A study about gender pay gap for nurses in Denmark
Is there a the gender wage gap for nurses in Denmark?
Author: Marcus Hansen Advisor: Chizheng Miao
Examinator: Mats Hammarstedt Date: 27/5- 2020
Subject: Economics Level: Bachelor Course code: 2NA11E
Abstract
This study investigates the gender earnings gap among nurses in Denmark years 2004-2016.
The data at hand will be from Luxembourg Income studies which provided 7078 observations. Furthermore, ordinary least squares method with gender as dummy variable will be conducted. The findings are a raw male-female annual wage gap of 13 percent. After
adding control variables, the gap decreased to 7.4 percent. The remaining wage gap can be due to unobservable characteristics. However, discrimination cannot be discounted.
Keywords
Gender, Earnings, Gap, Discrimination, Nurses, Denmark
Acknowledgments
First, I would like to address my greatest gratitude to my supervisor Chizheng Miao. Who has been a great support in the writing process and the statistical part. Without his guidance, this
thesis would not be possible.
Additionally, thanks to Mats Hammarstedt for his comments which contributed to this thesis.
Finally, I appreciated the constructive criticism from my discussant, Tilda Nilsson.
Contents
1. Introduction………..…………....…...5
2. Literature review……….………...…..7
3. Theoretical framework……….……….11
4. Data………..………14
5. Methodology………..…...…………...17
6. Results………..………18
7. Discussion………..………..……21
8. Conclusion………..……….23
References……….……….………...24
Appendix……….………...….………26
1. Introduction
The gender pay gap has been investigated for a long time. It is still an ongoing debate and remains an area of research. Regardless of remarkable gender convergence over the last century, gender pay gap is still substantial in many countries. The early literature regarding gender wage gap concentrated on human capital variables and discrimination (Altonji &
Blank 1999). However, the last century females have increased their productivity characteristics. The Human capital part explains the wage gap less today, meaning the explanation for gender inequality on the labour market is elsewhere (Blau and Kahn, 2016).
Denmark has one of the highest female work participation in the world with about 80 percent, in comparison to US with 70 percent (Kleven et al. 2018). Furthermore, Denmark has the highest number of nurses per capita in Europe. This is about double the EU average, 16,7 versus 8,4 nurses per 1000 citizen (European commission, 2020). There are about 35,216 nurses in Denmark (National board of health, 2019). This occupation is very female dominated. Although the number of male nurses is increasing in Denmark, it remains only around 10 percent.
General wage gap in Denmark is around 13 percent (Statistics Denmark, 2020). Some
researchers have shown that the gap is as high as 20 percent (Gallen et al.2019). Gender wage gap is common in many countries. However, there less study about the gender pay gap in the nursing sector, especially in the Danish context. Within the healthcare sector, nursing is the largest of all occupations. This is why policy makers should care about the pay gap in nursing sector, which a significant number of the nurses are women.
This is an interesting topic, as previous literatures show that men dominated professions usually have larger wage differentials between men and female. Further, Magnusson and Nermo (2020) have shown that the gender wage gap is worse in high prestige occupations, such as lawyers and university professors, in which require long educations. However, we are not sure whether there is a gender pay gap in the nursing sector in Denmark and whether the gap is smaller in the occupation which is less prestige.
This segment is important to study as gender equality is important, especially for more developed country as Denmark, in which can serve as a guideline for less developed countries. The Scandinavian countries are foremost at offering opportunities for female to
balance work life and family, in comparison to other countries (Kleven et al. 2018).
Additionally, because it provides better political decision making. When studies about degree of wage differentials are provided to politicians, they are able to formulate new policies. This is for a more equalized labour market.
Previous studies have analyzed the gender wage gap across the labour market and the main factors leading to it (Blau and Kahn, 2016). Additionally, literature regarding pay gap in medical sector is substantial (Lo Sasso et al., 2011). But there are few studies about the gender pay gap in the nursing occupation. Furthermore, to my knowledge, this is the first study on gender inequality on nurses in Denmark.
Therefore, in this study, I want to answer the following questions: What is the degree of the gender wage gap for nurses in Denmark and what are the influential factors contributing to the gap?
In this paper, I provide new empirical findings regarding gender wage gap in the segment of nurses. This study will contribute with more up to date data. Moreover, this study will be much more focused on the occupation nursing in Denmark than previous literature. I use the Danish data from Luxembourg income studies (LIS), from year 2004 to 2016. I perform ordinary least squares regression with gender dummy variable as the variable of interest. The annual wage of main job will be the dependent variable. Important explanatory variables will be included, that earlier literature proves to be of importance, such as year of education and number of children.
In section 2 you will find a review of carefully picked literature and previous research on which I will base my study on. Additionally, these previous studies will guide me to the results I can expect to receive. Section 3 provides theories about gender wage gap like discrimination theories and family situation. Data is presented in section 4. Methodology can be found in section 5 in which I explain the OLS method. Results from the regression model in section 6. Section 7 provides discussion which I discuss omission of variables and the annual data. Lastly, you will find conclusion, in section 8.
2. Literature review
In this section, I review and present relevant literatures that could later on be compared with my findings.
Blau and Kahn (2016) provide a comprehensive review about the gender wage gap.
Throughout the years 1980-2010 the wage differentials in US between men and female has decreased considerable. From a wage gap of 40 to 23 percent. In 2014, the full time working females earned 79 percent of what men did. The role of traditional human capital variables, such as education and working experience, in explaining the gender wage gap is declining.
The working sector and occupation explained 27 percent of the wage gap in 1980. In the same year, the human capital variables explained 28,6 percent of the gender wage gap.
However, in the year 2010, human capital variables explained only 14,8 percent while the working sector and occupation together explained 50,2 percent of the gender pay gap. The authors comments; “As women have increased their productivity enhancing characteristics and as they ‘look’ more like men, the human capital part of the wage difference has been squeezed out.” Furthermore, the gender wage gap declined a lot slower for high income earners than low income earners. By 2010 the wage differentials were particularly higher for the high-income earners.
Blau and Kahn (2016) suggest that discrimination towards women cannot be discounted.
Empirical evidence shows a negative relationship between children and female wages.
Previous research has shown women tend to carry more responsibility for nonmarket work, such as staying at home with their children. This traditional gender role negativity affects women's labour market outcome.
Blau and Kahn (2016) comments about new explanations for the gender wage gap. These new explanations are more directed to personality traits. The authors found that women have lower propensity to negotiate. This could be negotiation about wages, promotion or bonuses.
This can negatively affect the pay gap and enhance it. One explanation as why women in their study had lower propensity to negotiate is due to traditional gender roles; women feels pushier or overbearing when discussing salaries. Croson and Gneezy (2009) has shown that women are more risk averse in comparison to men, meaning men take more risk in the labour market than women do. This is connected to women having lower propensity to negotiate as
it might be seen as a risk.
One Swedish study by Magnusson and Nermo (2020) shows that gender wage gap is worst in high prestige occupations, such as lawyers and university professors. They claimthat the wage gap between men and female has decreased remarkably since 1980, except for high prestige occupation in which require long education and work life experience. The authors suggest that the wage gap for these occupations has not decreased like the rest of the labour market during the last decade. To quote the author Magnusson and Nermo; “we can now see that it is in the low and medium prestige professions, where women and men are still quite divided by occupation, that the gender wage gap has decreased”.
Magnusson and Nermo (2020) used the Swedish data from the “The Level living survey” and analyzed the hourly wages. They found that occupation with high prestige women as a group still earn less than men with the same occupation. In this type of occupation segment, the gender wage gap has been the same between 1968 and 2010. Contrary to similar research, Magnusson and Nermo (2020) concluded that the gender wage gap cannot be explained by experience, education or whatever the individual is employed in private or public
sector. Instead, they explained that women are more likely to have the main responsibility for their children at home. High prestige professions often require a lot of time and traveling.
This makes it highly inconvenient for women in which have responsibility for their children, to have a high prestige carrier path. The wage differentials in their latest survey 2010, was 15 percent. Additionally, the medium and low prestige occupation, had 10 percent wage gap.
This study shows that gender pay gap could differ across occupations.
Dietrich and Muench (2019) did a study regarding the gender pay gap for nurses in Germany.
Their theoretical framework was based on discrimination theories. Especially statistical discrimination, which is based on stereotypes. This kind of discrimination may be present on average. Because women tend to take more responsibility for their children. This means women spends more time outside the labour market.
They used cross sectional data collected 2006 and 2012 in which included 828 observations of actively working nurses. Their method was based on ordinary least squared and estimated the log of monthly salary. The model included many of the traditional variables when studying gender pay gap such as; human capital, education and years of nursing experience.
The authors also did a follow up review for alternative explanations for the wage
differentials, in which included motivation for working uncomfortable hours and years with the same employer.
Dietrich and Muench (2019) found that unadjusted monthly salary for full time male nurses were 30 percent higher than for female nurses. However, in the fully adjusted model. Which controls for many more factors. The pay gap was 9,3 percent in favor for the male nurses.
Lo sasso et al. (2011) study found a gender pay gap 1999-2008 for newly graduated physicians. Their study focused on New York and had a sample size of 8233 observations.
4918 men and 3315 women. Ordinary least squares method was used to estimate the differences in mean wages over time.
In 2008, newly trained physicians earned an average of 16,819 dollars more than newly trained female physicians. They conducted that the pay gap cannot be explained by specialty choice, worked hours or other characteristics. No other human capital variable could explain this either. Because all individuals in the study had the same education and work experience.
The study also shows that women tend to specialize in the less paying areas. For example, 14 percent of female are specialized in pediatrics, versus only 5 percent of men. This
specialization is the lowest paying. In this specialization, female had average wage of 116,950 dollars and men 125,343 dollars. The authors suggest that policy makers should rethink how they construct their working arrangements and how they pay.
One disadvantage with this study is the data. It lacks information regarding marital and family status, which are two important characteristics. Because women tend to take more responsibility for family and stay at home more.
Occupations associated with men have, on average, larger gender pay gap. However, even within the same occupation, we still observe the gender pay differentials. Wilson et al. (2017) analyze gender pay differences between two female dominated professions: registered nurses and teachers. Their sample included 427,080 nurses and 965,878 teachers in US. It was conducted years 2000-2013. Only 8 percent of the nurses and 23 percent of teachers were male. According to the authors, female experience gender pay gap for every education level and occupation category. Female dominated occupations have lower wages due to
occupational segregation. Excess labour force put downwards pressure on wages.
Nevertheless, this does not explain the within-occupation pay difference. The results were that male out earned females in both occupations. The authors comments that females are wage discriminated in almost all work sectors.
Kleven et al., (2018) analyze the impact of children on gender inequality in Denmark. In contrast to similar studies in which focused a lot on labour variables such as occupation and industry, the authors used Danish administrative data from 1980-2013 and conducted an event study approach. The paper suggests that the gender wage gap is not explained by human capital or discrimination, but is due to children. This is because female’s productivity attributes have increased massively during the last decade. Considering only human capital variables, men and females are today very similar. Moreover, the authors do not suggest that discrimination explains the gap, because of anti-discrimination policies. The authors focused on the “child penalty”, which is defined as the percentage by female gets reduced compared to men due to children. The overall gender pay gap in Danish labour market is about10-15 percent. However, the authors found that the long-run child penalty for females 1980-2013 is a wage gap of 20 percent. The main reasons for this wage gap are; lower wage rates, less labour participation and fewer hours worked. Additionally, the paper shows that the child penalty is passed down in generation within families, which is from mothers to daughters.
One Danish study by Gallen et al., (2019) investigated the wage gap between 1980-2010 and shows a comprehensive overview of the female labour market in Denmark. They used Danish administrative database IDA, which contains micro income information regarding the Danish population. Almost 1,6 million observations were conducted in 2010. Their results show that the last 30 years, gender segregation and the wage gap has declined drastically, which is mostly driven by the increase in working hours among females. Moreover, the wage gap has decreased because human capital variables such as experience, occupation and education explains less of the gap. In 1980 these variables explained more than half of the 30 log points to the wage gap. Compared to 2010, which these variables explained only one third of the 20 log points. However, gender earnings gap is present around 20 percent for the Danish labour force. According to the authors; “The combined effect of hours and wages is a more than 20 percent gender earnings gap”. Once again, the motherhood penalty serves as an explanation.
The authors show that women with children have 9 percent lower hours worked. The worked
hour gender gap is mostly driven by mothers. Variance in education, experience and
occupation proves to create higher gender earnings gap for mothers, in comparison with the rest of the population.
3. Theoretical framework
Many theoretical frameworks in literatures tries to explain the gender wage gap in the labour market. I present several theories that are relevant to my research question. The
discrimination theories are of great value for my study. Because it might offer explanation.
When all human capital variables are calculated, the remaining unexplained part can be due to discrimination. And many previous studies have shown that it cannot be disregarded. Blau and Kahn (2016) Additionally, theories of segregated workforce suits my study well. As mentioned before, the nursing sector is female dominated profession. Pay gap in this occupation, points to the vertically segregation.
3.1 Discrimination theories
Labour market discrimination is a broad concept. Is it defined as when one individual is equally productive in material or physical aspects. But is treated unequally due to characteristics such as race, gender or ethnicity. Moreover, when participant in the marketplace makes economics decisions because on one's race or sex.
If gender pay gap is present even if accounting for all human capital variables, labour market discrimination theories can offer an explanation. However, this is hard to prove. As this unexplained part can be due to productivity characteristics for both men and female. In which is hard to put a value on and therefore, hard to run a linear regression on. Yet, previous studies have shown that discrimination cannot be discounted (Blau and Kahn, 2016).
This economic theory is based on Beckers (1957) The economics of discrimination. Becker's study was highly influential. Many modern studies are still guided by the theoretical
framework set by Becker back in 1957. Therefore, i find it very appropriate to include Becker’s study.
3.1.1 Taste-based discrimination
This theory was introduced by Gary Becker 1957 and he wanted to explain discrimination by using a relevant measure. This discrimination translates racial prejudice into economic language, that is, monetary terms. If an employer discriminates a certain group, this individual makes economic decisions regardless of the profit maximizing. The individual would act like there was an indirect cost related to this discriminated group. Becker (1957) presented a taste parameter, that he called “coefficient of discrimination”. Which is a positive number. Suppose there are have two groups, men and female. An employer who
discriminates women, will to have to pay the female wage plus this taste parameter. While wages for men remain the same. This put female wages on a higher level. That means female labour force becomes more expensive, compared to men's. The employer will only hire men in this case. As he will only hire women if female wage plus taste parameter is lower or even equal to men wages.
From this type of discrimination, I would expect a discriminating employer to be prejudiced.
Hence the employer will either only hire men or hire female but with a lower wage. The firm is not profit maximizing and will achieve levels below this desired level.
3.1.2 Statistical discrimination
As previous studies show, marital status and number of children serves as an explanation of the wage gap. Because of this, statistical discrimination might be present.
Statistical discrimination is a theory regarding gender or racial inequality. It implies that individuals can be judged by the average performance of a group. The employer can
“statistically discriminate” by assuming workers skills, solely based on the group's average skill. In this case, the group is females (Becker ,1957). Companies can use easily visible characteristics such as gender and ethnicity to form expectations regarding the individual's performance. Especially if the company in question lacks information about the individual.
For example; employers have the knowledge that females tend to stay at home with their children more than men. This might result that newly hired female workers are not getting the same salary or the same opportunities as their male colleagues. Because the employer
assumes that the newly hired female will stay more at home, in the future.
3.2 Segregated workforce
Occupational segregation is the distribution of workers across or within jobs. Is it the idea that some occupations are overwhelmed by men like industry work. While other line of work is dominated by female, like nurses. In 2015, men had 98 percent of construction jobs in United States. (Washington center, 2017). The segregates workforce has two explanations.
The most common explanation for occupational segregation is due to personal preferences. It bases on that men and female have different preferences regarding what occupation they strive for. Women might simply choose line of work in which require less education and is less time-demanding (Reeves, 2010)
The other, more sophisticated, explanation for occupational segregation is human capital, such as education, experience and training. Research shows that male and female invest their own human capital in different ways. Women tend to invest less in human capital. This results in segregation as the employment outcome differs between men and female (Reeves ,2010).
3.3 Family situation
The family situation such as married or number of children can provide explanation for gender wage gap. As previous literatures show, the family situation is very important. Many studies show that the remaining part of the wage gap is due to children (E.g. Kleven et al., 2018; Magnusson and Nermo, 2020). The explanation of human capital variables part has been squeezed out, due to females enhancing productivity characteristics (Blau and Kahn ,2016). This suggests that the cause gender wage gap lies elsewhere and that family situation can serve as explanation.
Children effects labour market outcome in two ways, the first is pre-child effect; women who is expecting children may invest less time in education and choose more family friendly occupation, which provides less traveling and more comfortable hours. The second effect is post-child affect; females changing their occupation, firm or even industry because of motherhood (Kleven et al., 2018).
Whether a female is married or not proves to affect labour outcomes. Married women tends to make likewise decisions as mothers. This can be because of pre-or post-child effect
mentioned. If a female is newly married, expectation about children can be present and hence chooses to invest less time in education. Furthermore, choose occupation with certain
characteristics, such as close to home.
4. Data
In this section, a brief presentation of the data source will be presented. Additionally, the data selection of which the restrictions are shown. Lastly, I present the data description and the variables used.
4.1 Data source
To investigate the gender pay gap in the nursing sector in Denmark, I use the individual level data from Luxembourg income studies (LIS from now on) from years 2004, 2007, 2010, 2013 and 2016. The database contains cross national microeconomic income data made for research purpose only. This database includes 300 datasets from over 50 countries, many countries have been represented for 30 years. LIS does not conduct own surveys, but are acquired by their data provides. For Denmark, Ministry of finance and Statistics Denmark is responsible for the data. The database contains random samples of all Danish households and individuals. The sample size represents 1/30 of total population for every survey year,
approximately 180,000 individuals.
4.2 Data selection
To address the gender wage gap among the nurses in Denmark, I made several restrictions on the sample. I choose to focus on individuals whom main occupation is nursing. I also exclude any part time working. Secondly, as LIS contains both household and individual level data, in this study, I focus solely on individual level data. The original sample size of Danish data is 537,309 observations. However, I want to focus on individuals who has nursing profession as main occupation. This restriction leads me to sample size of 7,332 nurses. Also, I want to restrict my study to only nurses with positive labour income. This reduces the sample down to 7,195 observations, 6,471 females and 7,24 males. Furthermore, I want to focus on
working age population and not include any pensioners. Current retirement age in Denmark is 65 (Statistics Denmark 2020). After this correction, I end up with sample size of 7,078, of which 6381 females and 697 males.
The earnings variable is annual labour earnings. I used Danish consumer price index to adjust wage data into 2016 price level. Important to notice is that I use annual wage income and not hourly wage income. It is defined as payments from regular and irregular dependent work employment. It includes salary income and monetary supplements, such as overtime and bonuses. However, drawback of this annual wage income is the lack of information regarding any breaks in the labour force. Additionally, annual wage can also depend on total working hours. This may in fact, also cause gender wage gap.
The variable of interest is the gender. Traditional human capital variables are also included in this study. These are; age, years of education and immigrant. As previous literature shows, the family situation is very important. Therefore, the number of children and marital status is also included in this analysis.
The summary statistics for the data that will be used in the linear regression are presented below in table 1.
Table 1: Descriptive statistics
The summary statistics represents the mean and can be interpreted as the average. The table also includes the standard deviation of each variable, which shows how much the value of the variable can deviate from the average. The currency in table 1 is Danish crowns, as
mentioned before, adjusted to 2016 price level and it is measured in 1,000 kr.
Noteworthy is that the nursing occupation is very female dominated. Women represents 90 percent of the observations. Variable of interest are the wages: Females has an average annual wage income of 369.010kr and men has 539.856kr. This indicates for an annual earnings gap of 31,6 percent. Noticeable is that differences in the standard deviation is high, compared to the other variables. Men's standard deviation in annual wage income is
300.303kr versus females 119.166kr. Showing a bigger wage distribution for male nurses. In terms of the logarithm wage, females had an average logged annual wage income of 12.75, compared to men with 12.92. The average age of all the individuals are fairly similar. The average male is almost 2 years older than the average female. Men tend to invest more time in education, the average male in this study has 1 additional year in education than women.
Furthermore, the majority of men and women are married, which the share of married men and women are 65 and 68 percent respectively. By looking at the table, this study contains very few immigrants, only 6 percent of females are immigrant versus 12 percent male.
Females has an average of 1.19 children, compared to men in which has 1.04 children on average. This shows that females have 12.6 percent more children than men in this study
5. Methodology
In this section, I present the regression model specification to investigate the gender wage gap. The method used in this study to investigate gender earnings gap is Ordinary Least Square method (OLS from now). The individual weight has also been taken into
consideration when running the regression. This is used to increase the results to better reflect the total population in the given dataset.
I will use this following equation;
Yi= α1 +D1 gender + D2 marital status + D3 immigrant + α2 age + α3 children + α4 education + α 5 year dummies + µi
For details and definitions regarding the variables, see appendix.
The income variable is (Yi) which indicates annual logged earnings and will be the dependent variable. This will also improve the model because it reduces the problem of
heteroscedasticity, because the logged values are normally distributed.
The variable of interest will be the gender variable (D1). This is a dummy variable, which can only present two values, 0 if female and 1 if male. The gender dummy variable captures the degree of earnings gap between male and female workers in the nursing sector.
Furthermore, the family status is also included in the model. Firstly, by the married dummy variable, 1 indicates married and 0 indicates not married. Secondly, number of children which is a continuously variable. I have also other control variables in the regression, like age, education and immigrant. The second to last variable are the year dummies, so that it compares the wages within the same years.
Ui is the error term and is present when the model does not fully explain the actual relationship between independent variable and the dependent variables. This error term is used to demonstrate the uncertainty in the model.
5.2 Estimation
Firstly, I perform a regression with only the annual wage income as dependent variable and gender as independent. This is to see the raw and unadjusted gender wage gap and will give overall estimation of the earnings gap. Additionally, another regression takes place called adjusted. This specification controls for all the variables previously mentioned. Given earlier studies mentioned in section 2. I expect a decrease in the wage gap in the second regression.
Also, the data at hand in section 4 indicates for a gender wage gap. We can expect a small part of the unexplained part due to career breaks, because of the annual wage income.
Before section 6, there are two important variables to acknowledge in order to interpret the regression results. These are the coefficients and p-values. Firstly, the coefficients show the relationship between the independent variable and the dependent variables. It indicates how much the mean of the dependent variables shifts when the independent variable shifts one- unit. Hence, a positive coefficient shows if the value of independent variable increases, the mean of the dependent variable follows. Moreover, a negative coefficient means if the independent variable increases, the dependent variable will decrease. Also of importance is the p-value, which indicates if the coefficients are significant or not. This value tests if the independent variables have correlation with the dependent variable.
6. Results
In this section, I present the results from the OLS linear regression model.
Table 2 presents the main results of this study. First column shows the raw difference, which is the regression of log annual wage earnings on gender. This only controls for years and provides an overview of the wage gap. On average, females earn a lower annual wage income, by 13 percent, than men in the nursing sector in Denmark. Moreover, the p-value indicates that the relationship between annual earnings and gender is highly significant at 1 percent level.
In the second regression, I controlled for all the demographic- and human capital variables.
The inclusion of those variables reduces the gender earnings gap, from 13 percent to 7,4 percent. This significant drop indicates that demographic- and human capital variables
explain more than half of the gender earnings gap.
Age is significant correlated to the wages. If the age of one individual increased with one year, the annual earnings increase with 1.1 percent. The only negative coefficient in the results is marital status, indicating on a negative relationship. Therefore, being married negatively affects the annual wage earnings. Married individuals earn 2.6 percent less
compared to non-married. Being an immigrant proves to be insignificant, because the p-value is very high. Immigrants earn more than natives with an average of 2.9 percent. The variable with the highest coefficient is years of education. An increase of one more year of education would increase the wage by 11.5 percent. Having one child indicates a wage increase of 2.15 percent and the point estimate is significant at 1% level.
Kleven et al., (2018) showed a gender wage gap in Denmark of 10-15 percent. My result indicated on unadjusted earnings gap of 13 percent. Hence, those result are very similar.
However, Kleven et al., (2018) claims that the remaining gender earnings gap is due to children. According to my findings, education explains most of the gender earnings gap.
Years of education is associated with 11.5 percent wage increase and number of children 2.15 percent. Contrary to their study, children have smaller positive effect on wages. Gallen et al.(2019) paper also suggest that gender earnings inequality are mostly explained by children and shows a 20 percent gender wage gap in Denmark. My results suggest a lower wage gap of 13 percent.
My findings suggest that gender pay gap is lower in nursing sector than average gender pay gap in Denmark. The pay gap can be smaller in this occupation in comparison to others because of the female dominance, because male dominated professions shows to have more gender wage gap. Furthermore, the gender earnings gap can be smaller due to the nurses having lower wages if compared to the overall labour market in Denmark. Previous studies show, gender wage gap is more significant for high-income earners (Blau & Kahn, 2016).
Additionally, as mentioned before, gender earnings differentials are worst in high prestige occupation. This may also result in lower gender wage gap in the nursing sector. These are the characteristics of the occupation that may contribute a smaller gender pay gap.
Table 2: Regressions of log annual earnings
7. Discussion
7.1 Omission of experience & annual wage data
Because of the lack of working experience and annual wage data in my regression analysis, my findings can be biased. The remaining wage gap of 7.4 percent can be further decreased by adding these two components. The unexplained part can partly be due to career breaks, because of the annual wage income. Furthermore, annual wage may depend on the total hours worked within the year, this may also affect my findings. Ideally, one could use the hourly wage data. The omission of working experience has to be accounted for. Of course, this variable would have provided a better result. As the earnings gap can be contributed because of experience. However, previous studies mentioned in section 2, shows that working
experience explain very little to the wage gap in today.
7.2 Sensitivity regression check
Due to outliers, the results of earnings gap can be partly affected. Therefore, a sensitive analysis is conducted. This is done by running the quantile regression. By doing this, I can analyze the gender earnings gap for nurses with similar wage income. First is the 25th
percentile, which represents the low-income earners. Secondly, the 50th percentile, which can be viewed as the median income earners. Lastly is the 75th percentile, showing only the high- income earners. The results are presented in table 4.
Notable is the remarkable increase of the gender coefficient. For the lower income earners, the wage gap is 19.5 percent. However, the gap increases to 41.9 percent for the high-income nurses. This confirms (Blau and Kahn, 2016) findings, that gender gap earnings are
particularly higher for high income earners.
The coefficient for education remained similar throughout the wage distribution, also the significance remained the same. In line with this study previous results, marital status negatively affects the wages in all percentiles.
Table 3. Sensitivity regression check
8. Conclusion
This study aims to investigate whether there is a gender earnings gap for nurses in Denmark
Yes, there is an existing gender wage gap for nurses in Denmark. When not controlling for any demographic- and human capital variables, female nurses earns on average 13 percent less than male nurses. When controlling for traditional factors, the earnings gap substantially dropped to 7.4 percent. The explanatory variables used in this study have contributed to the reduction of the gender pay gap. However, I cannot claim which factor is more influential than the other.
The remaining unexplained part of the gender earnings gap can be due to discrimination.
Both statistical and taste-based, suggested by Gary Becker. However, the omission of working experience and the use of annual wage data can serve as explanation. Furthermore, unobserved factors in which are hard to measure can also explain the gap. These can be personality traits, motivation and ability to work.
I believe that future studies should include working experience variable and use hourly wage data. This would contribute remarkable to this study. Furthermore, the Oaxaca decomposition can be also conducted.
For policy implantations, I encourage the Danish society to keep striving for gender equality.
Because the earnings gap might be explained by discrimination, I suggest that Danish government investigates this further and perhaps add newer discrimination polices.
Moreover, I suggest that society should rethink the normative gender roles so we can have more men in the nursing sector.
9. References
Altonji, J. & Blank, R. (1999) Handbook of labour economics, North-Holland.
Beckers, G (1957) The economics of discrimination, University of Chicago press, Chicago.
Blau, F. & Kahn, L. (2017) "The Gender Wage Gap: Extent, Trends, and Explanations."
Journal of Economic Literature, 55 (3): 789-865.
Biblarz, T. J.; Bengtson, V. L.; Bucur, A. (1996). "Social mobility across three generations".
Journal of Marriage and the Family. 58 (1): 188–200
Borjas, G. (2012) Labour economics, Harvard university, Massachusetts.
Croson, R. & Gneezy, U. (2009) “Differences in preferences”. Journal of Economic Literature, 47 (2): 448-74.
European commission. (2020) State of Health in Europe, Denmark. Retrieved 20 May 2020:
http://www.euro.who.int/__data/assets/pdf_file/0009/355977/Health-Profile-Denmark- Eng.pdf?ua=1
Gallen, Y., Lesner, R. & Vejlin, R. (2019) “The labor market gender gap in Denmark: Sorting out the past 30 years”. Journal of Labour Economics, 56 (1): 58-67
Kleven, H., Landais,C & Søgaard, J. (2019) “Children and gender inequality: evidence from Denmark”. American Economic Journal: Applied Economics, 11(4): 181-209.
Lo Sasso, A., Richards, M., Chou, C. & Gerber, S. (2011) “The 16,819 dollars pay gap for newly trained physicians: the unexplained trend of men earning more than women”. Health Affairs. 30 (2): 193-201.
Luxembourg income studies. (2020) Retrieved April 14, 2020: https://www.lisdatacenter.org Magnusson, C. & Nermo, M. (2020) “The gender wage gap- worst in high-prestige
occupation” Swedish institute for social research.
Meara, K., Pastore, F. & Webster, A. (2019) “The gender pay gap in the USA: a matching study”. Journal of Population Economics, 33 (1): 271–305
Muench, U. & Dietrich H. (2019) “The male-female earnings gap for nurses in Germany: A pool cross- sectional study for the years 2016 and 2012.International Journal of Nursing studies. 2019; 89:125–131
National board of health (2019). Denmark. Retrieved April 16 2020:
https://www.socialstyrelsen.se/en/about-us/
Reeves, M. (2010) Women in business. Routledge, New York.
Statistiska centralbyrån (2018), Sjuksköterskor mer jämställda än läkare. Retrieved 2 April 2020: https://www.scb.se/hitta-statistik/artiklar/2018/sjukskoterskor-mer-jamstallda-an- lakare/
Statistics Denmark (2020), Gender earnings equality in Denmark. Retrieved 18 May 2020:
https://www.dst.dk/en/Statistik/emner/levevilkaar/ligestilling/ligestillingswebsite
Washington center for equitable growth (2017), Occupational segregation in United States.
Retrieved April 5 2020: https://equitablegrowth.org/fact-sheet-occupational-segregation-in- the-united-states/
Wilson, B., Butler, M., Butler, R., & Johnson, W. (2017) “Nursing gender pay differentials in the new millennium”. Journal of Nursing Scholarship, 50 (1): 102-108.
