Gender wage gap in the US in 2007 - 2016
Name: Mariia Sukhova
University: University of Amsterdam
Faculty: Faculty of Economics and Business Date: 26/06/2018
Supervisor: Melvin Vooren Bachelor: Economics and Finance Number of EC: 12
Statement of Originality
This document is written by Mariia Sukhova who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
In this paper the trends in gender wage gap dynamics in the US in 2007 – 2016 are analyzed based on the data from the Survey of Consumer Finances conducted by US Federal Reserve. According to Oaxaca and Blinder (1973) decomposition method, the gender wage gap has reduced by 3% during the last decade, from 23% in 2007 to 20% in 2016, but the gap is still substantial. Only about 2/3 of the gap can be explained by the differences in human capital between men and women, so there is still a significant part that cannot be explained by differences in the observed characteristics of individuals and may be attributed to discrimination. The increase in the level of education among women has had the most significant impact on narrowing the gender wage gap.
Content:
1. Introduction and research question 2. Literature review
2.1. Previous studies on gender wage gap 2.2. Determinants of gender wage gap
a. Education and its impact on wage gap b. Age and its impact on wage gap
c. Marital status and its impact on wage gap d. Children and their impact on wage gap 3. Methodology and Empirical analysis
3.1. Data and sample 3.2. The empirical model 3.3. Results
4. Conclusion 5. Reference list 6. Appendix
1. Introduction
It is a well-documented fact that women are still in a disadvantaged position relative to men: they participate less, they are subject to a higher degree of unemployment and labor informality and still earn less than men in many segments of the labor market. The wage gap may be explained in part by the differences in human capital between genders, but the unobserved factors or “wage discrimination” contribute to its existence as well - women are paid less than men just because they are women
(International Labor Organization, 2015). This serious issue many nations around the world are concerned about. In some countries, including the United States, there is an Equal Pay Day: this date symbolizes how many days in the new year women need to work extra to get the same amount as men earned last year (Berger, 1970).
The United Nations 2030 Agenda for Sustainable Development identified "decent work for all males and females, reduction in inequality and eliminating the discrimination as the key objectives of a new universal policy agenda" (Razavi, 2016). The problem of reducing gender pay gap is central to this agenda. Sustainable Development Goal (SDG) 8 calls for "promoting sustained, inclusive and sustainable economic growth, full and productive employment and decent work for all, indicates the importance of ensuring equal pay for work of equal value" (International Labor Organization, 2017). At a summit in Turkey the G20 also discussed the need to eliminate the inequalities, with the
understanding that they violate political and social cohesion and slowing down economic growth, worsening the welfare of the country (2015).
This paper investigates the change in the gender wage gap in the United States of in 2007 - 2016. The structure of the paper is as follows. A review of relevant literature is presented in section 2.
Several existing studies on wage gap from both the US and outside the US are discussed, as well as the possible determinants of wage gap. Section 3 introduces the dataset, the variables used in this paper and the research methodology. The results of the analysis are presented, followed by the discussion of the limitations of this paper, suggestions for further research and a conclusion.
2. Literature review
2.1 Previous studies on gender wage gap
The focus of this paper is the dynamics of the gender wage gap over time. Several existing works have identified the changes in the wage gap over the years. However, the results of analyzes for different countries were contradictory. Suh (2010) in his study investigated the changes in the gender wage gap between 1989 and 2005 in the U.S. He used Blinder - Oaxaca decomposition model by dividing the wage gap into two parts, a part that is “explained” by wage determinants, differences in personal human characteristics such as marital status or age and a residual part that cannot be explained by such differences in wage determinants. According to him, the gender pay gap decreased significantly during the period studied, from 74.0% of men's income to about 80%. Reduction of the gap was confirmed by the results found by M. Duraisamy and Duraisamy (2016), who examined gender wage gap in India in 1983 - 2012 for six segments of the labour market. They used the quantile regression method in order to identify the gap. The paper includes the estimation of 260 gender-specific quantile regressions. The authors conclude that during the last two decades the gap between wages of men and women was reduced by 20%, and the female to male wage ratio has increased significantly, from 0.49 to 0.69. On the other hand, the analysis of the gender pay gap in Italy from the mid-1990s to the mid-2000s, carried out by Mussida and Picchio in 2013, showed an increase in the wage gap over the decade analyzed. At the first stage of their study, they used a specific model for flexible estimates of wage distribution in the presence of covariates in the sample selection (developed in 2011). At the second stage, the authors estimated the wage structures and distributions of the individual characteristics of
females and males at the beginning and end of the analyzed period to understand the changes in various components of the gender pay gap. As a result of their analysis, they came to the conclusion that the average wage gap rose from 6 to 15% in the mid-1990s and from 5 to 18% in the mid-2000s. Therefore, assuming men and women had the same distribution of individual characteristics during the analyzed period, the wage inequality has increased from 15 to 18%. The research conducted by
Albrecht, Bjorklund, and Vroman (2003) using Blinder - Oaxaca decomposition model concluded that the gender pay gap has widened from 1976 to 1993, thus confirming the existence of the glass ceiling effect in Sweden. The term glass ceiling is used to describe the set of invisible obstacles that prevents women and other minorities from significantly climbing the career ladder and achieving high
positions, regardless of their qualifications and abilities (Cotter, Hermsen, Ovadia, & Vanneman, 2001).
I expect that there was a decrease in the wage gap in the United States in the analysed period of 2007 - 2016. This expectation is based on the increase in the level of education among women (DiPrete & Buchmann, 2013) and by the adoption of laws aimed at eliminating wage discrimination. The Lilly Ledbetter Fair Pay Act, signed by President Obama on January 29, 2009, restored the protection against pay discrimination that was stripped away by the Supreme Court’s decision in Ledbetter v. Goodyear Tire & Rubber Co. The Lilly Ledbetter Fair Pay Act helps women who face wage discrimination on the basis of race, sex or religion (Langdon & Klomegah, 2013).
H1: The gender wage gap has decreased
2.2 Determinants of the gender wage gap
a. Education and its impact on wage gap
Over the past few decades, women have made a significant progress in terms of education. Just a few decades ago, a smaller percentage of females than males had a school diploma or a degree. Nowadays,
the situation has changed completely(DiPrete & Buchmann, 2013).In the process of globalization,
education takes on an increasing market value. Education as an investment in human capital, just like investment in physical capital, can lead to economic growth. In addition, education can stimulate growth, contributing to innovation and technological change in the economy. People's awareness of the need for education is growing because, as a result of economic integration and technological changes,
new opportunities arise for skilled workers (Singh, 2004). Research shows that women are more
educated than men. The proportion of women in the US who finish a secondary education and get a
bachelor’s degree after that is higher than among men as represented in Histogram 1. There are some
fields which are less popular among women and in which they are not that successful as men, for instance science, engineering. However, despite this, females are gradually improving their skills,
achieving progress even in these areas (DiPrete & Buchmann, 2013). However, does education narrow
the gender wage gap? Angrist and Krueger (1991) have identified a causal relationship between education and wages. The increase in the share of women in the education increased the wages of women, but this did not completely eliminate the payment inequality, women with the same level of education as men still on average receive a wage lower than their male counterparts. Nevertheless, the gender wage gap has clearly declined (Histogram 2). Now, when women with higher education levels earn more, and the proportion of those women in the workforce increases, the proportion of women out-earning men in the same age group increases as well. If men had the same education as women, their salary would have increased by about 4% and the unemployment rate would have been 0.5%
lower than today. However, Bush (2006) found out the opposite is true regarding the increase in
education level and wage gap - one additional year of schooling widens the wage gap, and these results are in line with previous findings for US in the 80-90s (Buchinsky, 1994).
Awareness of the value of education and its advantages in the labor market stimulates parents to invest more in educating their children, including girls, hoping that this will provide them with a better future. This encourages women to obtain and education and join the ranks of the workforce as a result. The increase in the level of education among women is growing faster than among men, and, as mentioned earlier, this reduces the wage gap. Nevertheless, rise in education does not universally imply a higher probability of working. The increased level of education among women in Pakistan generated by the presence of a gender-specific school that was studied by Andrabi, Das, and Khwaja (2012) did not result in larger labor force participation. A similar situation of low women labor supply was observed is the Zomba by Baird, McIntosh, and Ozler (2016), where most of women do not work or work at home, less than 8 percent of females work outside the home. These results are related to the specific cultural characteristics of these countries.
My study is based on the US, a country concerned with gender discrimination, where women are fighting for their rights and equality. So I expect that the increase in education level among women might be the reason for the decrease in wage gap.
H2: education is positively correlated with wage for both women and men
Histogram 1.
Source: World Bank
b. Age and its impact on wage gap
Saying that the labor market values professionals, regardless of their gender and age, would be unrealistic. In the modern world, both gender and age are significant factors that have a direct impact on people's income. International statistics show: the gender gap in pay increases with the age of workers. For all EU countries, the averages change from 10.9% (for workers under 30) to 16.8% (for workers over 60) (Weichselbaumer & Winter-Ebmer, 2005).
As can be seen on the Histogram 3, where the trends of the gender pay gap by age groups in the US for 2007 - 2016 represented, the same trend the wage gap between women and men increasing with age also exists. According to Graph 1, in 2007 when women and men received secondary education and started working, their starting salary was not significantly different, however, from 22 to 45 years, the gender pay gap increases rapidly - from 11% to 30%, and continues to grow until it reaches its
0 10 20 30 40 50 60 70 80 90 100 2008 2009 2010 2011 2012 2013 2014 2015 2016
Gender statistics in educational attainment
Educational attainment, at least completed upper secondary, population 25+, male (%) (cumulative)
Educational attainment, at least completed upper secondary, population 25+, female (%) (cumulative)
Educational attainment, at least Bachelor's or equivalent, population 25+, male (%) (cumulative)
Educational attainment, at least Bachelor's or equivalent, population 25+, female (%) (cumulative)
maximum of about 37% at the age of 60. But, after 60 years, the gap declines to 27 - 28% for ages over 65. In 2016, for ages 16-27 the gap is relatively small, but after 27 years it increases rapidly. It is clear that the overall gender wage gap has decreased, for all age groups except for 65 years and older. This dynamic can be related to the following factors. The overwhelming majority of women give birth to children while still under the age of 40, which results, among other things, in a prolonged pre-school period, which requires more attention, care and presence of the mother at home near her child, as well as a reduction in the opportunities for a woman to study and improve their skills. During the ten years examined, a significant increase in the gender wage gap has shifted from 20 years to 25-27 years in. This can be explained by the fact that nowadays women are more concentrated on their career, get married later and give birth to children later.
That is, it turns out that both females’ and males’ employers in the majority believe that at the age of 16-25 years it is necessary to build a career, to try and get the highest possible salary. Priorities for ladies are different after 25-30 years; they are more concerned with family, not career (Evetts, 2014). It is important to note that the gender pay gap has reduced from the peak at the age of 55 (37%) to 65 years (30%), and continues to decrease.
Graph 1.
Source: US Bureau of Labor Statistics
H3: Age is negatively correlated with wage for both sexes
c. Marital status and its impact on wage gap
Marriage is also an important factor of the difference in wages between the sexes. Hundley (2000) studied the effects of marital status on wage of males and females. The results of this research show that marriage has a direct impact on wage; marriage for non-self-employed men increases wage by approximately twenty percent, but has a diametrically opposite effect on a woman's salary. Marriage results in decrease in women’s wage of, on average, eight percent. For men, unlike women, marriage often helps in building a career. Those who consider marriage a catalyst for a career claim that men who have their own family tend to work harder, because they need to support a family. For the same reason they treat their jobs more seriously than bachelors. Koreman and Neumark (1991) in their study concluded that married men are likely to perform better, which also increases their chances for a
0% 5% 10% 15% 20% 25% 30% 35% 40% 16 to 19
years 20 to 24 years 25 to 34 years 35 to 44 years 45 to 54 years 55 to 64 years and older 65 years
The gender pay gap by age groups in the US,
surveys 2007 and 2016 (in percentage)
2007 2016
promotion, because of the sense of responsibility towards the family. It is interesting that male employers more often than women indicate a more responsible attitude towards the job on the part of "family" people. But those men also demand a higher salary. And although the income does not depend on the time spent at work, married men try to work more, as they feel the need and the duty to provide for their family. Employers of the opposite sex are inclined to think that marriage gives the employee more solidity (Daniel, 1994).
However, others believe that a bachelor is more suitable for a non-standard working hours and business trips than a family man. If your beloved wife and children are waiting at home, then you are unlikely to be willing to linger at work in order to finish a complex project. You are also more likely to dedicate your weekends to your family, and any employer wants you to think about work issues even in your spare time. Of course, if you are not officially married, this does not mean that you are not in a relationship and will not behave like a married person. But, if it is necessary to work overtime, if there is a need to go on a business trip, then the employer may prefer a bachelor employee. If the work is to be more measured, the attention of the employer will be more likely drawn to a family man, as a more responsible, judicious and reliable one. Since family life presupposes stability and reliability, the candidate is likely to possess these qualities and will be less prone to taking unjustified risks (Koreman & Neumark, 1991).
The woman's marital status also leads to an ambiguous attitude. For an employer, an unmarried woman is a potentially married woman with young children who can get sick, and a less concentrated and efficient worker, who potentially can go on a maternity leave for several years. On the other hand, the burden of family responsibilities does not lie on the shoulders of an unmarried woman, and she can devote herself entirely to her career (Joslin, 2015). Women often pay more attention to children than men. Therefore, the presence of children may lead to an increased number of delays and days off at work. In spite of this, a married woman with children may be a more attractive candidate, albeit with the same probability of maternity leave and sick leave (Dias, Elming & Joyce, 2016). However, an advantage of married woman is that the thought of personal life will not distract her from work, and the family and children make a person more organized and responsible.
Marriage affects health. Studies from around the world show that a marriage improves mental health, contributes to longevity and reduces the likelihood of cardiovascular disease, which directly affects the salary (Chandra, Szklo, Goldberg & Tonascia, 1983). In comparison with single people, married people have a higher level of mental health, they are happier and they are less prone to stress and depression. Lonely people, on the contrary, are more prone to stress, however, they are more independent and they are more likely to achieve personal growth (Wilson & Oswald, 2005). H4: Marriage is negatively correlated with wage for women and positively correlated for men
d. Children and their impact on wage gap
The presence of children has a negative impact on the career for women. The motherhood imposes a certain penalty - the birth of a child is a factor in the reduction of women's income, which results in the wage gap between men and women. Popular stereotypes contribute to the discrimination. Employers often do not favor working mothers, treating them as "unprofitable". It is believed that they have lower productivity, are less committed to their job and are often absent from work due to illness and
education of children, and, as a consequence, are less reliable than their male counterparts. Thus, female employees with children can be discriminated against in pay, both at the stage of hiring and in the course of their employment work (Correll, Benard & Paik, 2007). Traditionally it is believed that a woman should pay more attention to the family and the household responsibilities, so she can work part-time or choose a job that is easier to combine with the responsibilities of caring for the children, which affects her income. Long maternity leave and the inability to find a suitable kindergarten push
women to choose a more "comfortable" profession. Nevertheless, according to statistics, married women without children also receive lower wages. So the motherhood penalty applies to all women. This happens because of the expectation that women will get married and have a child one day, go on a maternity leave or resign. Even if they want to return to work after the childbirth, they would need flexible working hours, perhaps they would prefer to work at home or have a part-time job, which is much less profitable (Edin & Kefalas, 2011). Therefore, women are considered less reliable and stable employees.
However, kids might help men to climb the career ladder. Men with children are hired more willingly than the childless, and the salary, as a rule, increases with the growth of the family. This difference is not affected either by strict accounting of working hours or by real indicators of labor productivity - employers often believe that a male worker will work diligently for the benefit of the family
(“fatherhood bonus”), while the woman will be distracted from work (Hodges & Budig, 2010). Professor of Sociology Michel Badig of the University of Massachusetts, as a result of a study of the relationship between the presence of children and gender pay gap concludes that with the birth of a child, a man's salary grows by an average of 5.8%, while a woman's salary is reduced by 3.9%. Investigating the reasons for this gap Badig found that fathers really do start working harder to ensure a family that has grown with the birth of a child is provided for, but this only accounts for 15% of their additional income. Mothers do choose work with a more flexible schedule or work at home, so that they can combine work with childcare, but this explains only a quarter to a third of a woman's income decline (Dias, Elming & Joyce, 2016).
Graph 2.
Source: Dias, Elming & Joyce, 2016
According to data from the World Bank presented in Graph 3, in 2007 - 2016 the birth rate has decreased from 2.1 to 1.8 births per woman. This can affect the difference in pay between men and women, and in fact reduce the gender wage gap.
Graph 3.
Source: World Bank
H5: The presence of a child is negatively correlated with women’s wages, but positively correlated with men’s wages
3.
Methodology and Empirical analysis
3.1 Data and sample
The United States was chosen as a country of interest for the purposes of this study because the US economy is the largest in the world and the socio-economic processes taking place in it have a
significant impact on the development of the entire global economy (Pryor, 2000). A study on changes in the gender income gap in the United States will be conducted for the period of 2007 - 2016 and the results of the analysis will answer the main question: What happened to the gender wage gap in the US during these ten years? Information for the study is taken from the official site of the Federal reserves. The US Federal reserve conducts a Survey of Consumer Finances (SCF) every three years. SCF is a
cross-sectional survey of U.S. families, that contains the respondents’ key labor market indicators and
personal information, which is necessary for analyzing the gender wage gap at any given point in time or period. 18123 respondents over 16 years old were interviewed. After processing the data, the sample was reduced (due to some respondents providing incomplete information) to 16210
respondents, including 13560 men and 2710 women. The follow-up survey that was conducted in 2016 after the correction contains 22255 observations. This overall sample consists of 17930 male and 4325 female respondents.
3.2 The empirical model
In order to analyze the changes in wage differences between men and women over the period of ten years the Blinder–Oaxaca Decomposition method (developed in 1973) was used. The Blinder–Oaxaca approach was used for the first time in the 1970s, and after that numerous authors have used it as a basic model for identifying the gender pay gap. Some researchers have extended the decomposition model, however the initial model is still considered the most reliable. That is why the Blinder–Oaxaca approach was used for our meta-study. The wages are taken in the form of natural logarithms as the dependent variable in the regression; other variables are taken as independent variables, unexplained
1,6 1,7 1,8 1,9 2,0 2,1 2,2 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Fertility rate in the US, total (births per woman)
part of gender wage gap is also taken into account (residual U). It allows the wages to be estimated separately for individuals in different groups, males and females (Blinder, 1973). The personality variables (more detailed descriptive statistics represented in the Table 1) included in the analyses are:
– Education. Respondents were asked about their highest level of completed education (primary, secondary (lower, upper), college, bachelor, master, PhD etc.) transferred into years of
schooling.
– Age. In full years at the time of the interview.
– Marital status. Such as married, living with partner, separated, divorced, widowed, never married.
– Children. Respondents were asked if they have any children and how many of them are under 18 years.
For the purpose of the gender wage gap analysis the following regression was used: ln wit = Xit*βit + εit,
where wit denotes the natural logarithm of monthly wages for an individual i during the year t, Xit denotes a set of observed characteristics (as independent variables, the equation includes marital status, age, education level and number of children), βit denotes the regression coefficients, and εit is a random error term.
In order to investigate the sources of gender differentials in detail, researchers view men’s and women’s wage functions separately:
ln wmit =Xmit*βmit + εmit ln wfit=Xfit*βfit + εfit ,
where m represents men and f - women.
According to Oaxaca–Blinder decomposition, the gender income gap is expressed as: ln 𝑤m – ln 𝑤f = 𝑋 m β m - 𝑋 f β f (men as reference group)
Where bars above variables mean the estimation of the coefficients at average values, the mean log wages of women and men. Then, adding and subtracting the term 𝑋 f β m gives the following expression:
ln 𝑤m – ln 𝑤f = ( 𝑋 m -𝑋 f )β m + (β m - β f ) 𝑋 f Ξ E+ U
the term ( 𝑋 m -𝑋 f )β m corresponds to the part of the wage gap explained by differences in the
observed characteristics of individuals, that is the endowment effect ( E ). On the other hand, the term (β m - β f ) 𝑋 f reflects the unexplained part of the gap, which is due to differences in the coefficients of Xi. The unexplained portion of the mean outcome gap has often been attributed to discrimination, but may also be a result of the influence of unobserved variables ( U ) (Oaxaca & Ransom, 1999).
After that, the correlations between wage and each of the determinants were analyzed, using the same variables, but the martial status was taken as a dummy variable. Single is represented by zero, and married is represented by one. Dummy variable is used because the aim is to identify how the marital status of men and women correlates with their wages. This is very important since we only try to estimate the effect of being single or being married. Other marital statuses such as divorced or widowed may influence the results since they might affect the income in various ways, such as
alimony payments or subventions.
The model used to estimate the general effects of these factors on wage formation is:
lnwage women = a + 𝛽1 education + 𝛽2 married (dummy) + 𝛽3 age + 𝛽4 children + εt lnwage men = a + 𝛽1 education + 𝛽2 married (dummy) + 𝛽3 age + 𝛽4 children + εt
3.3 Results
Our primary concern was to check the regression model on multicollinearity. As the degree of multicollinearity increases, the regression model estimates of the coefficients become unstable. As a rule of thumb, a variable which has VIF larger than 10 should have a further investigation (Kumar, 1975). According to the results, presented in Table 2, there are no VIF values larger than 10, so there are no problems with multicollinearity.
Table 2.
2007 2016
VIF 1/VIF VIF 1/VIF Education 1.07 0.932491 1.05 0.956624 Marit_st. 1.82 0.550778 1.04 0.964701 Children 1.09 0.919548 1.11 0.901854 Age 1.20 0.833953 1.15 0.873297 Male 1.72 0.582015 1.87 0.534759 Mean VIF 1.38 2.24
Following the Oaxaca and Blinder (1973) decomposition method, the gross difference between
females and males can be attributed to differences in characteristics and an unexplained residual that is normally ascribed to discrimination. The results of the decomposition are presented in Table 3 for 2007 and 2016. The Blinder-Oaxaca decomposition output reports the mean predictions by groups and their difference in the first panel. In our sample, the mean of the log wages is 3.49 for men and 3.25 for women, yielding a wage gap of 0.23 in 2007. In 2016 the mean of the log wages is 3.44 for men and 3.24 for women and the wage gap is 0.20. So this indicates that the gender wage gap in 2007 was approximately 23% and it has reduced to 20% in 2016, by 3%. As the results of decomposition show, unexplained residual appears to account for a much smaller portion of the pay gap than the differences in individual characteristics between men and women. If we consider the whole population of 16+ years of age, the unexplained part of the wage gap, which is traditionally attributed to discrimination, constitutes 36.28% of the total (23.64%) for 2007 and approximately 43.61% of the total for 2016. In 2007 the explained part of the gender wage gap was 15.05%, while the unexplained was the remaining 8.57%. Differences in the marital status contribute the most to the explanation of the wage gap, which is in line with Badig’s research results (2010), and explain about 66% of the whole explained part. The next important factor is the presence of children. This explains about 7% of the gender wage gap. And the last factor is age, which explains about 2% of the whole gap. As expected, differences in level of education reduce the gender gap in wages (-4.46%).
According to the results for 2016, the wage gap has decreased by 3%, but the explained part of the gap has also decreased - only 56% of the gap can be explained. Overall, the explained part of the gender wage gap is 11.51%, while the unexplained is the remaining 8.91%. Children are the most significant factor in the explanation of the gap (6.6%), which is different from 2007, however the marital status is
still highly influential, accounting for 5.8% of the total. Age determines about 9% of the explained part of the gap, about 2% of the total. Just as in 2007, education reduces the gender gap in wages. As discussed in the previous section, the education level among American citizens increases and the rapid rise in numbers of women with higher education degrees has certainly contributed to narrowing the gender wage gap in 2016 by 2.0%.
Table 3. Oaxaca-Blinder decomposition method output 2007 lnwage 2007 lnwage Male Male Coeff SE Coeff SE Male 3.4868 0.011753 3.44192 0.0100222 Female 3.2504 0.0206095 3.23761 0.01617 Difference 0.236404 0.0237252 0.2043034 0.019024 Explained 0.1505264 0.0231608 0.1151901 0.0184063 Unexplained 0.0857776 0.0298409 0.0891133 0.0242303 Explained Education -0.0446271 0.0105926 -0.0200949 0.007676 Age 0.0205414 0.0042044 0.0103333 0.0024081 Married 0.0998947 0.0208494 0.0584959 0.0170589 Children 0.0747174 0.0013243 0.0664558 0.0008309 Unexplained Overall 0.0857776 0.0298409 0.0891133 0.0242303 N of male and female 16,270 22,255 N of male 13,560 17,930 N of female 2,710 4,325
Table 4 represents the Ordinary Least Squares (OLS) estimates of the log monthly wage equation for the male and female samples for 2007 and 2016 respectively. Coefficients and standard errors are also represented. All coefficients are significantly strong, meaning the statistic is reliable.
Table 4. OLS Regression outputs with SE, 2007, 2016
2007 lnwage 2016 lnwage
Male Female Male Female
Coeff SE Coeff SE Coeff SE Coeff SE
Age 0.00305*** -0.00084 -0.01471*** -0.00158 -0.00323*** -0.0007 -0.01389*** -0.00118 Married 0.0363*** -0.00771 -0.0218*** -0.0186 0.0396*** -0.00601 -0.0238*** -0.0123 Education 0.0347*** -0.00755 0.0228*** -0.0142 0.0385*** -0.00332 0.0306*** -0.00612 Children 0.0454*** -0.0136 -0.0222*** -0.0317 0.03121 -0.0126 -0.0959*** -0.024 _Cons 10.64*** -0.0441 10.22*** -0.128 9.536*** -0.0474 9.057*** -0.108 N 13,560 2,710 17,930 4,325 R-sq 0.203 0.128 0.213 0.179 * p < 0.05, ** p < 0.01, *** p < 0.001
a. Higher education and wage
In the previous section of the paper, we have discussed that the level of education among American citizens has increased and the percentage of women with higher education has already excessed the percentage of men. Table 4 shows the coefficient of education with a value of 0.347 for men, a value significantly different from zero. Since a logarithm-linear model is used, a change of one unit in years of schooling will result in a 0.0347*100%=3.47% increase in wage for men. In case of women, one additional year of schooling will result in 2.28% increase in wage. The findings for 2016 are different. The coefficient for men is 0.0385, for women it is 0.0306, so one additional year of schooling will result in 3.85% and 3.06% increase for men and women respectively. During the 10 years, the positive effect of a higher level of education on wages has increased for both genders, from 3.47% to 3.85% and from 2.28% to 3.06% for men and women respectively.
b. Children and wage
As discussed previously, there is a positive correlation between having children and wages for men and a negative correlation for women. In 2007, the wages for men increased by 4.54% with one additional child and decreased for women by 2.22%. In 2016, the correlation between the wages and number of children was even stronger. The coefficients represented in Table 4 are 0.0312 for men and 0.0959 for women, so one additional child will result in an increase in male worker’s wage of 3.12% and a decrease of 9.59% for women. Having a child results in larger reduction in wage for women in 2016, 9.59% compared to 2.22% in 2007. However, the benefit of having a child for men has
decreased over time, the wage increase reduced from 4.54% to 3.12% over the 10 years.
c. Marriage and wage
The results obtained by Hundley (2000) in his study were confirmed by our results presented in Table 4. Women’s wages are negatively correlated with marriage, but men’s wages show a positive
correlation. The coefficient of the regression (2007) implies that marriage has a positive effect of 3.63% for male respondents, so the wage increases by 3.63% when a man changes his marital status, but this change results in 2.18% reduction in wages of women. Similarly, in 2016, married men earn 3.96% more, and married women earn 2.38% less.
d. Age and wage
There is a negative correlation between the age of a woman and her wage, but there is a positive correlation between the age of a man and his income. Based on the 2007 results, as a man gets older his salary is increases by 0.3% every year, but there is a completely opposite situation for woman. With every year as a she gets older her wage is reduced by 1.47%. Compared to the situation 10 years later, the indicators are approximately the same. Each one year results in an increase of 0.3% in wage for a man and a reduction by 1.38% in wage for a woman.
4. Conclusion
In this paper the trends in gender wage gap dynamics in the US in 2007 – 2016 are analyzed based on the data from the Survey of Consumer Finances conducted by US Federal Reserve. Women in the US are in a disadvantaged position relative to men. According to Oaxaca and Blinder (1973)
decomposition method, the gender wage gap has reduced by 3% during the last decade, from 23% in 2007 to 20% in 2016, but the gap is still substantial. However, only about 2/3 of the gap can be explained, so there is still a significant part that cannot be explained by differences in the observed
characteristics of individuals and may be attributed to discrimination. The wage, the dependent variable, is influenced by many factors, including education, age, marital status and the presence of children. Additional years of education have a positive effect on the wages of both sexes, men and women, but men's investment in education is slightly more rewarded. Based on the results of our analysis, the increase in the level of education among women has had the most significant impact on
narrowing the gender wage inequality. The results also show that one additional year of education will
result in higher wages for both genders, so education and wage are positively correlated. However, there is a positive correlation between wages of men and negative correlation with women’s wage, the same holds for the correlation between marriage and age and wage.
As was mentioned above, there is still a significant part of the wage gap that cannot be explained by the aforementioned factors. This unexplained wage gap might be a result of discrimination, gender differences in lifestyle, the fact that women often just choose less well-paid jobs, such as a teacher or a nurse, or simply by labor market preferences. This should be addressed in future research. In addition, our results show a necessity for further exploration of the effects on the wage gap of other possible determinants of gender preferences such as, for example, competitiveness and sphere of occupation. This study has some limitations. It is based on a survey that includes roughly twenty thousands
respondents. It is just a random sample from the population of the US; the research is not based on the whole population. The analysis of personal human characteristics is based on data that respondents of the survey reported and the personality traits are considered after the respondents have entered the labor market, so we cannot examine the effects of their labor market experience on these traits. Also the results of our analysis might be subject to omitted variable bias. We could miss out the important variable that is correlated with both the dependent variable (lnwage) and one or more of the included independent variables (Clarke, 2005).
Reference list
Albrecht, J., Björklund, A., & Vroman, S. (2003). Is there a glass ceiling in Sweden?. Journal of Labor economics, 21(1), 145-177.
Angrist, J. D., & Keueger, A. B. (1991). Does compulsory school attendance affect schooling and earnings?. The Quarterly Journal of Economics, 106(4), 979-1014.
Berger, C. G. (1970). Equal Pay, Equal Employment Opportunity and Equal Enforcement of the Law for Women. Val. UL Rev., 5, 326.
Blinder, A. S. (1973). Wage discrimination: reduced form and structural estimates. Journal of Human resources, 436-455.
Buchinsky, M. (1994). Changes in the US wage structure 1963-1987: Application of quantile regression. Econometrica: Journal of the Econometric Society, 405-458.
Clarke, K. A. (2005). The phantom menace: Omitted variable bias in econometric research. Conflict Management and Peace Science, 22(4), 341-352.
Correll, S. J., Benard, S., & Paik, I. (2007). Getting a job: Is there a motherhood penalty?. American journal of sociology, 112(5), 1297-1338.
Cotter, D. A., Hermsen, J. M., Ovadia, S., & Vanneman, R. (2001). The glass ceiling effect. Social Forces, 80(2), 655-681. doi:10.1353/sof.2001.0091
Dias, M. C., Elming, W., & Joyce, R. (2016). The Gender Wage Gap. Institute for Fiscal studies DiPrete, T. A., & Buchmann, C. (2013). The rise of women: The growing gender gap in education and
what it means for American schools. Russell Sage Foundation.
Duraisamy, M., & Duraisamy, P. (2016). Gender wage gap across the wage distribution in different segments of the Indian labour market, 1983–2012: exploring the glass ceiling or sticky floor phenomenon. Applied Economics, 48(43), 4098-4111.
Edin, K., & Kefalas, M. (2011). Promises I can keep: Why poor women put motherhood before marriage. Univ of California Press.
Evetts, J. (2014). Women and career: themes and issues in advanced industrial societies. Routledge. Geneva, Switzerland: International Labour Office. (2017) Global Wage Report 2016/17: Wage
inequality in the workplace. Retrieved from: http://www.ilo.org/wcmsp5/groups/public/---dgreports/---dcomm/---publ/documents/publication/wcms_537846.pdf
Hodges, M. J., & Budig, M. J. (2010). Who gets the daddy bonus? Organizational hegemonic masculinity and the impact of fatherhood on earnings. Gender & Society, 24(6), 717-745. International Labour Organization (2015): “Global Wage Report 2014/15, Wages and income
inequality”. International Labour Office, CH-1211 Geneva 22, Switzerland Joslin, C. G. (2015). Marital Status Discrimination 2.0. BUL Rev., 95, 805.
Kumar, T. K. (1975). Multicollinearity in regression analysis. The Review of Economics and Statistics, 57(3), 365-366.
Langdon, D. L., & Klomegah, R. (2013). Gender wage gap and its associated factors: An examination of traditional gender ideology, education, and occupation. International Labour Review, 173-203.
Martins, P. S., & Pereira, P. T. (2004). Does education reduce wage inequality? Quantile regression evidence from 16 countries. Labour economics, 11(3), 355-371.
Mussida, C., & Picchio, M. (2014). The trend over time of the gender wage gap in Italy. Empirical Economics, 46(3), 1081-1110.
Oaxaca, R. (1973). Male- female wage differentials in urban labor markets. International Economic Review, 14(3), 693-709. doi:10.2307/2525981
Oaxaca, R. L., & Ransom, M. R. (1999). Identification in detailed wage decompositions. Review of Economics and Statistics, 81(1), 154-157.
Pryor, F. L. (2000). The millennium survey: How economists view the U.S. economy in the 21 st century. American Journal of Economics and Sociology, 59(1), 333. doi:10.1111/1536
-7150.00002
Razavi, S. (2016). The 2030 Agenda: challenges of implementation to attain gender equality and women's rights. Gender & Development, 24(1), 25-41.
Singh, P. (2004). Globalization and education. Educational theory, 54(1), 103-115.
Suh, J. (2010). Decomposition of the change in the gender wage gap. Research in business and economics journal, 1, 1.
Weichselbaumer, D., & Winter-Ebmer, R. (2005). A meta-analysis of the international gender wage gap. Journal of Economic Surveys, 19(3), 479-511.
Wilson, C., & Oswald, A. J. (2005). How does marriage affect physical and psychological health? A survey of the longitudinal evidence
Zheng, H. (2009). Rising US income inequality, gender and individual self-rated health, 1972–2004. Social Science & Medicine, 69(9), 1333-1342.
Appendix
Histogram 2.
Source: US Bureau of Labor Statistics
Histogram 3.
Source: US Bureau of Labor Statistics
0% 5% 10% 15% 20% 25% 30% 35% 40% 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
The gender pay gap by educational
attainment in the US,
2007–2016
Less than a high school diploma High school graduates, no college
Some college or associate's degree
Bachelor's degree and higher
0% 5% 10% 15% 20% 25% 30% 35% 40% 16 to 19
years 20 to 24 years 25 to 34 years 35 to 44 years 45 to 54 years 55 to 64 years and older 65 years
The gender pay gap by age groups in the US,
surveys 2007-‐2016 (in percentage)
2007 2008 2009 2010 2011 2012 2013 2014 2015
Table 1. Descriptive statistics
2007 2016
Variable Description Mean SD Mean SD
Wage 967.54 979.48 1102.14 1111.52
Marital status
Married 1 if individual is married, 0 otherwise 0.39 0.47 0.41 0.51 Living with partner 1 if individual is living with partner, 0 otherwise 0.25 0.38 0.28 0.41 Divorced 1 if individual is divorced, 0 otherwise 0.22 0.29 0.25 0.33 Never married 1 if individual never married, 0 otherwise 0.18 0.27 0.13 0.25 Widowed 1 if individual is widowed, 0 otherwise 0.11 0.21 0.09 0.19 Education Years of education attained 9.46 11.15 12.87 14.16 Sex 1 if individual is a man, 0 if woman 0.82 0.77 0.79 0.73 Kids Amount of children the individual has, continuous 0.47 0.62 0.39 0.58 Age Age of the person, continuous form 16 to 80 38.17 11.96 40.02 12.66
