Persistent effects of immigration

People with non-Iberian surnames earn higher salaries and achieve higher educational levels than other Brazilians

Guilherme Gaensly (1843-1928)/Energy Heritage Foundation of São Paulo/Memorial to the Immigrant Europeans newly arrived in São Paulo, assembled in the central patio of the Immigrants Welcome Center, around 1890Guilherme Gaensly (1843-1928)/Energy Heritage Foundation of São Paulo/Memorial to the Immigrant

The son of a Bolivian who was an adolescent when he arrived in Rio de Janeiro more than 60 years ago and became a physician, economist Leonardo Monasterio is investigating whether current descendants of European and Japanese immigrants who arrived here between the final decades of the 19th century and the first decades of the 20th enjoy, even today, economic and educational advantages associated with their ancestry. According to a study he published on May 5, 2017 in the scientific journal PLOS ONE, Brazilian workers who have been contracted formally and have at least one Japanese, Italian, German, or Eastern European surname earn significantly more and exhibit a slightly higher educational level than those who bear Iberian family names that originated in Portugal or Spain. The disparity is even greater when the comparison includes the earnings and number of years of formal education achieved by blacks, browns, and descendants of Brazilian Indians—groups that represent about 55% of Brazil’s population.

The average earnings and years of schooling of individuals who bear Japanese surnames were the highest in the sample, which employed surname analysis to determine the heritage of formally-employed workers in Brazil. According to the article, workers of Japanese origin earned an average of R$73 an hour, more than twice the sum for workers with Iberian surnames and almost three times as much as blacks and browns. Descendants of Asians attend school for an average of 13.6 years, about three years longer than blacks and browns. Ranking next on both the financial and educational questions are descendants of Italians, Germans, and Eastern Europeans, always ranked in that order, (see chart). “We cannot yet say what is the cause of the inequality. When they arrived, the immigrants were more literate than the Brazilians, and some of them received subsidies,” explains Monasterio. He is affiliated with the Institute for Applied Economic Research (IPEA) and a professor of graduate courses in economics at the Catholic University of Brasília (UCB). “But historical and contemporaneous discrimination or even cultural differences could also explain the salary bonus granted to non-Iberians.”

To conduct the study, the economist had access to the 2013 edition of the Annual Social Information Report (RAIS), the result of a set of socioeconomic data that employers supply to the Ministry of Labor. That year, the RAIS contained information about 46.8 million Brazilians age 23-60 who were working at least 40 hours a week. The earnings of the many wealthy persons who do not live on their salaries, and the poorest, whose jobs are not registered in a carteira assinada, i.e., who are “off the books,” do not figure in the ministry’s survey. “The RAIS gives us a good idea of the level of wages and salaries in the middle strata of the Brazilian population,” Monasterio explains.

Aided by automated methods that use algorithms to identify the origins of surnames, the researcher found slightly more than a half million different family names in that enormous data base. However, the five most common family names (Silva, Santos, Oliveira, Souza and Pereira) were used as either second or last surname by 14 million workers, almost one-third of those recorded in the RAIS. (see chart on surname frequency). When an individual has more than one surname, only the last one was considered in the analyses of schooling and financial earnings.

The finding that current descendants of Japanese, Italian, German and Eastern European immigrants earn more than whites of Iberian origin and much more than blacks, browns, and indigenous peoples was not unexpected. It reflects a socioeconomic inequality that has persisted for decades in Brazil. Between the end of the 19th century and beginning of the 20th, the recruitment of European immigrants as well as people from Japan was part of a Brazilian government policy that sought to replace the former black and slave laborers, or those recently freed, with white workers.

“Immigrants, followed by their descendants, were the first to gain access to education and formal employment in Brazil,” observes sociologist Rosana Baeninger from the Population Studies Center (NEPO) at the University of Campinas (Unicamp), a scholar of the immigration process as it pertains to São Paulo. “They also had skills that were important for the development of cities and they conveyed the idea that they took work seriously.” In short, history shows that the best opportunities were offered to the segments of the Brazilian population who were of European (or Japanese) origin, while descendants of freed slaves and indigenous peoples were relegated to second place.

Iberian, but only somewhat
Determining the ancestry of an individual would seem to be a simple task provided one has access to his surname. But there are obstacles and limitations to be overcome. In the case of the work carried out by the IPEA economist, the first problem was quantitative. It is not feasible to classify a half million different surnames manually, by referring only to historical sources. There are, however, software programs that do this automatically. That is how Monasterio was able to distribute the family names into five categories: Japanese, Italian, German, Eastern European, and Iberian; the latter responsible for the bulk of the migratory flow of 3.2 million foreigners who entered Brazil at the turn of the 20th century.

In the first four cases, the association is almost always direct and immediate. Someone with a Japanese surname descends from Japanese who probably arrived in Brazil about 100 years ago, when the flow of immigrants peaked. This strategy works in most of the cases, although it is not perfect. Strictly speaking, a Teutonic surname would indicate a descendant of Germans, but can also apply to Austrians or German-speaking Swiss. Surnames of peoples whose immigration to Brazil was less significant (specifically, the case of Arabs) end up being classified imperfectly.

The biggest challenge encountered during the study was to associate Iberian surnames with the flow of migrations from Portugal or Spain. The algorithms that Monasterio used did not differentiate the surnames of Portuguese origin from those originating in Spain because the spelling is so similar. The answer was to adopt the “umbrella” solution of an Iberian surname. But the problems did not end there. Besides encompassing whites who did indeed descend from Portuguese and Spanish immigrates, this classification also included blacks, browns, and indigenous peoples whose families were probably required to adopt an Iberian surname in the past. Therefore, bearing a typically Luso surname like Silva or Oliveira in Brazil does not necessarily imply that one is a son, grandson, or great-grandson of Portuguese forebears. In order to separate the bearers of Iberian surnames who are white, descending from Europeans, from those who are black and brown and whose African ancestors were brought here by force and enslaved in Brazil, the economist had to insert the criterion of color/race as recorded in the RAIS. “In those cases, in order to determine ancestry, we had to create a hybrid index that took into account both the surname and the self-declared color,” the economist explains.

The influence of surnames

The 46.8 million workers recorded in the RAIS bear 531,009 distinct surnames. In that contingent, equivalent to almost one quarter of the Brazilian population, the immense majority bear family names from the Iberian Peninsula: the second or final surname of 88.1% of those registered is of Portuguese or Spanish origin. There are, for example, six million Silvas, 3.5 million “Santos” and 1.9 million Oliveiras. Next in the ranking are the Italian surnames (7.2% of employees), followed by German (3.2%), Eastern Europeans (0.8%) and Japanese (0.7%). Since it has been more than a century since Brazil received influxes of immigrants, the current presence of surnames from other countries is quite modest. This is not a universal pattern. In Spain, the list of the 500 currently most common surnames includes one Indian (Singh), and one Chinese surname (Chen).

The greatest concentration of non-Iberian surnames, according to the Leonardo Monasterio study, occurs in the geographical zone that starts in north central Rio Grande do Sul, passes through Santa Catarina, and ends in south central Paraná. Interestingly, the presence in a region or country of a large contingent of workers who have Italian, German, East European and Japanese surnames seems to have favorable effects on all employees in that place, according to the preliminary findings of work being done by Monasterio and his colleague Philipp Ehrl, a UCB economist. “A 10% increase in the percentage of Brazilian workers of non-Iberian ancestry, as estimated based on surname, causes a 2.2% increase in the earnings of all of them,” Monasterio estimates. That effect is believed to be more noticeable in cities that are ethnically more diverse and function as magnets for workers who possess new skills. The two economists are now testing their theory by working with RAIS data about Rio Grande do Sul.

Scientific article
MONASTERIO, L. Surnames and ancestry in Brazil. PLOS ONE. May 8, 2017