O professor Archibald Haller, que formou gerações de sociólogos brasileiros na Universidade de Wisconsin, Madison, nos envia a seguinte contribuição:
Research using PNAD data of the last quarter of the 20th Century has yielded several results that may be of some importance in the current debates about education and about race.
These analyses were carried out by teams at the University of Wisconsin-Madison USA, and included (among others) researchers now at several other universities: USP, UFMG, the Australian National University, and the University of Iowa. Different specific studies used one or another of the labor force PNADs. But they share certain methodological characteristics. Each uses tightly controlled statistical procedures resulting in conclusions which, though surprizing, would be hard to challenge. All but one set of the findings summarized below has been published.
1. In 1992 the late Helcio Saraiva and I used 1973 and 1982 data. For working men the average income increment per each additional year of education was about 9% in 1973 and 7% in 1982. For working women the corresponding rates were about 8% and 7%. So the greater the number of additional years of education one gains, the greater the already substantial effect of one’s education on one’s income.
This analysis also checked the income increment for each individual year of additional education for those starting at zero years and getting one year, those starting with one year and getting a second year and so on up to those who at 14 years would add a 15th. This excercize showed that — as many have predicted — credentials count: the income increment is higher if one starts at zero and finishes year 1 than if one starts at 1 year and then finishes year 2. Finishing each of the standard termination years (yr.4, yr.8, etc) is more fruitful than stopping in between them. However: contrary to the skeptics, adding a year over what one already has raises income EVEN if that year is between normal termination years. This seems to mean that the LEARNING one gets from education pays off (again, contrary to the skeptics).
2. In 2005 Jorge Alexandre Neves (UFMG) published an article along the lines of the above, but using only the nation’s rural farm personnel. He used the PNADs of 1973,1982,1nd 1988. Contrary to previous research (badly done) and common belief, the income increment to each additional year of education was around 9% in 1973 and 1982, and 5% in 1988.
3. In 2001 Jonathan Kelley (ANU) and I published an analysis of the effect industrial development on income, comparing the less developed Northeast with the more developed South (our definition of these regions [see Haller 1982: Geographical Review]). All classes of workers gain, and in about equal proportions. 40% of the gains were due to education, 10% to occupational upgrading, and 50% to better paying jobs.
In a still unpublished paper (Kelley, myself, and W. Haller [Clemson University]), we retested the same hypothesis by comparing Brazil as a whole (except Amazonia) as industrialization proceeded between 1973 and 1988. Results: Practically identical to those of the 2001 paper. The pay of each occupational group, lowest to highest, grew at about in about the same proportion as every other group: 1% to 2% per year. About 40% of the income growth came from better pay per job, about 10% from occupational upgrading, and about 50% from educational growth.
4. In 2005 Danielle C. Fernandes (UFMG) published an analysis of race, socioeconomic development and education, using age cohorts from the 1988 PNAD. She concluded 1. ‘that the transformations brought about by industrialization have not decreased the effect of the socioeconomic determinants of educational stratification in Brazil’. 2. that race shows its strongest effects at both the lower and the higher levels of educational attainment, and its weakest in the middle. 3. At least as importan, the ‘transformations brought about by industrialization have lessened neither the effects of socioeconomic origins nor of race. Indeed there is compelling evidence that the negative effects of being Black or Mullato have increased’.
Haller and Saraiva (1992). The income effects of education in a developing country: Brazil–1973 and 1982. RESEARCH IN SOCIAL STRATIFICATION AND MOBILTIY 11: 295-336.
Neves (2005). Labor force classes and the earnings determination of the farm poulation in Brazil: 1073, 1982, and 1988. THE SHAPE OF SOCIAL INEQUALITY: STRATIFICATION AND ETHNICITY IN COMPARATIVE PERSPECTIVE. RESEARCH IN SOCIAL STRATIFICATION AND MOBILITY 22: 424-475. (Oxford: Elsevier.)
Kelley and Haller (2001). Working class wages during early industrialization: Brazilian evidence. RESEARCH IN SOCIAL STRATIFICATION AND MOBILITY 18: 119-161.
Fernandes (2005) Race, socioeconomic development and the educational stratification process in Brazil. THE SHAPE OF SOCIAL STRATIFICATION AND ETHNICITY IN COMPARATIVE PERSPECTIVE. RESEARCH IN SOCIAL STRATIFICATION AND MOBILITY 22: 365-422. (Oxford: Elsevier.)