Of economic theory known such a thing as discrimination in the labor market . Discrimination on the labor market is the practice of hiring , which leads to differences in wages for equally productive workers . In the presence of wage discrimination against certain groups of workers is lower than that of other groups doing the same work with the same qualifications. Entrepreneurs often accused of discrimination on grounds of race , sex, age , disability , religious beliefs or ethnic origin .
Discrimination affects the functioning of labor markets . It generates differences in equilibrium wages , which do not appear in the result of these different income groups of the marginal product of their labor and the marginal cost of labor of the workers themselves . Discrimination does not recognize the equality of opportunities for certain groups of workers to find a job.
It is hard to believe that women and still earn less than men, but it is. On average, the difference between the salaries ranged from 13 to 23 percent. And this average statistics is worldwide, regardless of their country of residence and level of education.
Incomes Data Services ( Data Service Revenue ) based on the socio- available statistical data of Eurostat , the International Labour Organization and Sites Fund WageIndicator ( indicator salary) conducted a study on the difference between the salaries of 63 countries , 30 of which - European .
Here's a brief report on the study :
The average difference in wages between men and women around the world is 15.6 percent. In Europe , Oceania and Latin America, it is slightly smaller than in Asia and Africa . On the situation in Asia and Africa, while much less information available.
Registered salary data site WageIndicator, which filled about 400 000 people , showed that the difference in the salary survey participants ranged from 13 to 23 percent.
According to the study , the level of education of women the same as men , or even higher . However, getting a woman of higher education does not necessarily lead to a reduction in the wage gap . Nevertheless , the study showed that the level of women's education , still affected by this difference.
The salary gap between men and women are traditionally more in occupations that are considered "feminine" - health, education , social. While in traditionally "male" occupations , this difference is less.
Especially significant difference is observed in industries such as mining , services and financial sector.
Public administration , social activities and the provision of individual services show the smallest difference in salaries between men and women .
Data Set Construction
We use a fictitious sample data for this project. As data selected average wages to men and women in the same positions in a certain large corporation. All other data affecting the wages are considered equal. The sample consists of 30 data on salaries of men and women's wages - a total of 60 observations in the two samples.
The data is given in the table below:
The purpose of the narrative (descriptive) statistics is the processing of empirical data, their classification, visual representation in the form of graphs and tables, as well as their quantitative description by the main statistical indicators.
Descriptive Statistics: Men’s monthly wage; Women’s monthly wage
Variable N N* Mean SE Mean StDev Variance Minimum
Men’s monthly wage 30 0 2911,3 64,9 355,4 126301,0 2199,0
Women’s monthly wage 30 0 2434,0 36,7 201,1 40421,4 2111,0
Variable Q1 Median Q3 Maximum
Men’s monthly wage 2687,0 2848,0 3176,3 3556,0
Women’s monthly wage 2297,3 2413,5 2626,3 2879,0
As we may see, the average (mean) value of men’s wage (2911.3) is higher than women’s (2434.0), but the men’s data is much dispersed (the standard deviation is 355.4 compared to the women’s 201.1).
Let’s construct frequency histograms of the given data.
Both data data sets are left-skewed, the data is not bell-curved.
Now construct the confidence intervals for the population means for both data sets. Confidence interval is the term used in mathematical statistics for interval (as opposed to point) statistical evaluation of parameters, preferably with a small volume of sample. Confidence interval is called, which covers the unknown parameter with a given reliability.
We do it in Minitab:
Variable N Mean StDev SE Mean 90% CI
Men’s monthly wage 30 2911,3 355,4 64,9 (2801,1; 3021,5)
Women’s monthly wage 30 2434,0 201,1 36,7 (2371,6; 2496,3)
Variable N Mean StDev SE Mean 95% CI
Men’s monthly wage 30 2911,3 355,4 64,9 (2778,6; 3044,0)
Women’s monthly wage 30 2434,0 201,1 36,7 (2358,9; 2509,0)
Variable N Mean StDev SE Mean 99% CI
Men’s monthly wage 30 2911,3 355,4 64,9 (2732,5; 3090,1)
Women’s monthly wage 30 2434,0 201,1 36,7 (2332,8; 2535,1)
The last column represents the confidence intervals (CI) for population means for men and women with 90%, 95% and 99% level of significance.
Hypothesis Testing and Conclusion
Now it is time to perform the statistical test for comparing means between the data samples to check our hypothesis about that the men’s average wage is significantly higher than women’s average wage.
We do this in Minitab using 2-sample Student’s t-test. This test is one-tailed.
Two-Sample T-Test and CI: Men’s monthly wage; Women’s monthly wage
Two-sample T for Men’s monthly wage vs Women’s monthly wage
N Mean StDev Mean
Men’s monthly wage 30 2911 355 65
Women’s monthly wage 30 2434 201 37
Difference = mu (Men’s monthly wage) - mu (Women’s monthly wage)
Estimate for difference: 477,3
90% CI for difference: (352,1; 602,5)
T-Test of difference = 0 (vs not =): T-Value = 6,40 P-Value = 0,000 DF = 45
As we can see, the observed t-score is 6.40 with the significance p-vaue of 0.000 (interpreted as “Lesser than 0.001”), hence, we can reject the null hypothesis about the mean equality at all levels of significance (1%, 5% and 10%) and state, that there is a significant difference between the average wage values for men and women.
Our expectation about labor discrimination in wage levels seems to be true.