Heterogeneous labor impacts of migration across skill groups: The case of Costa Rica

Blyde, Juan S. Working Paper Heterogeneous labor impacts of migration across skill groups: The case of Costa Rica IDB Working Paper Series, No. IDB-WP-1145 Provided in Cooperation with: Inter-American Development Bank (IDB), Washington, DC Suggested Citation: Blyde, Juan S. (2020) : Heterogeneous labor impacts of migration across skill groups: The case of Costa Rica, IDB Working Paper Series, No. IDB-WP-1145, Inter-American Development Bank (IDB), Washington, DC, https://doi.org/10.18235/0002595 This Version is available at: https://hdl.handle.net/10419/234714 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/3.0/igo/legalcode Heterogeneous Labor Impacts of Migration Across Skill Groups: The Case of Costa Rica Juan S. Blyde IDB WORKING PAPER SERIES No IDB-WP-1145 Inter-American Development Bank Migration Unit August 2020 Heterogeneous Labor Impacts of Migration Across Skill Groups: The Case of Costa Rica Juan S. Blyde Inter-American Development Bank Migration Unit August 2020 Cataloging-in-Publication data provided by the Inter-American Development Bank Felipe Herrera Library Blyde, Juan S. Heterogeneous labor impacts of migration across skill groups: the case of Costa Rica / Juan S. Blyde. p. cm. — (IDB Working Paper Series ; 1145) Includes bibliographic references. 1. Foreign workers-Labor productivity-Costa Rica. 2. Immigrants-Costa Rica. 3. Labor market-Costa Rica. 4. Labor market-Nicaragua. 5. Job analysis-Costa Rica. 6. Job analysis-Nicaragua. I. Inter- American Development Bank. Migration Unit. II. Title. III. Series. IDB-WP-1145 http://www.iadb.org Copyright © 2020 Inter-American Development Bank. This work is licensed under a Creative Commons IGO 3.0 Attribution- NonCommercial-NoDerivatives (CC-IGO BY-NC-ND 3.0 IGO) license (http://creativecommons.org/licenses/by-nc- nd/3.0/igo/legalcode) and may be reproduced with attribution to the IDB and for any non-commercial purpose, as provided below. No derivative work is allowed. Any dispute related to the use of the works of the IDB that cannot be settled amicably shall be submitted to arbitration pursuant to the UNCITRAL rules. The use of the IDB’s name for any purpose other than for attribution, and the use of IDB’s logo shall be subject to a separate written license agreement between the IDB and the user and is not authorized as part of this CC-IGO license. Following a peer review process, and with previous written consent by the Inter-American Development Bank (IDB), a revised version of this work may also be reproduced in any academic journal, including those indexed by the American Economic Association’s EconLit, provided that the IDB is credited and that the author(s) receive no income from the publication. Therefore, the restriction to receive income from such publication shall only extend to the publication’s author(s). With regard to such restriction, in case of any inconsistency between the Creative Commons IGO 3.0 Attribution-NonCommercial-NoDerivatives license and these statements, the latter shall prevail. Note that link provided above includes additional terms and conditions of the license. The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the Inter-American Development Bank, its Board of Directors, or the countries they represent. Heterogeneous Labor Impacts of Migration Across Skill Groups: The Case of Costa Rica∗ Juan S. Blyde† Inter-American Development Bank August, 2020 Abstract Popular empirical strategies that examine the labor impacts of migrants, like the skill-cell approach, are frequently used to measure the effects of immigrants from a particular skill group on native-born workers with similar skills. I use an augmented version of the skill-cell approach to examine the impacts of immigrants on native workers with similar skills but also across skill groups. I apply this approach to the case of Nicaraguan immigrants in Costa Rica. I find large positive employment and wage effects on high-skilled women arising from low-skilled migrants. These positive effects are derived from both the household channel and the complementary-skills channel. I also find negative but small effects on low-skilled native workers. The results show that immigrants can have complex labor market effects on native workers with own and cross elasticities that can be quite different. Keywords: International migration, skill-cell, employment JEL Classification: J60, J61, F22 ∗I would like to thank Anna Maria Mayda, Ana Maria Iba˜nez, Marisol Rodriguez Chatruc, Denisse Pierola, Emmanuel Abuelafia, Camila Cort´es and Fernando Morales for comments and suggestions to an earlier version of this paper. Camila Cort´es provided excellent research assistance. The views and interpretations in this study are strictly those of the author and should not be attributed to the Inter-American Development Bank, its Board of Directors, or any of its member countries †Correspondence address: Juan Blyde. Inter-American Development Bank, 1300 New York Ave., NW, Washington DC, 20755, U.S. Phone: (202) 623-3517, E-mail: juan[email protected] 1 Introduction The effect of immigration on native’s labor market outcomes depends on how substitutable or complementary immigrants and natives are in the labor market. Immigrants are substitutes for native-born workers when they compete for similar jobs which can cause displacement and/or lower wages. Immigrants are complements when they increase the demand for complementary production tasks and skills of the native workers. Popular empirical strategies that examine the labor impacts of migrants, like the skill-cell approach, are frequently used to measure the effects of immigrants on native-born workers with similar skills. But the same immigration shock can both cause substitutability and complementarity in the labor market depending on the skill of the native worker. Consider, for example, low-skilled immigrants working as fruit pickers in the strawberry industry. The immigrants might be willing to supply their labor at lower wages than natives thus lowering the natives’ salaries but also the strawberry grower’s production costs. The growers might then increase output which may require hiring more high-skilled managers to supervise the expansion of production. In this example, the same group of migrants harmed low-skilled native workers by reducing their salaries but benefited high-skilled native individuals by increasing their employability. In this paper I use a skill-cell approach to disentangle this type of differentiated impacts. The traditional skill-cell approach pioneered by Borjas (2003) divides the labor market into skill groups, and the change in immigrant inflows to skill groups is compared with the change in wages within those skill groups. I employ an augmented version of the skill-cell approach to examine the labor market effects that immigrants exert on native workers with similar skills but also across skill levels.1 The analysis is focused on to the case of Nicaraguan migration to Costa Rica; therefore, the study also contributes to expand the literature on South-South migration. Studies that evaluate the impact of migration across developing countries are rare in comparison to analyses that examine migration flows across developed countries or from developing to developed countries. The lack of studies looking at migration flows between developing countries leaves a vacuum in the migration literature because migrants from developing countries arriving to other developing countries might not necessarily have the same labor market effects as migrants arriving to developed countries which typically exhibit deeper labor markets, more mature industries and more resilient institutions. Therefore, this study contributes to a growing number of papers that provide evidence on the impact of South-South migration flows (Hatton and Williamson,2005;Biavaschi et al.,2018;Gindling, 2009). Two recent studies examine the Nicaraguan migration in Costa Rica in terms of the labor market (Gindling,2009;Mora and Guzm´an,2019). This study is mostly related to Gindling (2009) who examines the impact of Nicaraguan immigrants on the earnings of Costa Rican workers. The analysis in Gindling (2009) relies on the pure skill-cell approach and thus it assesses whether immigrants are substitutes or complements to the native population within the same skill level but not across skill levels. As mentioned before, I adopt a novel empirical strategy that allows me to explicitly explore the impact of migrants on natives across skill-cells. As it will be shown below, this is a large effect in the case of Costa Rica and is generally a type of impact that is overlooked 1Borjas (2003) uses a structural approach to measure the wage effects of immigrants on natives with different skills. In this analysis I do not need to rely on a structural model 2 in the empirical analyses that rely on the skill-cell approach. Another difference from Gindling (2009) is that I evaluate the impact of immigration not only on earnings but also on employment. Looking at the employment rate is important because the labor market effects of migration can be absorbed not only through wage changes but also through employment changes. In fact, as the labor supply elasticity increases, the wage effects of migration tend to become more muted while the employment effects become larger (Dustmann et al.,2016). Thus, by looking at the response of earnings only, one might fail to detect important labor market adjustments from migration. I find large positive employment and earnings effects on high-skilled women arising from lowskilled migrants. For example, a 1 percentage point increase in the share of low-skilled Nicaraguan workers raises the likelihood of being employed by about 4.5 percentage points for high-skilled women. These positive effects are derived from both the household channel and the complementaryskills channel. I also find negative but small employment impacts on low-skilled native workers arising from similar skill levels. In general, the findings indicate that migrants can have complex labor market effects on native workers with own and cross elasticities that can be quite different. The rest of the paper is organized as follows. Section 2 discusses the immigration flows between Nicaragua and Costa Rica. Sections 3 presents the empirical strategy. Section 4 describes the datasets employed. Section 5 discuss the results and section 6 provides concluding remarks. 2 Nicaraguan migration to Costa Rica Nicaraguans have migrated to Costa Rica for years; nevertheless, the share of Nicaraguan immigrants in the total population of Costa Rica was still below 2% by the early 1980s. Significant increases in migration took place starting from this period. With the outbreak of the armed conflict between the Sandinista government and the Contra forces in 1984 a period of migration predominantly for political reasons took place until the end of Nicaragua’s civil war in 1990 (Otterstrom, 2008). After this period marked by military conflict, the causes for migration turned more economic in nature, particularly after Nicaragua implemented drastic structural adjustment policies between 1993 and 1997 (IOM,2001). Catastrophic flooding from the slow motion of hurricane Mitch in 1998 also triggered an additional wave of immigration during this time. Accordingly, the share of Nicaraguan immigrants in the total population of Costa Rica increased sharply from 1.95% in 1984 to 5.9% in 2000. Nicaraguan migrants continued to be attracted to Costa Rica after 2000 given the country’s political stability and much higher living standards than Nicaragua’s. The share of Nicaraguans in Costa Rica’s total population increased to 6.13% by 2010 and to 7.01% by 2018.2 Most notably, the share of migrants from Nicaragua in the working-age population increased even more during this period, from 7.82% in 2010 to 9.05% in 2018 (see Figure 1).3In this study I focus on this 2010-2018 period and examine the potential labor market impacts from this rise in predominantly economic migrants. I also concentrate the analysis to this period because this allows me to address potential endogeneity concerns that I discuss in section 3 below. To get a sense of the Nicaraguan immigrants in Costa Rica, table 1 (upper panel) shows some basic demographic characteristics of this population based on Costa Rica’s household survey. The most notable characteristic is that a large percentage of these immigrants are low skilled. In 2018, 2Nicaraguans represent around 75% of the stock of all immigrants in Costa Rica 3The data source of Figure 1 is the Costa Rican household survey, ENAHO 3 for example, 78.6% of the working-age migrants had secondary education incomplete or less, and only 2% had completed tertiary education. This contrasts with the working-age native population in which 58.3% show an incomplete secondary education or less, while 9% possess a tertiary education degree (see lower panel of table 1). Given the low levels of education, Nicaraguans tend to work relatively more than Costa Ricans in low skilled jobs. Table 2 shows, for example, that 50% of Nicaraguans are employed in elementary occupations.4The corresponding figure for Costa Ricans is 22%. Conversely, while 27% of Costa Rican workers are employed as managers, professionals or technicians only 6% of the Nicaraguans hold this type of jobs. The evidence in these tables indicates that because of the relatively low levels of education relative to the Costa Rican population, the majority of the Nicaraguan immigrants tend to be absorbed in low-skilled jobs. The main questions that this paper seeks to address are: how the employment of Nicaraguans in low-skilled jobs affects the native population with similar low skills and whether this low-skilled migration has additional impacts on higher skills Costa Ricans. 3 Empirical methodology To measure the impact of migrants on the labor market I exploit variation of immigrant shares across geographic space and skill-cells (education and experience). This is called the mixture approach (Card,2001;Dustmann et al.,2016) which combines the pure spatial approach led by Altonji and Card (1991) with the skill-cell approach pioneered by Borjas (2003). The mixture approach has been used recently by Llull (2018), Borjas (2006), Card and Peri (2016), Biavaschi et al. (2018). The baseline empirical specification of the typical mixture approach relies on the following functional form: Yijrt =α0+β·mjrt +¯ Xijrt ·γ+αj+αr+αt+ (αj·αr)+(αj·αt)+(αr·αt) + eijrt (1) with: mjrt =Mjrt Mjrt +Njrt (2) where Yijrt is the labor outcome of interest for a native individual iin skill group jin region r at time t;Mjrt is the number of Nicaraguan workers in skill group jin region rat time t;Njrt is the corresponding number of natives; ¯ Xijrt is a vector of individual characteristics, including gender and marital status; αj,αr, and αtare skill-cell, region and year fixed effects, respectively, and eijrt is the error term, which are clustered at the skill-region-year level. Our paramenter of interest is β. Borjas (2003) defines this parameter the ‘own elasticity’ of substitution/complementarity because it measures the labor market impact of a migrant on a native that exhibits the same skills. Below, I introduce additional terms to equation (1) to measure also the impact of a migrant on a native that exhibits different skills (cross-elasticity). The migration shares in (2) are defined at the skill-region-year level.5In order to provide 4Among the elementary occupations, many Nicaraguans work in activities of households as employers, like cleaners or nannies. The share of Nicaraguan workers in these occupations is 18% 5Throughout the paper all the shares are expressed in percent 4 some intuition behind this calculation, I start by describing the calculation of the migration shares by skill group at the national level, mjt =Mjt/(Mjt +Njt). First, I select the number of cells so that there is a high enough number of migrants in each cell. Since I am using a household survey to calculate these shares, having too many cells could lead to a too few observations of Nicaraguan migrants in the survey for some of the cells, particularly for the higher skill cells. This would make the calculation of the migrant shares in those cells noisy. Accordingly, I selected 10 cells that combine different levels of education and experience. In particular, the 10 cells consist of 5 education groups (primary incomplete, primary complete, secondary incomplete, secondary complete and tertiary complete) and two experience levels (below or equal and above 15 years). I calculate these migrant shares for each cell at the national level using the yearly household surveys. For comparison purposes, I also compute these same shares using the 2011 census and contrast them with those from the 2011 household survey. Table 3 shows that the shares from both data sources are very similar which give us confidence that the survey is doing an adequate job capturing the Nicaraguan migration. Now I calculate the migrant shares in (2). For this, I need a measure for the number of migrant workers that varies by skill, region and year, Mjrt. I obtain Mjrt as follows: Mjrt =Mjr1984 Mj1984 ·Mjt (3) where Mjr1984/Mj1984 is the share of Nicaraguan workers in skill group jin region rthat I take from the 1984 census, and Mjt is the number of Nicaraguan workers of skill jat time tat the national level that I obtained from the household surveys.6In expression (3), Nicaraguan migrants of skill group jin year tare apportioned across geographic space according to the historical distribution of Nicaraguan immigrants in the year 1984. Note that this is the basis for the popular shiftshare instrument which interacts national inflows of immigrants with the geographic distribution of the immigrants in the past. The instrument, introduced by Altonji and Card (1991) and further developed by Card (2001) is often employed to address the fact that the location of immigrants across regions is not random as there could be factors that affect the current location decision of immigrants across region that also affect current outcomes Yijrt. If this is the case, the estimates of the migration effects on the labor market could be biased. The instrument rely on the fact that migrants tend to locate in areas where there are already settlements of their co-nationals (Bartel, 1989); therefore, if the settlements formed in the past are uncorrelated with the current outcomes, distributing the migrants by such settlements can address the endogeneity problem. There could be, however, local conditions in the past that could be persistent affecting the location of migrants. For instance, if places that have better employment opportunities attract more migrants and the correlation overtime is strong, this channel can bias the estimates. Therefore, following Mayda et al. (2018), I perform a falsification exercise by regressing the change in the subsequent share of migrants on past outcomes. In particular, I regress the change in the share of migrants at the region-skill level during the treatment period (2010-2018) on changes in two outcome variables (employment rate and average earnings) in a pre-treatment period (2000-2009). In both cases, the coefficients are actually negative but not statistically significant.7 6In principle I could have used the information from the household surveys which is representative at the regional level, but the calculation of the migrant shares for some of the skill-region-year cells could be potentially noisy because there are only a few observations of Nicaraguan migrants in some of these cells 7The estimated coefficients for the employment rate and for the average earnings are -0.143 and -0.008, respectively, with p-values of 0.46 and 0.85, respectively 5 References Altonji, J. G. and Card, D. (1991). The effects of immigration on the labor market outcomes of less-skilled natives. In Abowd, J. M. and Freeman, R. B., editors, Immigration, Trade and the Labor Market, page 201–234. University of Chicago Press. Bartel, A. P. (1989). Where do the new u.s. immigrants live? Journal of Labor Economics, 7(4):371–391. Biavaschi, C., Facchini, G., Mayda, A. M., and Mendola, M. (2018). South-south migration and the labor market: evidence from South Africa. Journal of Economic Geography, 18:823–853. Blanchard, O. J. and Katz, L. F. (1992). Regional evolutions. Brookings Papers on Economic Activity, 1:1–77. Blau, F. D. and Kahn, L. M. (2015). Immigration and the distribution of incomes. In Chiswick, B. and Miller, P., editors, Handbook on the Economics of International Migration, pages 793–843. Elsevier. Borjas, G. J. (2003). The labor demand curve is downward sloping: reexamining the impact of immigration on the labor market. The Quarterly Journal of Economics, 40:1335–1374. Borjas, G. J. (2006). Native internal migration and the labor market impact of immigration. Journal of Human Resources, 41(2):221–258. Card, D. (2001). Immigrant inflows, native outflows, and the local labor market impacts of higher immigration. Journal of Labor Economics, 19:22–64. Card, D. and Peri, G. (2016). Immigration economics: A review. Unpublished paper, University of California. Cort´es, P. and Tessada, J. (2011). Low-skilled immigration and the labor supply of highly skilled women. American Economic Journal: Applied Economics, 3(3):88–123. Dustmann, C., Sch¨onberg, U., and Stuhler, J. (2016). The impact of immigration: Why do studies reach such different results? Journal of Economic Perspectives, 30(4):31–56. Farr´e, L., Gonzalez, L., and Ortega, F. (2011). Immigration, family responsibilities and the labor supply of skilled native women. The BE Journal of Economic Analysis and Policy, 11(1). Gindling, T. H. (2009). South–south migration: the impact of Nicaraguan immigrants on earnings, inequality and poverty in Costa Rica. World Development, 37:116–126. Hatton, T. J. and Williamson, J. G. (2005). What fundamentals drive world migration? In Borjas, G. and Crisp, J., editors, Poverty, International Migration and Asylulm, pages 15–38. Palgrave–MacMillan. Hiller, T. and Chatruc, M. R. (2020). South–south migration and female labor supply in the dominican republic. Unpublished Document. IOM (2001). A binational study: the state of migration flows between Costa Rica and Nicaragua. International Organization for Migration (IOM). 12 Jaeger, D. A., Ruist, J., and Stuhler, J. (2018). Shift-share instruments and the impact of immigration. NBER, Working Paper N. 24285. Llull, J. (2018). The effect of immigration on wages: Exploiting exogenous variation at the national level. Journal of Human Resources, 53(3):608–662. Mayda, A. M., Peri, G., and Steingress, W. (2018). The political impact of immigration: Evidence from the united states. NBER, Working Paper N. 24510. Mora, A. and Guzm´an, M. (2019). Aspectos de la migraci´on nicaraguense hacia Costa Rica e impacto en el mercado laboral. Unpublished paper, Banco Interamericano de Desarrollo. Otterstrom, S. M. (2008). Nicaraguan migrants in Costa Rica during the 1990s: Gender differences and geographic expansion. Journal of Latin American Geography, 7(2). Peri, G. and Sparber, C. (2009). Task specialization, immigration and wages. American Economic Journal: Applied Economics, 1(3):135–169. 13 Figure 1: Immigrants from Nicaragua 14 Table 1: Descriptive statistics of individuals aged 15-65 (percent) Nicaraguans 2010 2014 2018 Age 33.9 36.0 37.5 Female (%) 54.3 54.7 54.2 Secondary incomplete or less (%) 84.2 78.8 78.6 Secondary complete (%) 14.9 19.1 19.3 Tertiary complete or more (%) 0.9 2.1 2.2 Costa Ricans 2010 2014 2018 Age 35.6 36.8 37.6 Female (%) 51.3 51.1 51.6 Secondary incomplete or less (%) 64.2 58.9 58.3 Secondary complete (%) 28.6 33.8 32.6 Tertiary complete or more (%) 7.2 7.3 9.1 15 Table 2: Distribution of employed individuals by occupation, 2018 (percent) Occupation (ISCO-08) Costa Ricans Nicaraguans Managers 2.0 0.4 Professionals 13.9 1.9 Technicians and Associate Professionals 10.7 3.4 Clerical Support Workers 9.0 2.5 Services and Sales Workers 21.6 21.8 Skilled Agr., Forestry and Fishery Workers 3.4 3.6 Craft and Related Trades Workers 10.2 12.7 Plant and Machine Operators and Assemblers 7.6 3.5 Elementary Occupations 21.6 50.0 Not specified 0.1 0.3 16 Table 3: Migration shares by education and experience (percent) Education Experience Survey 2011 Census 2011 Primary incomplete years ≤15 34 33 years >15 23 25 Primary complete years ≤15 14 13 years >15 7 9 Secondary incomplete years ≤15 13 12 years >15 13 12 Secondary complete years ≤15 7 7 years >15 6 7 Tertiary complete years ≤15 2 2 years >15 3 3 17 Table 4: Summary statistics, main variables 2010 2014 2018 Employment: Mean 0.92 0.91 0.91 Employment: Std. dev 0.27 0.29 0.28 Log of real earnings: Mean 6.02 6.11 6.19 Log of real earnings: Std. dev 0.79 0.83 0.78 Migration shares: Same skill cell: Mean 8.23 8.80 9.72 Same skill cell: Std. dev 7.47 7.46 7.56 Lower skill cells: Mean 15.77 16.63 18.47 Lower skill cells: Std. dev 12.22 11.34 12.95 Higher skill cells: Mean 4.13 4.66 4.80 Higher skill cells: Std. dev 2.60 2.58 2.59 Lower skill cells within high-skilled category: Mean 15.77 16.63 18.47 Lower skill cells within high-skilled category: Std. dev 12.22 11.34 12.95 Higher skill cells within low-skilled category: Mean 4.13 4.66 4.80 Higher skill cells within low-skilled category: Std. dev 2.60 2.58 2.59 18 Table 5: Baseline results Employment Earnings All Men Women All Men Women Migration share: (1) (2) (3) (4) (5) (6) Same skill cell -0.0035*** -0.0025*** -0.0055*** -0.0012 0.0006 -0.0023 (0.0009) (0.0009) (0.0019) (0.0022) (0.0026) (0.0050) R-squared 0.0418 0.0322 0.0536 0.3217 0.3072 0.3483 Observations 133,054 82,417 50,637 111,125 69,410 41,715 Notes: The dependent variable in (1)-(3) is a dummy equal to 1 if the individual is employed and 0 if unemployed. The dependent variable in (4)-(6) is the log of real hourly earnings of the employed individuals. The main explanatory variable is the share of Nicaraguan workers in the labor force. Additional controls include gender and marital status and fixed effects for skill (education and experience), year, region and any two-way interaction FE. Robust standard errors adjusted for clustering at the skill-region-year level are in parentheses *** ; ** ; * significant at the 1%, 5% and 10% level respectively 19 Table 6: Migration impacts differentiated by educational groups Employment Earnings Low-skilled High-skilled Low-skilled High-skilled Men Women Men Women Men Women Men Women Migration share: (1) (2) (3) (4) (5) (6) (7) (8) Same skill cell -0.0028*** -0.0044** 0.0003 -0.0065 0.0028 -0.0048 -0.0133 -0.0187 (0.0010) (0.0021) (0.0031) (0.0051) (0.0026) (0.0059) (0.0154) (0.0153) R-squared 0.0323 0.0538 0.0301 0.0506 0.0581 0.0312 0.2462 0.2581 Observations 53,921 24,717 28,496 25,920 43,807 19,304 25,603 22,411 Notes: The dependent variable in (1)-(4) is a dummy equal to 1 if the individual is employed and 0 if unemployed. The dependent variable in (5)-(8) is the log of real hourly earnings of the employed individuals. The main explanatory variable is the share of Nicaraguan workers in the labor force in the same skill-cell as the individual in the dependent variable. Additional controls include gender and marital status and fixed effects for skill (education and experience), year, region and any two-way interaction FE. Robust standard errors adjusted for clustering at the skill-region-year level are in parentheses *** ; ** ; * significant at the 1%, 5% and 10% level respectively 20 Table 7: Migration impacts from similar and across skills Employment Earnings Low-skilled High-skilled Low-skilled High-skilled Men Women Men Women Men Women Men Women Migration shares: (1) (2) (3) (4) (5) (6) (7) (8) Same skill cell -0.0028*** -0.0044** 0.0002 -0.0022 0.0028 -0.0051 -0.0095 -0.0084 (0.0010) (0.0021) (0.0031) (0.0046) (0.0027) (0.0059) (0.0156) (0.0152) Lower skill cells -0.0010 0.0006 0.0077 0.0448*** 0.0003 -0.0011 0.0220 0.0715*** (0.0008) (0.0012) (0.0061) (0.0074) (0.0015) (0.0036) (0.0248) (0.0229) Higher skill cells 0.0033 0.0017 -0.0044 0.0005 -0.0240 -0.0408 0.0103 0.0341 (0.0065 (0.0129) (0.0046) (0.0058) (0.0164) (0.0321) (0.0233) (0.0252) R-squared 0.0324 0.0538 0.0301 0.0510 0.0581 0.0313 0.2463 0.2582 Observations 53,921 24,717 28,496 25,920 43,807 19,304 25,603 22,411 Notes: The dependent variable in (1)-(4) is a dummy equal to 1 if the individual is employed and 0 if unemployed. The dependent variable in (5)-(8) is the log of real hourly earnings of the employed individuals. The main explanatory variables are the share of Nicaraguan workers in the labor force in the same skill-cell as the individual in the dependent variable (first row), in lower skill-cells (second row) and in higher skill-cells (third row). Additional controls include gender and marital status and fixed effects for skill (education and experience), year, region and any two-way interaction FE. Robust standard errors adjusted for clustering at the skill-region-year level are in parentheses *** ; ** ; * significant at the 1%, 5% and 10% level respectively 21