information

Rural and Urban Migrants in India:

Rural and Urban Migrants in India: Rural and Urban Migrants in India: 1983-2008 Viktoria Hnatkovska and Amartya Lahiri July 2014 Abstract This paper characterizes the gross and net migration flows between rural and urban areas in India

More information

The Impact of Unionization on the Wage of Hispanic Workers. Cinzia Rienzo and Carlos Vargas-Silva * This Version, December 2014.

The Impact of Unionization on the Wage of Hispanic Workers. Cinzia Rienzo and Carlos Vargas-Silva * This Version, December 2014. The Impact of Unionization on the Wage of Hispanic Workers Cinzia Rienzo and Carlos Vargas-Silva * This Version, December 2014 Abstract This paper explores the role of unionization on the wages of Hispanic

More information

Explaining the 40 Year Old Wage Differential: Race and Gender in the United States

Explaining the 40 Year Old Wage Differential: Race and Gender in the United States Explaining the 40 Year Old Wage Differential: Race and Gender in the United States Karl David Boulware and Jamein Cunningham December 2016 *Preliminary - do not cite without permission* A basic fact of

More information

Labor market consequences of trade openness and competition in foreign markets

Labor market consequences of trade openness and competition in foreign markets Labor market consequences of trade openness and competition in foreign markets Daniel Chiquiar Enrique Covarrubias Alejandrina Salcedo Banco de México January 2016 We analyze the labor market consequences

More information

RESEARCH BRIEF: The State of Black Workers before the Great Recession By Sylvia Allegretto and Steven Pitts 1

RESEARCH BRIEF: The State of Black Workers before the Great Recession By Sylvia Allegretto and Steven Pitts 1 July 23, 2010 Introduction RESEARCH BRIEF: The State of Black Workers before the Great Recession By Sylvia Allegretto and Steven Pitts 1 When first inaugurated, President Barack Obama worked to end the

More information

Industrial & Labor Relations Review

Industrial & Labor Relations Review Industrial & Labor Relations Review Volume 60, Issue 3 2007 Article 5 Labor Market Institutions and Wage Inequality Winfried Koeniger Marco Leonardi Luca Nunziata IZA, University of Bonn, University of

More information

Computerization and Immigration: Theory and Evidence from the United States 1

Computerization and Immigration: Theory and Evidence from the United States 1 Computerization and Immigration: Theory and Evidence from the United States 1 Gaetano Basso (Banca d Italia), Giovanni Peri (UC Davis and NBER), Ahmed Rahman (USNA) BdI-CEPR Conference, Roma - March 16th,

More information

Wage Structure and Gender Earnings Differentials in China and. India*

Wage Structure and Gender Earnings Differentials in China and. India* Wage Structure and Gender Earnings Differentials in China and India* Jong-Wha Lee # Korea University Dainn Wie * National Graduate Institute for Policy Studies September 2015 * Lee: Economics Department,

More information

Trade Liberalization in India: Impact on Gender Segregation

Trade Liberalization in India: Impact on Gender Segregation Trade Liberalization in India: Impact on Gender Segregation ARTNeT/UNDP Workshop on Trade and Gender Linkages 15 th -17 th September 2010 Shilpi Kapur The Energy and Resources Institute OVERVIEW Motivation

More information

Explaining the Deteriorating Entry Earnings of Canada s Immigrant Cohorts:

Explaining the Deteriorating Entry Earnings of Canada s Immigrant Cohorts: Explaining the Deteriorating Entry Earnings of Canada s Immigrant Cohorts: 1966-2000 Abdurrahman Aydemir Family and Labour Studies Division Statistics Canada aydeabd@statcan.ca 613-951-3821 and Mikal Skuterud

More information

Policy brief ARE WE RECOVERING YET? JOBS AND WAGES IN CALIFORNIA OVER THE PERIOD ARINDRAJIT DUBE, PH.D. Executive Summary AUGUST 31, 2005

Policy brief ARE WE RECOVERING YET? JOBS AND WAGES IN CALIFORNIA OVER THE PERIOD ARINDRAJIT DUBE, PH.D. Executive Summary AUGUST 31, 2005 Policy brief ARE WE RECOVERING YET? JOBS AND WAGES IN CALIFORNIA OVER THE 2000-2005 PERIOD ARINDRAJIT DUBE, PH.D. AUGUST 31, 2005 Executive Summary This study uses household survey data and payroll data

More information

Rethinking the Area Approach: Immigrants and the Labor Market in California,

Rethinking the Area Approach: Immigrants and the Labor Market in California, Rethinking the Area Approach: Immigrants and the Labor Market in California, 1960-2005. Giovanni Peri, (University of California Davis, CESifo and NBER) October, 2009 Abstract A recent series of influential

More information

Residual Wage Inequality: A Re-examination* Thomas Lemieux University of British Columbia. June Abstract

Residual Wage Inequality: A Re-examination* Thomas Lemieux University of British Columbia. June Abstract Residual Wage Inequality: A Re-examination* Thomas Lemieux University of British Columbia June 2003 Abstract The standard view in the literature on wage inequality is that within-group, or residual, wage

More information

IS THE MEASURED BLACK-WHITE WAGE GAP AMONG WOMEN TOO SMALL? Derek Neal University of Wisconsin Presented Nov 6, 2000 PRELIMINARY

IS THE MEASURED BLACK-WHITE WAGE GAP AMONG WOMEN TOO SMALL? Derek Neal University of Wisconsin Presented Nov 6, 2000 PRELIMINARY IS THE MEASURED BLACK-WHITE WAGE GAP AMONG WOMEN TOO SMALL? Derek Neal University of Wisconsin Presented Nov 6, 2000 PRELIMINARY Over twenty years ago, Butler and Heckman (1977) raised the possibility

More information

Skilled Immigration and the Employment Structures of US Firms

Skilled Immigration and the Employment Structures of US Firms Skilled Immigration and the Employment Structures of US Firms Sari Kerr William Kerr William Lincoln 1 / 56 Disclaimer: Any opinions and conclusions expressed herein are those of the authors and do not

More information

WhyHasUrbanInequalityIncreased?

WhyHasUrbanInequalityIncreased? WhyHasUrbanInequalityIncreased? Nathaniel Baum-Snow, Brown University Matthew Freedman, Cornell University Ronni Pavan, Royal Holloway-University of London June, 2014 Abstract The increase in wage inequality

More information

English Deficiency and the Native-Immigrant Wage Gap in the UK

English Deficiency and the Native-Immigrant Wage Gap in the UK English Deficiency and the Native-Immigrant Wage Gap in the UK Alfonso Miranda a Yu Zhu b,* a Department of Quantitative Social Science, Institute of Education, University of London, UK. Email: A.Miranda@ioe.ac.uk.

More information

The wage gap between the public and the private sector among. Canadian-born and immigrant workers

The wage gap between the public and the private sector among. Canadian-born and immigrant workers The wage gap between the public and the private sector among Canadian-born and immigrant workers By Kaiyu Zheng (Student No. 8169992) Major paper presented to the Department of Economics of the University

More information

Immigrant-native wage gaps in time series: Complementarities or composition effects?

Immigrant-native wage gaps in time series: Complementarities or composition effects? Immigrant-native wage gaps in time series: Complementarities or composition effects? Joakim Ruist Department of Economics University of Gothenburg Box 640 405 30 Gothenburg, Sweden joakim.ruist@economics.gu.se

More information

Cross-State Differences in the Minimum Wage and Out-of-state Commuting by Low-Wage Workers* Terra McKinnish University of Colorado Boulder and IZA

Cross-State Differences in the Minimum Wage and Out-of-state Commuting by Low-Wage Workers* Terra McKinnish University of Colorado Boulder and IZA Cross-State Differences in the Minimum Wage and Out-of-state Commuting by Low-Wage Workers* Terra McKinnish University of Colorado Boulder and IZA Abstract The 2009 federal minimum wage increase, which

More information

International Import Competition and the Decision to Migrate: Evidence from Mexico

International Import Competition and the Decision to Migrate: Evidence from Mexico DISCUSSION PAPER SERIES IZA DP No. 11346 International Import Competition and the Decision to Migrate: Evidence from Mexico Kaveh Majlesi Gaia Narciso FEBRUARY 2018 DISCUSSION PAPER SERIES IZA DP No. 11346

More information

English Deficiency and the Native-Immigrant Wage Gap

English Deficiency and the Native-Immigrant Wage Gap DISCUSSION PAPER SERIES IZA DP No. 7019 English Deficiency and the Native-Immigrant Wage Gap Alfonso Miranda Yu Zhu November 2012 Forschungsinstitut zur Zukunft der Arbeit Institute for the Study of Labor

More information

Rural and Urban Migrants in India:

Rural and Urban Migrants in India: Rural and Urban Migrants in India: 1983 2008 Viktoria Hnatkovska and Amartya Lahiri This paper characterizes the gross and net migration flows between rural and urban areas in India during the period 1983

More information

Inequality in Labor Market Outcomes: Contrasting the 1980s and Earlier Decades

Inequality in Labor Market Outcomes: Contrasting the 1980s and Earlier Decades Inequality in Labor Market Outcomes: Contrasting the 1980s and Earlier Decades Chinhui Juhn and Kevin M. Murphy* The views expressed in this article are those of the authors and do not necessarily reflect

More information

US Exports and Employment. Robert C. Feenstra University of California, Davis and NBER

US Exports and Employment. Robert C. Feenstra University of California, Davis and NBER US Exports and Employment Robert C. Feenstra University of California, Davis and NBER National Press Club, Washington, D.C., October 4, 2018 Global Decline in Manufacturing Employment in manufacturing

More information

Impacts of International Migration on the Labor Market in Japan

Impacts of International Migration on the Labor Market in Japan Impacts of International Migration on the Labor Market in Japan Jiro Nakamura Nihon University This paper introduces an empirical analysis on three key points: (i) whether the introduction of foreign workers

More information

Raymundo Miguel Campos-Vázquez. Center for Economic Studies, El Colegio de México, and consultant to the OECD. and. José Antonio Rodríguez-López

Raymundo Miguel Campos-Vázquez. Center for Economic Studies, El Colegio de México, and consultant to the OECD. and. José Antonio Rodríguez-López INTERNATIONAL COLLABORATIVE INITIATIVE FOR TRADE AND EMPLOYMENT (ICITE) ICITE REGIONAL CONFERENCE, SANTIAGO, CHILE SESSION 2, PAPER 4 TRADE AND OCCUPATIONAL EMPLOYMENT IN MEXICO SINCE NAFTA Raymundo Miguel

More information

Can free-trade policies help to reduce gender inequalities in employment and wages?

Can free-trade policies help to reduce gender inequalities in employment and wages? Janneke Pieters Wageningen University, the Netherlands, and IZA, Germany Trade liberalization and gender inequality Can free-trade policies help to reduce gender inequalities in employment and wages? Keywords:

More information

EPI BRIEFING PAPER. Immigration and Wages Methodological advancements confirm modest gains for native workers. Executive summary

EPI BRIEFING PAPER. Immigration and Wages Methodological advancements confirm modest gains for native workers. Executive summary EPI BRIEFING PAPER Economic Policy Institute February 4, 2010 Briefing Paper #255 Immigration and Wages Methodological advancements confirm modest gains for native workers By Heidi Shierholz Executive

More information

GLOBALISATION AND WAGE INEQUALITIES,

GLOBALISATION AND WAGE INEQUALITIES, GLOBALISATION AND WAGE INEQUALITIES, 1870 1970 IDS WORKING PAPER 73 Edward Anderson SUMMARY This paper studies the impact of globalisation on wage inequality in eight now-developed countries during the

More information

I'll Marry You If You Get Me a Job: Marital Assimilation and Immigrant Employment Rates

I'll Marry You If You Get Me a Job: Marital Assimilation and Immigrant Employment Rates DISCUSSION PAPER SERIES IZA DP No. 3951 I'll Marry You If You Get Me a Job: Marital Assimilation and Immigrant Employment Rates Delia Furtado Nikolaos Theodoropoulos January 2009 Forschungsinstitut zur

More information

Determinants of Return Migration to Mexico Among Mexicans in the United States

Determinants of Return Migration to Mexico Among Mexicans in the United States Determinants of Return Migration to Mexico Among Mexicans in the United States J. Cristobal Ruiz-Tagle * Rebeca Wong 1.- Introduction The wellbeing of the U.S. population will increasingly reflect the

More information

Bank Deregulation and Racial Inequality in America

Bank Deregulation and Racial Inequality in America Critical Finance Review, 2014, 3: 1 48 Bank Deregulation and Racial Inequality in America Ross Levine 1,YonaRubinstein 2, and Alexey Levkov 3 1 Haas School of Business, University of California, Berkeley

More information

Online Appendix: Robustness Tests and Migration. Means

Online Appendix: Robustness Tests and Migration. Means VOL. VOL NO. ISSUE EMPLOYMENT, WAGES AND VOTER TURNOUT Online Appendix: Robustness Tests and Migration Means Online Appendix Table 1 presents the summary statistics of turnout for the five types of elections

More information

The Employment of Low-Skilled Immigrant Men in the United States

The Employment of Low-Skilled Immigrant Men in the United States American Economic Review: Papers & Proceedings 2012, 102(3): 549 554 http://dx.doi.org/10.1257/aer.102.3.549 The Employment of Low-Skilled Immigrant Men in the United States By Brian Duncan and Stephen

More information

Benefit levels and US immigrants welfare receipts

Benefit levels and US immigrants welfare receipts 1 Benefit levels and US immigrants welfare receipts 1970 1990 by Joakim Ruist Department of Economics University of Gothenburg Box 640 40530 Gothenburg, Sweden joakim.ruist@economics.gu.se telephone: +46

More information

How does international trade affect household welfare?

How does international trade affect household welfare? BEYZA URAL MARCHAND University of Alberta, Canada How does international trade affect household welfare? Households can benefit from international trade as it lowers the prices of consumer goods Keywords:

More information

Latin American Immigration in the United States: Is There Wage Assimilation Across the Wage Distribution?

Latin American Immigration in the United States: Is There Wage Assimilation Across the Wage Distribution? Latin American Immigration in the United States: Is There Wage Assimilation Across the Wage Distribution? Catalina Franco Abstract This paper estimates wage differentials between Latin American immigrant

More information

Poverty Reduction and Economic Growth: The Asian Experience Peter Warr

Poverty Reduction and Economic Growth: The Asian Experience Peter Warr Poverty Reduction and Economic Growth: The Asian Experience Peter Warr Abstract. The Asian experience of poverty reduction has varied widely. Over recent decades the economies of East and Southeast Asia

More information

Globalization, Gender, and the Family

Globalization, Gender, and the Family Globalization, Gender, and the Family Wolfgang Keller and Hâle Utar University of Colorado, CESIfo, CEPR and NBER Bielefeld University, CESIfo This draft: April 2018 Using population register data on all

More information

ARTNeT Trade Economists Conference Trade in the Asian century - delivering on the promise of economic prosperity rd September 2014

ARTNeT Trade Economists Conference Trade in the Asian century - delivering on the promise of economic prosperity rd September 2014 ASIA-PACIFIC RESEARCH AND TRAINING NETWORK ON TRADE ARTNeT CONFERENCE ARTNeT Trade Economists Conference Trade in the Asian century - delivering on the promise of economic prosperity 22-23 rd September

More information

Immigration, Offshoring and American Jobs

Immigration, Offshoring and American Jobs Immigration, Offshoring and American Jobs Gianmarco I.P. Ottaviano, (Universita Bocconi and CEPR) Giovanni Peri, (University of California, Davis and NBER) Greg C. Wright (University of California, Davis)

More information

The Effects of Trade Policy: A Global Perspective

The Effects of Trade Policy: A Global Perspective The Effects of Trade Policy: A Global Perspective Nina Pavcnik Dartmouth College and NBER Conference on Firms, Trade and Development Stanford Center on Global Poverty and Development December 6, 2018 Public

More information

Trends in inequality worldwide (Gini coefficients)

Trends in inequality worldwide (Gini coefficients) Section 2 Impact of trade on income inequality As described above, it has been theoretically and empirically proved that the progress of globalization as represented by trade brings benefits in the form

More information

Cleavages in Public Preferences about Globalization

Cleavages in Public Preferences about Globalization 3 Cleavages in Public Preferences about Globalization Given the evidence presented in chapter 2 on preferences about globalization policies, an important question to explore is whether any opinion cleavages

More information

UNION COLLEGE DEPARTMENT OF ECONOMICS, FALL 2004 ECO 146 SEMINAR IN GLOBAL ECONOMIC ISSUES GLOBALIZATION AND LABOR MARKETS

UNION COLLEGE DEPARTMENT OF ECONOMICS, FALL 2004 ECO 146 SEMINAR IN GLOBAL ECONOMIC ISSUES GLOBALIZATION AND LABOR MARKETS UNION COLLEGE DEPARTMENT OF ECONOMICS, FALL 2004 ECO 146 SEMINAR IN GLOBAL ECONOMIC ISSUES GLOBALIZATION AND LABOR MARKETS The Issues wage inequality between skilled and unskilled labor the effects of

More information

Export markets and labor allocation in a low-income country 1. First version: July 2011 This version: November Abstract

Export markets and labor allocation in a low-income country 1. First version: July 2011 This version: November Abstract Export markets and labor allocation in a low-income country 1 Brian McCaig Wilfrid Laurier University bmccaig@wlu.ca Nina Pavcnik Dartmouth College BREAD, CEPR, and NBER Nina.pavcnik@Dartmouth.edu First

More information

CENTRO STUDI LUCA D AGLIANO DEVELOPMENT STUDIES WORKING PAPERS N April Export Growth and Firm Survival

CENTRO STUDI LUCA D AGLIANO DEVELOPMENT STUDIES WORKING PAPERS N April Export Growth and Firm Survival WWW.DAGLIANO.UNIMI.IT CENTRO STUDI LUCA D AGLIANO DEVELOPMENT STUDIES WORKING PAPERS N. 350 April 2013 Export Growth and Firm Survival Julian Emami Namini* Giovanni Facchini** Ricardo A. López*** * Erasmus

More information

The Impact of Legal Status on Immigrants Earnings and Human. Capital: Evidence from the IRCA 1986

The Impact of Legal Status on Immigrants Earnings and Human. Capital: Evidence from the IRCA 1986 The Impact of Legal Status on Immigrants Earnings and Human Capital: Evidence from the IRCA 1986 February 5, 2010 Abstract This paper analyzes the impact of IRCA 1986, a U.S. amnesty, on immigrants human

More information

The Impact of Interprovincial Migration on Aggregate Output and Labour Productivity in Canada,

The Impact of Interprovincial Migration on Aggregate Output and Labour Productivity in Canada, The Impact of Interprovincial Migration on Aggregate Output and Labour Productivity in Canada, 1987-26 Andrew Sharpe, Jean-Francois Arsenault, and Daniel Ershov 1 Centre for the Study of Living Standards

More information

Is inequality an unavoidable by-product of skill-biased technical change? No, not necessarily!

Is inequality an unavoidable by-product of skill-biased technical change? No, not necessarily! MPRA Munich Personal RePEc Archive Is inequality an unavoidable by-product of skill-biased technical change? No, not necessarily! Philipp Hühne Helmut Schmidt University 3. September 2014 Online at http://mpra.ub.uni-muenchen.de/58309/

More information

Economic assimilation of Mexican and Chinese immigrants in the United States: is there wage convergence?

Economic assimilation of Mexican and Chinese immigrants in the United States: is there wage convergence? Illinois Wesleyan University From the SelectedWorks of Michael Seeborg 2012 Economic assimilation of Mexican and Chinese immigrants in the United States: is there wage convergence? Michael C. Seeborg,

More information

The Impact of Foreign Workers on the Labour Market of Cyprus

The Impact of Foreign Workers on the Labour Market of Cyprus Cyprus Economic Policy Review, Vol. 1, No. 2, pp. 37-49 (2007) 1450-4561 The Impact of Foreign Workers on the Labour Market of Cyprus Louis N. Christofides, Sofronis Clerides, Costas Hadjiyiannis and Michel

More information

Accounting for the role of occupational change on earnings in Europe and Central Asia Maurizio Bussolo, Iván Torre and Hernan Winkler (World Bank)

Accounting for the role of occupational change on earnings in Europe and Central Asia Maurizio Bussolo, Iván Torre and Hernan Winkler (World Bank) Accounting for the role of occupational change on earnings in Europe and Central Asia Maurizio Bussolo, Iván Torre and Hernan Winkler (World Bank) [This draft: May 24, 2018] This paper analyzes the process

More information

LABOUR-MARKET INTEGRATION OF IMMIGRANTS IN OECD-COUNTRIES: WHAT EXPLANATIONS FIT THE DATA?

LABOUR-MARKET INTEGRATION OF IMMIGRANTS IN OECD-COUNTRIES: WHAT EXPLANATIONS FIT THE DATA? LABOUR-MARKET INTEGRATION OF IMMIGRANTS IN OECD-COUNTRIES: WHAT EXPLANATIONS FIT THE DATA? By Andreas Bergh (PhD) Associate Professor in Economics at Lund University and the Research Institute of Industrial

More information

5. Destination Consumption

5. Destination Consumption 5. Destination Consumption Enabling migrants propensity to consume Meiyan Wang and Cai Fang Introduction The 2014 Central Economic Working Conference emphasised that China s economy has a new normal, characterised

More information

Boston Library Consortium Member Libraries

Boston Library Consortium Member Libraries ' M.I.T. LfBRARFES - DEWEY Digitized by the Internet Archive in 2011 with funding from Boston Library Consortium Member Libraries http://www.archive.org/details/exportersskillupoobern working paper department

More information

Government data show that since 2000 all of the net gain in the number of working-age (16 to 65) people

Government data show that since 2000 all of the net gain in the number of working-age (16 to 65) people CENTER FOR IMMIGRATION STUDIES June All Employment Growth Since Went to Immigrants of U.S.-born not working grew by 17 million By Steven A. Camarota and Karen Zeigler Government data show that since all

More information

Outsourcing Household Production: Effects of Foreign Domestic Helpers on Native Labor Supply in Hong Kong

Outsourcing Household Production: Effects of Foreign Domestic Helpers on Native Labor Supply in Hong Kong Outsourcing Household Production: Effects of Foreign Domestic Helpers on Native Labor Supply in Hong Kong Patricia Cortes Jessica Pan University of Chicago Graduate School of Business October 31, 2008

More information

Small Employers, Large Employers and the Skill Premium

Small Employers, Large Employers and the Skill Premium Small Employers, Large Employers and the Skill Premium January 2016 Damir Stijepic Johannes Gutenberg University, Mainz Abstract I document the comovement of the skill premium with the differential employer

More information

Trade, Technology, and Institutions: How Do They Affect Wage Inequality? Evidence from Indian Manufacturing. Amit Sadhukhan 1.

Trade, Technology, and Institutions: How Do They Affect Wage Inequality? Evidence from Indian Manufacturing. Amit Sadhukhan 1. Trade, Technology, and Institutions: How Do They Affect Wage Inequality? Evidence from Indian Manufacturing Amit Sadhukhan 1 (Draft version) Abstract The phenomenon of rising income/wage inequality observed

More information

Online Appendices for Moving to Opportunity

Online Appendices for Moving to Opportunity Online Appendices for Moving to Opportunity Chapter 2 A. Labor mobility costs Table 1: Domestic labor mobility costs with standard errors: 10 sectors Lao PDR Indonesia Vietnam Philippines Agriculture,

More information

Non-Voted Ballots and Discrimination in Florida

Non-Voted Ballots and Discrimination in Florida Non-Voted Ballots and Discrimination in Florida John R. Lott, Jr. School of Law Yale University 127 Wall Street New Haven, CT 06511 (203) 432-2366 john.lott@yale.edu revised July 15, 2001 * This paper

More information

The Rise of the East and the Far East: German Labor Markets and Trade Integration

The Rise of the East and the Far East: German Labor Markets and Trade Integration University of Zurich Department of Economics Center for Institutions, Policy and Culture in the Development Process Working Paper Series Working Paper No. 310 The Rise of the East and the Far East: German

More information

Competitiveness: A Blessing or a Curse for Gender Equality? Yana van der Muelen Rodgers

Competitiveness: A Blessing or a Curse for Gender Equality? Yana van der Muelen Rodgers Competitiveness: A Blessing or a Curse for Gender Equality? Yana van der Muelen Rodgers Selected Paper prepared for presentation at the International Agricultural Trade Research Consortium s (IATRC s)

More information

The Impact of Immigration on Wages of Unskilled Workers

The Impact of Immigration on Wages of Unskilled Workers The Impact of Immigration on Wages of Unskilled Workers Giovanni Peri Immigrants did not contribute to the national decline in wages at the national level for native-born workers without a college education.

More information

Changes in Wage Inequality in Canada: An Interprovincial Perspective

Changes in Wage Inequality in Canada: An Interprovincial Perspective s u m m a r y Changes in Wage Inequality in Canada: An Interprovincial Perspective Nicole M. Fortin and Thomas Lemieux t the national level, Canada, like many industrialized countries, has Aexperienced

More information

Department of Economics Working Paper Series

Department of Economics Working Paper Series Accepted for publication in 2003 in Annales d Économie et de Statistique Department of Economics Working Paper Series Segregation and Racial Preferences: New Theoretical and Empirical Approaches Stephen

More information

NBER WORKING PAPER SERIES IMMIGRANTS' COMPLEMENTARITIES AND NATIVE WAGES: EVIDENCE FROM CALIFORNIA. Giovanni Peri

NBER WORKING PAPER SERIES IMMIGRANTS' COMPLEMENTARITIES AND NATIVE WAGES: EVIDENCE FROM CALIFORNIA. Giovanni Peri NBER WORKING PAPER SERIES IMMIGRANTS' COMPLEMENTARITIES AND NATIVE WAGES: EVIDENCE FROM CALIFORNIA Giovanni Peri Working Paper 12956 http://www.nber.org/papers/w12956 NATIONAL BUREAU OF ECONOMIC RESEARCH

More information

Inequality in the Labor Market for Native American Women and the Great Recession

Inequality in the Labor Market for Native American Women and the Great Recession Inequality in the Labor Market for Native American Women and the Great Recession Jeffrey D. Burnette Assistant Professor of Economics, Department of Sociology and Anthropology Co-Director, Native American

More information

NBER WORKING PAPER SERIES IMMIGRATION, JOBS AND EMPLOYMENT PROTECTION: EVIDENCE FROM EUROPE. Francesco D'Amuri Giovanni Peri

NBER WORKING PAPER SERIES IMMIGRATION, JOBS AND EMPLOYMENT PROTECTION: EVIDENCE FROM EUROPE. Francesco D'Amuri Giovanni Peri NBER WORKING PAPER SERIES IMMIGRATION, JOBS AND EMPLOYMENT PROTECTION: EVIDENCE FROM EUROPE Francesco D'Amuri Giovanni Peri Working Paper 17139 http://www.nber.org/papers/w17139 NATIONAL BUREAU OF ECONOMIC

More information

NBER WORKING PAPER SERIES MEXICAN ENTREPRENEURSHIP: A COMPARISON OF SELF-EMPLOYMENT IN MEXICO AND THE UNITED STATES

NBER WORKING PAPER SERIES MEXICAN ENTREPRENEURSHIP: A COMPARISON OF SELF-EMPLOYMENT IN MEXICO AND THE UNITED STATES NBER WORKING PAPER SERIES MEXICAN ENTREPRENEURSHIP: A COMPARISON OF SELF-EMPLOYMENT IN MEXICO AND THE UNITED STATES Robert Fairlie Christopher Woodruff Working Paper 11527 http://www.nber.org/papers/w11527

More information

I ll marry you if you get me a job Marital assimilation and immigrant employment rates

I ll marry you if you get me a job Marital assimilation and immigrant employment rates The current issue and full text archive of this journal is available at www.emeraldinsight.com/0143-7720.htm IJM 116 PART 3: INTERETHNIC MARRIAGES AND ECONOMIC PERFORMANCE I ll marry you if you get me

More information

Prospects for Immigrant-Native Wealth Assimilation: Evidence from Financial Market Participation. Una Okonkwo Osili 1 Anna Paulson 2

Prospects for Immigrant-Native Wealth Assimilation: Evidence from Financial Market Participation. Una Okonkwo Osili 1 Anna Paulson 2 Prospects for Immigrant-Native Wealth Assimilation: Evidence from Financial Market Participation Una Okonkwo Osili 1 Anna Paulson 2 1 Contact Information: Department of Economics, Indiana University Purdue

More information

Economic Origins of Authoritarian Values. Evidence from Local Trade Shocks in the United Kingdom

Economic Origins of Authoritarian Values. Evidence from Local Trade Shocks in the United Kingdom The : Evidence from Local Trade Shocks in the United Kingdom Cameron Ballard-Rosa Mashail Malik Stephanie Rickard Kenneth Scheve University of Texas, Austin International Political Economy Society 2017

More information