Global Journal of Human Social Science, E: Economics, Volume 23 Issue 3

appreciated activities on labor market, as domestic work (MACHADO; OLIVEIRA; WAJNMAN, 2005). Gender inequality and occupational segregation were observed in occupations typically identified as “feminine”, as domestic work, office activities, esthetic services, hotel and food services, commerce, health care, education, apparel and decoration sectors, among others (SOARES; OLIVEIRA, 2004). More recently, from 2002 to 2012, it was possible to identify a reduction of the differences between women and men, for example, regarding the tax of vacancy and also the proportion of unpaid jobs being higher among women (KUBRUSLY, 2015). III. D ata and M ethods Aiming to analyze changes in the structure of occupations on Brazilian labor market, we used the Pesquisa Nacional por Amostra de Domicílios – PNAD (Brazilian National Household Sample Survey) for the period between 1985 and 2015. The PNAD is nationally representative and annual household survey, covering both rural and urban areas, fielded by the Instituto Brasileiro de Geografia e Estatística – IBGE (Brazilian Census Bureau). Seeking to homogenize the databases and allow a comparison over the time studied, we also used hourly earnings in Real (R$) for workers aged 15 to 64 who participated on the labor force full-time, working at least 36 hours a week. The variables were kept in Real values and were calculated through a deflator based on the Índice Nacional de Preços no Consumidor – INPC (National Consumer Price Index) fielded by the IBGE, having 2015 as year base. As in Acemoglu and Autor (2011), we discarded the military sector, but the agricultural workers were kept as they are significantly representative. While in the military case significant changes are not expected over the period studied, the second group was represented a significant portion of jobs at the beginning of the period. To analyze employee qualifications, six educational levels were considered: illiterate or people with less than one year of formal education (ILL), people with incomplete elementary school (LSS), people with complete elementary school but incomplete high school (SSG), people with complete high school (HSG), people with incomplete higher education (LTC) and people with higher education or graduation (CSG). Regarding the structure of the occupations, eight occupational groups were considered: leaders (MSG), professionals and technicians (PAT), administrative employees (CRC), sales employees (SLS), production employees (PAO), agricultural workers (AGR), domestic workers (HHS) and other service providers (OTS). Besides the descriptive analyses to identify changes in the employment structure of labor market, in order to discuss income distribution, it was used some inequality measurements as Gini and Theil indexes and income proportion in tenths and cents of income distribution. In addition to, income equations were estimated to explain hourly earning determinants ( Y i ), that can be described as: = + � =1 +� =1 + Among many variables that can explain it are initially the seven occupational groups, then M = 7 , in which the reference category is domestic employees (HHS). Moreover, empiric literature suggests that there are many relevant elements to explain income inequality besides occupational groups. To Mincer (1974) in his seminal study, income is a function from human resources, especially education and the individual’s experience. Subsequently, studies have been incorporating other determinants that also explain income differentials on labor market. Besides education and experience that represent human resources it is possible to find gender as personal characteristic. Among work characteristics that can also result in different income it can be found the sector and the formalization. Lastly, there are the special characteristics, as place of residence in larger areas, urban areas or countryside and metropolitan or non- metropolitan areas. IV. J ob P olarization on B razilian L abor M arket Brazilian labor market’s composition can be observed on Table 1, in which it can be verified an increase in woman participation at occupation from 30.4% in 1985 to 41.8% in 2015. The increase in women's participation in the labor market can be explained by factors such as reduced fertility, greater availability of technology in domestic equipment and rising wages (Borjas, 2010). The increase in educational level is seen when reducing participation from less qualified individuals with incomplete elementary school and also an increase of other levels, especially for more qualified individuals with at least higher education, which increases from 4.6% to 16.0%. The improvements in access to education and changing skill requirements in the labor market were important factors for these results. Considering employment according to occupational groups, it can be noticed increase on the five starting groups and reduction on the last three. While the starting ones include occupations with more abstract tasks, the latter ones are manual. © 2023 Global Journals Volume XXIII Issue III Version I 13 Global Journal of Human Social Science - Year 2023 ( )E Occupational Polarization in Brazilian Labor Market, 1985-2015