Excerpt
Contents
1
Introduction
3
2
Literature
4
3
Data
6
4
Models
10
5
Results
12
5.1
Model 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12
5.2
Model 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
5.3
Model 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
6
Conclusion
14
7
Appendix
15
2
1
Introduction
My intention was to investigate the income structure of full-time employees
in Austria based on the returns to schooling. I also focused on the so-called
"gender gap", i.e. whether there are differences in the income of male and
female workers. Due to historical reasons, I also controlled for the so-called
"east-west divide in Austria" by looking at the effect of the individual Aus-
trian federal states. For this study, the Mincer Equation comes in handy as
a tool of analysis. The Mincer Equation is an empirical model which was
introduced by Jacob Mincer in 1974, originally called "The Human Capital
Earnings Function". Mincer's work had a vast and lasting influence on the
field of empirical labour economics. It was the first and most popular model
of investment in human capital examining changes in an individuals income
by considering its education and its experience (Lemieux, 2003).
For this project, I test slightly different forms of the original Mincer Equa-
tion. First of all, I want to see how well the original Mincer equation describes
the data 35 years after it was introduced. Within the explanatory variable
Education I also want to look at the difference between the different types
of High School Diplomas. Austria is one of the last countries that still of-
fers Vocational High Schools that last one year longer (13 years of schooling
total) in comparison to "General High Schools" that last 4 years (12 years
total). The rest of the models include a sex-variable to control for gender
specific income differences. In the third model I included the federal state to
examine the east-west divide of incomes in Austria.
In the second section I will give a short review of the findings about the
two factors that I will add in my models. In the following, I introduce the
data and its application including some descriptive analysis to give a short
overview on the dataset and possible relations. The fourth section contains
the three different models followed by the results of my estimations. In each
specification I estimated the models with OLS, robust standard-errors, and
GLS and as I find heteroscedasticity in my OLS specification. I also assume
that endogeneity could be a problem. Therefore, it would have been better
to use panel data to cut out time invariant effects like "abilities" by using
first differences or fixed effects. However, Statistik Austria was not able to
send me the data of other years before the deadline of this project. Thus,
this remains for further research. All tables can be found in the appendix.
3
2
Literature
With a EU gender gap indicator of 25.5% in 2008, the gender gap in Aus-
tria is very large compared to other developed countries (Commission, 2010).
Based on gross hourly wages, the EU gender pay gap indicator shows the av-
erage difference between men's and women's earnings. Within the European
Union, only the Czech Republic shows a bigger gap than Austria. However,
the indicator does not account for differences in labour market experience,
education, or other productivity related variables. Boeheim et al. (2011)
therefore used new matched employer-employee data from 2007 to account
for these differences and they also considered different measures of experi-
ence and work interruptions. They look at the gap decomposing it into the
parts which are attributed to different characteristics and the corresponding
returns to these characteristics. They find that Austrian women have, on av-
erage, less workplace experience but that they are better formally educated.
Still, the authors obtain that about 50% of the gender wage gap cannot be
attributed to observable characteristics. However, the gender gap was found
to be much smaller in the public sectors, which is still dominated by trans-
parent pay scales whereas in top-paying jobs in the private sector, individual
bargaining is the norm. They assume that part of this difference could be
caused by unobserved characteristics, e.g., attitude and commitment. How-
ever, their conclusion is that it is plausible that (part of) this difference is
caused by discrimination against women.
Maria A. Davia (2009) focuses on the returns of Education based on the
EU-SILC of 2005 which is the only year of the survey that includes gross
monthly wages. They find that Austria is one of the two only countries
where women get a lower return from education than men on every level of
schooling and in every region of the country. In their OLS Robust regres-
sions she even finds some of the education effects to be negative and/or not
significant for women, while returns for men turn out to be significant on
every level and in every model specification. The World Economic Forum
(2014) ranks Austria as number 34 regarding gender specific wage differences
which is far behind the countries that Austria is normally compared to, e.g.
Finland, Norway, Sweden (rank 2-4) or our immediate neighbour and closest
partner Germany (rank 12). In addition to that, Gruenberger and Zulehne
(2009) find that gender specific differences have not changed since the end
of the 1990s in Austria which is also atypical for developed countries. Be-
fore that, however, "Gleichbehandlungsgesetze" (laws that obligate firms to
treat men and women equally) improved the possibilities for women which
reduced the gender gap. The authors also point out the interesting fact that
married men earn more than single men which is not the case for the female
4
workforce. They assume that this could also account for some part of the
gender specific difference in wages. In addition to that, they find women to
be less "aggressive" in their wage bargaining and that especially high wage
bargaining happens on an individual basis (in contrast to low wage jobs,
where collective bargaining contracts exist).
Another Austrian "specialty" are the differences between East and West
Austria, most probably caused by our history and/or the geographic sur-
rounding. If we exclude the capital Vienna, it is known that Austria's GDP
comes mostly from western regions where the GDP per capita is strictly
higher than 30.000 EUR whereas regions like Burgenland could only reach
a GDP per capita of up to 20.000 EUR in 2007. Data also shows a higher
mortality, more health issues and lower quality of education in the east of
Austria (Austria, 2010). The PISA test of the OECD also states a strong
decrease in quality of education from East to West (Presse, 13.12.2010). This
effect could be "hidden" in an east-west divide as the Mincer Equation only
measures the years but not the quality of education.
5
3
Data
For this project I used the EU-SILC data (European Statistics on Income
and Living Conditions) of 2007 to examine the income differences between
different groups (men-women and federal states). The dataset is a free micro-
dataset for external users provided by Statistics-Austria. It contains 3080
observations usable for my purpose, with more than 400 variables. In the
following I will give a quick overview over our dependent variables.
Education as a categorical variable with ten categories:
Experience in years:
Sex as a Dummy Variable:
6
Excerpt out of 19 pages
- Quote paper
- Katharina Bergant (Author), 2014, The Mincer Equation. The Gender Gap in Austria, Munich, GRIN Verlag, https://www.grin.com/document/299953
Similar texts
Publish now - it's free
✕
Excerpt from
19
pages
Comments