Impacts of Forced Immigration: The Venezuelan Diaspora and the Brazilian Labor Market

ABSTRACT

This article estimates the impact of immigrant influxes on local labor markets, taking advantage of a natural experiment generated by a massive wave of forced migration of Venezuelans to Roraima, a state in northern Brazil, from 2016 onward. Estimates show that a 1% increase in the share of migrants in the state population due to immigration resulted in 1.1% higher earnings for native workers in the formal sector, 0.3% fewer hours worked weekly, and 2.3% higher wages. Results are confirmed by employing a synthetic control method.

KEYWORDS:

• Brazil
• migration
• native workers
• natural experiment
• synthetic control method
• Venezuela
• wages

Acknowledgments

The author thanks Joan Hamory for outstanding research advice and Daniel Hicks, Gregory Burge, Chunbei Wang, and Fábio Costa Morais de Sá e Silva for helpful suggestions. I am grateful to Renata Alfena Zago for help with data collection and to other key informants for their input on the data and background for this project. Any errors are my own.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/08853908.2024.2350711

Notes

¹ For more information about studies of the impacts of immigration on labor markets, see Altonji and Card (1991); Card (2001 2009); Jaeger, Ruist, and Stuhler (2018); Borjas (1999).

² This set of studies claims that these phenomena can be classified as natural experiments, arguing that migrants were forced to leave their locations. Forced migration “is less of a choice and less voluntary than economic migration […] Forced migrants tend to move to destinations based on proximity and security criteria rather than personal networks, although networks can occasionally play a role” (Verme and Schuettler 2021).

³ The difference comes from the fact that the Mariel supply shock disproportionately affected low-skilled occupations since it increased the number of high school dropouts in Miami by 20% (Borjas 2017).

⁴ The authors review 59 empirical studies derived from 19 major forced displacements between 1922 and 2018, and 972 outcomes resulted from them. Results indicate that impacts on employment and wages are often non-significant. Those which show some level of significance often point out that decreases in employment and wages are more likely to occur in the short run in middle-income countries with female, young, and informal workers being more affected. Nevertheless, food and rent prices are positively affected by the increase in supply shocks in the short run.

⁵ Caruso et al. (2021) find that a 1% increase in immigration from Venezuela reduces the wages of the informal sector in the country by 10% points in urban areas. However, recent studies do not find any effect on the labor market outcomes in Colombia.

⁶ These studies find that young, low-educated Ecuadorian workers in high-inflow regions experienced increases in informality (5%) and a 13% decrease in earnings compared to native workers with similar characteristics located in areas with low or non-influxes of Venezuelans, and that Venezuelans are experiencing significant occupational downgrading relative to their employment prior to emigration.

⁷ This is likely a lower bound on the actual number of people who migrated to northern Brazil. According to the country’s Ministry of Interior, at least 199,365 Venezuelans passed through the area in 2017 and 2018, but not all remained in northern Brazil, and others entered the country without being detected.

⁸ The literature has shown that immigrants tend to earn lower wages upon arrival than locals (Huang and Anderson 2019).

⁹ Almost 1.3 million Venezuelan migrants were registered in Colombia in June 2019, as well as 768,000 in Peru, 288,000 in Chile, 263,000 in Ecuador, and 130,000 in Argentina (IOM 2019).

¹⁰ In that year, the petrol barrel reached its lowest price in more than a decade, at $34.7. See Figure A2 in the online Appendix.

¹¹ According to the 2010 census, Roraima is 13th among 27 states plus the Federal District in the Human Development Index ranking in Brazil. With a population of over half a million people in 2018, it averaged a nominal household income per capita of approximately $310.71 monthly in 2017 values (IBGE 2020).

¹² Although Amazonas also borders Venezuela, the Amazon forest makes crossing this part of the border an almost impossible task. The only road (Troncal 10 in Venezuela and BR-174 in Brazil) connecting both countries is the one that goes from Santa Elena de Uairén (Venezuela) to Pacaraima (Roraima, Brazil).

¹³ Pacaraima had a population of 15,580 people in 2018, the average monthly earnings in the city reached $411.07 (about 1.7 the minimum salary in Brazil), and 46.5% of the population earned up to half the minimum salary ($120.90 monthly) in 2017 values.

¹⁴ Although another Venezuelan state also borders the west of Roraima, the Amazon forest makes crossing this part of the border an almost impossible task. See Footnote 18 for more information.

¹⁵ Fewer than 1% reported no schooling.

¹⁶ The distribution of schooling levels among the unemployed is similar to the full sample.

¹⁷ About 7.4 million establishments that filled RAIS in 2009.

¹⁸ “Admission” means every entry of a worker into the establishment during the year, whatever its origin, and by “dismissal” during that year, every departure of a person whose employment relationship with the establishment ended during the year for any reason (dismissal, retirement, and death), either on the initiative of the employer or the employee. Inflows and outflows through transfers are included, respectively, in admissions and dismissals.

¹⁹ This data alleviates concerns with endogeneity coming from previous findings in the literature that migrants tend to settle in places with a higher share of compatriots.

²⁰ “Parda” represents mixed-race, a descent from Black and White, Black and Indigenous, or White and Indigenous parents.

²¹ This variable is constructed by multiplying the number of weekly hours worked by 4 and dividing monthly earnings by it.

²² The variable is constructed as follows: $\frac{\mathrm{number}\mathrm{of}\mathrm{migrants}\mathrm{in}\mathrm{Roraima}\mathrm{int}}{\mathrm{population}\mathrm{in}\mathrm{Roraima}\mathrm{int}}\mathrm{x}100$.

²³ Except when it is the dependent variable.

²⁴ Because the data is composed by workers that are not necessarily followed over time and due to the individual characteristics used in the set of control variables, individual fixed effects are not included in the model.

²⁵ It should also be noted that the Brazilian government, together with the army, started an operation named “Operação Acolhida” - Operation Welcome – as an answer to the massive migration of Venezuelans in the country. This program included the provision of financial aid to Roraima, as well as resettlement plans for some immigrants to other states in Brazil. In addition, this operation did not randomly choose those who would be relocated, nor their destination. It only did so with those who wanted to and to the states/cities they chose to go to. This would also create a nonrandomness factor when dealing with the effect of migration in the entire country. However, the number of people taken to other states reached only 5,482 (OIM 2019) (see Figure A3 in the online Appendix), which would likely lead to a much more diffused impact on other states.

²⁶ For more information on the SCM, see Abadie and Gardeazabal (2003) and Abadie, Diamond, and Hainmueller (2015).

²⁷ It still could be the case that Venezuelans who are leaving Roraima are not in the formal sector. However, given the financial situation of those who entered the country through Roraima, it is hard to expect that a high number of migrants would leave the state of entry to other states.

²⁸ Although, it is still possible that Venezuelans are departing Roraima to other states through land.

²⁹ These are the official numbers of people who requested refugee status, rather than the total number who entered the country. As already highlighted, it is hard to measure with precision the exact number of people who crossed the border and stayed in the country or in Roraima. Because of that, these results are likely a lower bound of the true effect.

³⁰ This calculation considers the average of R$1,602.50 for real earnings, as shown in Table 1.

³¹ This calculation considers the average wage of R$12.00 shown in Table 1 for Brazilians.

³² Until 2019, it is estimated that government spending increased by about $100 million in 2017 values, which corresponds to 11.4% of the state expenditure in 2018. Most of it is being spent on consumer goods (FGV DAPP 2020).

³³ From 2018 to 2019, Venezuela increased its participation in Roraima’s exports to 46.7%. The most exported products to the country were manufactured goods and rice (in grains) (FGV DAPP 2020).

³⁴ Notice, however, that, although there is an increase in the unemployment rate that is higher in proportion if compared to the increase in the labor force, this data does not allow us to distinguish nationals from migrants. Thus, it is hard to say whether it indicates a worsening of job market conditions for nationals or only an increase in the number of workers in the workforce.

³⁵ Which can be derived from factors such as xenophobia. In fact, a lot of movements of natives against the arrival of Venezuelans in the state were reported in Roraima (Costa and Brandão 2018; Marques 2018; Pontes 2018; Ramalho 2019).

³⁶ This result could be driven specifically by the military, expressing the effect of a higher presence of the army in the state. Because this operation is voluntary, it could be the case that most of the military personnel that relocated to Roraima have lower military ranks and, therefore, earn less. Notice that there are no effects, neither in hours worked nor in probability of termination, for this category, and the magnitude of the decrease in wages and earnings is approximately the same. Besides, expenditures in public safety decreased by 7.6% in the state in 2018 (FGV DAPP 2020).

³⁷ Although exports of rice increased by 126% in comparison of 2016 to 2017 and 2018 to 2019, and planted area increased by 28.9% from 2017 to 2018 (FGV DAPP 2020), results show that workers in agriculture and fishing are earning less, specifically due to lower wages, which may be a result of the higher competition for jobs.

³⁸ The sector corresponds to 15% of the state GDP and is led especially by construction.

³⁹ Data shows an increase in the number of kids and teenagers enrolled in school in the state as well as the number of adults enrolled in professional education courses (FGV DAPP 2020).

⁴⁰ See Table 2.

⁴¹ See Figure A6 in the online Appendix for the weights used in the pool of donor states.