The data set used in my research was collected by the National Bureau of Statistics (SSB) for the years 1998-2003 and constitutes a representative sample of Norwegian households. Income data is obtained from different administrative registers and other statistical data sources as of 31th December of the fiscal year. This data set gave me an opportunity to study variations in total annual work related income for both genders, a topic there has been written few articles about in contrast to the amount of research that exists on hourly wages gender gap and gender differences in labour supply.
Variations in total income may comprise effects of both variations in hours worked and variations in the hourly wage itself, in addition to effects of such factors as education and work experience. Due to the nature of the data material available to me, the effects of changes in hourly wages and hours worked could not be singled out. Thus I was not able to investigate how much of the income gap stems from the fact that women work fewer hours (part-time) than men and how much of it can be explained by women on average earning lower hourly wages. Neither could I investigate to which extent variations in wages or hours worked can be explained by factors such as education, age or number of children. On the other hand, this data gave me the chance to examine the total income gender gap, the total income impact of children on women in relation to the impact of potential work experience and the percentage of the overall gender income gap caused by women taking one year of maternity leave per child.
It is commonly found in research on the subject of economic gender inequality that children have a negative impact on womens hourly wages. I show that this is also true for total income from work for the years and age group examined in my research. There are two possible reasons why children have a negative impact on women's average income while they have a positive impact on men's average income: either women who have children experience gender discrimination (in this case there exists an income gap between women and men with identical profiles and same number of children and a smaller or no income gap between women and men with identical profiles and no children), or children affect women's income-related choices in a different way than they affect men's (in which case the gender income gap that exists between women and men with the same numberchildren can be explained by differences in variables that affect income directly). As the former has in earlier research been shown to be practically non-existant in Norway, it is assumed that the latter best describes the dynamics that lead to gender differences in returns to children.
In my research I have estimated, by use of a multivariate linear regression model, the impact of the fact that women (not men) take one year of maternity leave per child. In estimating coefficients in my models I took use of SPSS statistics software.My results signify that the fact that women are away from the work force on average for one year for each child they have, carries on average a penalty of 10 thousand NOKs for the specific age group examined (21-40 years). This constitutes around 10 % of the income gap between men's and women's average incomes for the years examined. Considering these results it is reasonable to believe that a more equal division of the parental leave between the parents would decrease the income gender gap ceteris paribus by 10 %.Other results include the finding that women lose 4 - 8 % of average income due to the effect of each child, depending on the year in question. There is a clear decreasing trend, with the effect in 2002 half of the effect in 1999.
Compared to the potential work experience premium, the effect of each child is by 2002 for women about the double (in absolute value) of the effect of each year of potential work experience. This indicates that women on average lose more income from having each child than they gain from each year of potential work experience. This is a logical result, as many women (in contrast to men) adjust their working time also after the one year of maternity leave has passed, as discussed in the theory overview of this paper.
The issue of the gender income gap is of current interest in contemporary Norway, both through existing political will to decrease gender inequality and the increasing demand for scarce labour in a heated economy where many women work part-time. Since, as discussed in my paper, women's labour supply is affected by children, more female participation in the labour market can be attained through political initiatives such as full kinder garden coverage. This will increase the number of hours on average worked by women, and decrease the income gender gap through higher incomes from work for women, both because of anincrease in hours worked and due to women acuiring more work experience than before. Concerning the current political initiative of a more equal division of parental leave between the parents, my research indicates that given that this initiative suceeds, it will decrease ceteris paribus the gender income gap by 10 %.