Analysis of the Problem of Discrimination against Women in the Banking Sector - Research Paper Example

Is there Women discrimination in the banking industry This inform about discrimination problem is based on data of a certain bank. This bank provides information about relevant variables such as gender, employment category, education level, previous experience in moths, beginning salary, current salary, minority classification, age in months, and months since hire. We have based the discriminatory analysis based on the next premise. Women discrimination, in relation to our research, exists when the difference in the beginning salary and current salary between men and women is statistically significant, after controlling for employment category, educational level, previous experience, months since hire, minority classification and age in months. Considering the above premise, we have run different statistics analyses. Firstly, in order to have an idea of raw data, we have run descriptive statistics of the above variables presented. Secondly, we run ANOVA models, comparing beginning and current salaries between men and women in order to search for statistically significant differences. Thirdly, in order to go deeper in the analysis of discrimination, we use multiple regression analysis (MRA). Using MRA, we intend to get aware about the real differences of salaries between men and women after controlling for a number of variables. Finally, we conclude the report with discussion of the findings. Below are presented the descriptive statistics for the variables used in the analysis. Gender. As it can be seen in table 1, 54,4% of the employees in the bank are men. Table 1: Frequencies of Gender variable Frequency Percent Valid Percent Cumulative Percent Valid male 258 54,4 54,4 54,4 female 216 45,6 45,6 100,0 Total 474 100,0 100,0 Employment Category. Table 2 shows that 76, 6% of the employees are clerical, 5,7% - custodial and 17, 7% are managers. Table 2: Frequencies of Employment Category variable Frequency Percent Valid Percent Cumulative Percent Valid Clerical 363 76,6 76,6 76,6 Custodial 27 5,7 5,7 82,3 Manager 84 17,7 17,7 100,0 Total 474 100,0 100,0 Descriptive statistics of table 3 shows that the mean educational level of the workers in the sample bank is nearly 14 years. The mean previous experience of the employees is almost 96 months. The mean age is roughly 35 years. The beginning salary is $17,016 and the average current salary is $34,419. Table 3: Descriptive statistics for quantitative variables N Minimum Maximum Mean St. dev. Educational Level (years) 474 8 21 13,5 2,885 Beginning Salary 474 $9,000 $79,980 $17,016.09 $7,870.638 Current Salary 474 $15,750 $135,000 $34,419.57 $17,075.661 Months since Hire 474 63 98 81,11 10,061 Previous Experience (months) 474 0 476 95,86 104,586 age in month 473 240,62 744,61 412,7001 141,34965 Valid N (listwise) 473 In table 4, we can see the distribution of men and women across the different employment categories in the bank. We can observe that in the clerical category the participation of men and women is similar. However, more women are employed as clerical. For the custodial and managerial category, the majority of the employees are men. Table 4: Gender and Employment Category Employment Category Total Clerical Custodial Manager gender male Count 157 27 74 258 % within Employment Category 43,3% 100,0% 88,1% 54,4% female Count 206 0 10 216 % within Employment Category 56,7% ,0% 11,9% 45,6% Total Count 363 27 84 474 The results of ANOVA analysis are presented in table 5. We can observe that there are statistically significant differences between the beginning and current salaries of men and women. The average beginning salary of men is $20,301, and the average beginning salary of women is $13,091. For current salary the averages are $41,441 for men and $26,031 for women. The two averages are statistically significantly different. The F statistic for the beginning salary shows that the difference between men and women is statistically higher than that of current salary. However, with the ANOVA analysis we cannot get insight about the variables that account for these differences. Accordingly, we build up various regression models for understanding the women discrimination in the bank. Table 5: Difference between the salaries of men and women in the bank N Mean Std. Deviation Std. Error Beginning Salary male 258 $20,301.40 $9,111.781 $567.275 female 216 $13,091.97 $2,935.599 $199.742 Total 474 $17,016.09 $7,870.638 $361.510 Current Salary male 258 $41,441.78 $19,499.214 $1,213.968 female 216 $26,031.92 $7,558.021 $514.258 Total 474 $34,419.57 $17,075.661 $784.311 ANOVA Sum of Squares df Mean Square F Sig. Beginning Salary Between Groups 6110780779,007 1 6110780779,007 124,376 ,000 Within Groups 23190124186,448 472 49131619,039 Total 29300904965,454 473 Current Salary Between Groups 27918533028,833 1 27918533028,833 119,798 ,000 Within Groups 109997962407,508 472 233046530,524 Total 137916495436,340 473 Our main regression model is related to exploring the differences in the beginning salary between women and men. The model is presented below: Beginning salary = f (gender, employment category, education level, previous experience, minority, age) In order to run regression models, dependent variable has to follow a normal distribution. We present below (see figure 1) the normal curve of the beginning salary variable considered for the analysis. Figure 1: Normal distribution of beginning and current salary As we can see in figure 1, beginning salary does not follow a normal distribution. Therefore, we transformed them in Logarithmic variable (hereafter Log variable). Additionally, as employment category variable is a multinomial variable, we transform it as Dummy variables. Therefore, the models we will test are: E(Log Beginning salary) = b0 + b1gender + b2clerical + b3managerial + b4education level + b5previous experience + b6minority + b7age + b8ageeduc + error In order to get insight about the variables that influence the beginning salary, we first run a stepwise method of regression analysis. Model of Beginning Salary: Method of Stepwise As shown in the model summary table of stepwise method of regression, we can observe that model 8 is the one that best fit. Its R-square is 73.5% and it is statistically significant at 99% of confidence level. According to model 8 of stepwise method, the variables that explain log beginning salary variable are: gender, square educational level, educational level, minority classification, previous experience, square previous experience, age, and square age. In that sense, the final and statistically significant model of Log beginning salary after running stepwise regression analysis is: E(Log Beginning salary) = b0 + b1gender + b2square educational level + b3educational level + b4minority classification +b5previous experience + b6square previous experience + b7age + b8squage +error After we have found the variables that account for log beginning salary variable, we search for which of that variables explain beginning salary when employees have the employment category of "Managers" and when those are "Clerical". Accordingly, we run two regression models, one for managers and the other for clerical employees. Additionally, we employ Enter method of regression. Model for beginning salary of Managers: Method of Enter As we can see in managers model summary, r-square is 60% and it is statistically significant different from zero at 99% of confidence level. The beginning salary of managers is accounted for by gender at 99% and minority classification at 90%. The effect of gender on beginning salary means that women managers earn1 79% of men managers' salary. Moreover, minority classification indicates that managers that belong to minority classification receive2 121% salary of that of managers that do not belong. Model for beginning salary of Clericals: Method of Enter As we can observe in clericals model summary, r-square is almost 55% and it is statistically significant different from zero at 99% of confidence level. The beginning salary of clericals is explained by gender, square educational level, educational level, minority classification, previous experience, square previous experience, age, and square age at 99% of confidence level, except for age in month at 95%. The effect of gender on beginning salary means that clerical women earn3 87% of men clericals' salary. Additionally, educational level variable denotes that as a clerical employee has one additional month of experience his/her salary increases in $0.914. Moreover, minority classification indicates that clericals that belong to minority classification receive5 91.2% of salary of clericals that do not belong. Summarizing our work, we have designed two steps in order to get insights about women discrimination in a selected bank. The first step concerns to select the variables that explain beginning salary of employees through stepwise method of multiple regressions. After running it, we found that gender, square educational level, educational level, minority classification, previous experience, square previous experience, age, and square age, affect beginning salary. After that, in the second step, we run two regression models, one related to explaining beginning salaries of managers and the other to accounting for that of clericals. We found that, for beginning salaries of managers, gender and minority classification explicate their beginning salaries. However, with respect to salaries of clericals, we came across that not only gender and minority classification explain beginning salaries but also square educational level, educational level, previous experience, square previous experience, age, and square age. As regards the model fit, both models are statistically significant at 99% of confidence level. Moreover, in relation to explanatory power, the model of managers has an R-square equals to 60% and the one of clerical model equals to 55%. Consequently, both models reveal validity and consistent evidence. Furthermore, the clerical model explaining the differences in beginning salary has better fit, since its F statistics is bigger than the one of the manager model. Its F is 54,421. Our conclusion, as revealed by the multiple regression models, is that, there is a discrimination of women not only for beginning salary of managers but also for that of clericals. Read More