THE DUAL MARKET:
An analysis of the presence of market gender segmentation and the impact of the motherhood in the wages determination process
Methodology
The mapping of the data aimed to verify the presence of gender segmentation in the Brazilian labor market and to analyze how motherhood affects women’s engagement and her earnings over time. For this purpose, the model was created with datas from the Pesquisa Nacional por Amostra de Domicílios (National Household Sample Survey - PNAD / IBGE) 2015 and 2014, and to achive the results were used Ordinary Least Squares (OLS) and Propensity Score Match (PSM) as econometric’s methods of estimation.
One of the arguments that supports the wage gap is the existence of a discriminatory factor for economic agents, imposing a different wage for individuals with the same marginal productivity. The duality of the labor market can be defined, in general, as a scenario in which equally qualified workers receive different wages just because they belong to a different gender. For the analysis of the existence of duality in the Brazilian labor market, the research method sought to measure the wage differences observed among workers that are not explained only by differences in their productive attributes.
Another argument is that women have more maternal responsibilities than men, bring on less involvement in the labor market. So, when women considers the optimal choice between leisure and work, there is an internal choice, associaded with this others resposabilites, who ponders more for utility in form of leisure than utility in form of work, resulting in a reduced labor supply.
The economic reasoning behind this approach is that an individual obtains satisfaction with consumption of goods $(C )$ and leisure $(L)$. In this model, leisure includes all activities carry out outside the labor market, such as activities related to househood and domestic duties. The supply of hours $(h)$ for the labor market occurs at the point of greatest individual utility among them. The total time $(T)$ is the combination of hours worked and the chosen leisure amount $T = H + L$. Thereafter, maximizing the worker’s utility depends on non-work income $(V)$, the number of hours worked $(H)$ and the wage per hours $(w)$. This process can be defined by the following equation:
The first order condition of consumption and leisure finds the marginal rate of substitution between goods. The number of hours worked and the quantity of goods consumed occurs at the point of tangency between the indifference curve and the problem’s restriction. Thus, the wage gap is not associated with a market failure, but with the different slopes of the utility curve between men and women.
For the he empirical approach of the motherhood impact, control groups were created, comparing the performance of women outside the period of maternity and mothers in the labor market. As a consequence, since they both belong to the same discrimination group and are affected in a similar way by socio-cultural factors, such as the high burden of household duties, when comparing women who had children with women who did not have children, the difference observed would be, then, the impact of motherhood on the mother’s job offer and, consequently, on her salary.
Being i equal to the mother’s group referring to the child’s age and $\mathbf{X}$ the matrix of characteristics of the individuals in the sample, which are: age, age squared, years of study, years of study squared, south, southeast, north , northeast, black, white, brown, yellow, lives with spouse, potential experience and natural income logarithm of the house, when $Y$ is equal to the amount of hours offered, or quantity of hours offered, when Y is equal to natural logarithm of income, and $\beta$ the slope parameter determined by the model.
Results
Assuming non-appearance of discrimination factors in the process of determining employers’ own wages, the remuneration of the production would be genuinely dependent on the productive characteristics and the preferences of individuals. However, when comparing the behavior of normal wages with the entrepreneur class, the appearance of the gender wage gap was noted.
Source: Own Elaboration based in Data from PNAD 2014 and 2015.
As a result, the argument of structural market segmentation or differentiation on the part of employers has weakened, pointing out that there are others dominant factors determining wages in contradistinction to the theory of discrimination by the contracting firm. The method adopted showed the magnitude of the wage differences between workers and employers that are not explained only by differences in their observable productive attributes. The purpose of this methodology was to investigate structural wage inequality, between genders and as well as between women with and without children.
The distinction between factors attached to market segmentation and factors attached to individual preferences is essential to determine what is the best strategy to reduce the wage gap between men and women. Based on the results obtained, the persistence of the gender pay gap in the entrepreneur class would then be due to other factors, such as the intrinsic factors boarded in the neoclassical theory, determined by the process of maximizing the utility of individuals. Thefore, the presence of a dual market caused by a discriminatory agent when determining wages is weakened and does not appear to be able to explain the causes of the gender pay gap.
The results found show that the effect of motherhood reduces the amount of hours offered by the mother, in the short and medium run. Despite the low magnitude of the reduction in the work hours, whether in the short or medium run, motherhood proved to be a relevant factor in the final determination of women’s wages. It is also observed that in the long run, the difference in the average treatment effect in the treaties is not significant at 5 \%. This may mean that mothers, in the long run, rebuild themselves in the labor market.
Source: Own Elaboration based in Data from PNAD 2014 and 2015.
The results indicate the non-parity of wages between groups in the short, medium and long run. The wage gap between women who had children against women who did not have children stands out and grows in the long run. In other words, motherhood has long-lasting effects that modify the woman’s salary not only in the period when the mother chooses to offer less hours of work, but also, later, when she returns to the same number of hours offered in the labor market. This results is either observed in the sample complete and in the subsample of wome employers.
However, the question of causality remains. Would mothers who offer fewer hours of work during the week be less engaged in the labor market because of their children, or would they naturally be more willing to dedicate themselves less to work because of their ambitions for the family? Due to the results measured by the PSM technique, the mother starts from an initially higher salary, in the subsample of women employers, and over the years she suffers a salary reduction due to the effects of motherhood. In the long run, the effects of this reduction are around $-10 \%$ of income in the complete sample and $-28 \%$ in the employers subsample.
This relationship brighten the argument of the natural wage gap between men and women and guides the public policies to contain wage inequalities between genders. Motherhood proved to be a more relevant factor in determining engagement and wages than the presence of a discriminating agent. These results may indicate that policies to support mothers, such as day-care centers and a more legal equal paternity leave, may mitigate the negative effect of the gender pay gap.