The gender pay gap is a worldwide challenge that has persisted despite political will and interventions. Comparably qualified women performing similar work as men continue to earn less. There are conflicting views in the literature regarding the status of the gender pay gap.
The purpose of the study was to determine status of the gender pay gap among employees in the same salary band and to establish whether men and women receive similar pay for similar work in the study population.
The status of the gender pay gap would establish the progress made towards closing the gap and guide necessary adjustments to interventions.
A quantitative analysis was conducted on the pay information of 217 902 employees collected in a survey from over 700 companies, across 10 job families and 6 industries.
Men’s pay was consistently higher than that of women in all salary bands except at the 75th and 95th percentile in sub-bands B-lower and B-upper and 25th percentile in sub-band E-upper. The gender pay gap ranged from 8% in band A to 27.1% in sub-band F-upper. The gaps observed in the salary bands were statistically significant (
Government’s efforts seemed to have produced minimal results as women are represented in all job families, industries and salary bands. The pay of men and women in senior and top management levels was similar. However, more still needs to be done to achieve the 50% target representation of women in senior management and close the gap at all levels.
The number of women at management levels is still very low when compared to their male counterparts. However, the gender pay gap in senior to top management positions are converging towards similar pay for work of similar value.
The pay of some women remains less than that of men, even when the women have the same level of education or perform work of the same value as men (Bosch,
There are contrasting reports on the status of the gender pay gap in South Africa. Steyn (
The gender pay or wage gap refers to the difference between men’s pay and women’s pay as a percentage of men’s pay (New Joint Committee for Higher Education Staff (JNCHES),
Previously, the gap was attributed to women not being suitably qualified in terms of education and work experience (Fransen et al.,
The current status of the gender pay gap is not certain, making it very difficult to chart the way forward and measure progress towards the goal of gender pay equity for work of equal value. Hence, the proposed study will focus on determining the status of gender pay gap in a sample of the South Africa population.
The purpose of the research study was to establish the status of the gender pay gap among employees doing similar work, with the hope of generating evidence that could lead to policy revision and strengthening of the interventions being implemented. This was achieved by interrogating the pay information of employees in salary bands and investigating the effects of associated independent variables on income. The research questions were as follows:
What is the gender pay gap in salary bands or among employees performing similar work?
What is the effect of selected variables (e.g. gender, race, job family and industry) on the gender pay gap in the study sample?
The literature review was focused on development of the key concepts of the gender pay gap and interventions aimed to close the gap. The literature review is in four parts including gender evolution, the root cause of the gender pay gap, an overview of gender pay gap and responses to the gender pay gap.
‘Gender’ refers to societal roles (e.g. men as the head of families and breadwinners and women if married as the primary caregiver to their husbands, children and extended family members), while ‘sex’ refers to a person’s biological classification. Gender is related to dynamic power relationships between men and women (Bendl & Schmidt,
Zebracki (
The terms ‘gender sensitivity’ and ‘gender mainstreaming’ have become very important, because of the unique needs of women and men. There is a distinct advantage to gender mainstreaming over approaches proposed and implemented previously (Zebracki,
Gender inequality is as old as human existence (Bussin & Nienaber,
Gender inequality and the pay gap favours men (Steyn & Jackson,
Low human capacity development among women, relative to that of men, is one of the reasons postulated for women earning less than men (Addabbo & Favaro,
An article published by Aláez-Aller et al. (
Bhorat and Goga (
Casale and Posel (
Gender equality as a policy objective has gained universal acceptance (Witkowska,
Blau and Kahn (
The research objective was achieved with descriptive study design and quantitative analysis.
The pay information used was obtained from 21st Century (Pty) Ltd, one of the largest reward consultancies in Africa. The pay information was obtained in a survey from respondents in over 700 companies, spread across 10 job families in 6 industries. The job families included compliance and risk; executive management; financial and accounting; human capital; information technology; logistics and procurement; marketing and sales; operational; secretarial; and technical and specialist. The industries represented in the study were extractive; transformative; distributive services; producer services; social services; and personal services. Descriptive and inferential analysis was conducted on the pay information to establish status of gender pay in the salary bands. Also, the effect of variables – race, gender, job family and industry – on the guaranteed package was assessed with a regression analysis.
The sample consisted of women and men, 18 years and older, employed in the formal sector, mostly made up of companies in the private sector, which had participated in the survey conducted by 21st Century. The database consisted of 264 660 respondents’ pay information. However, analysis was conducted on 217 902 of the respondents, who had a complete set of information.
The research is a cross-sectional or snap-shot study of the difference between pay earned by men and women employed in the formal sector. This is because the data used represent the pay situation at a point in time. Quantitative methods were used for this research study, because it was aimed at answering the question ‘how much’, which is best answered using quantitative methods (McCusker & Gunaydin,
The unit of analysis was gender pay, as the aim was to establish the status of the gender pay gap and describe the extent of the gap. ‘Pay’ or ‘income’ refers to the ‘guaranteed package’ of the respondents. The guaranteed package is the basic salary plus the benefits received by an employee per year. This includes car benefits, which could be in form of the cost to company for providing a company car or a car allowance plus reimbursement. Other benefits include housing, low-cost loans, club fees, professional fees, subscriptions, cell phone allowance, computer allowance and other similar benefits (Bussin, Nicholls, & Nienaber,
The pay information was received in a Microsoft Excel spreadsheet and Stata 14.1 (StataCorp,
The Shapiro–Wilk test of normality established that the guaranteed packed distribution was heavily skewed. Also, the data did not achieve normality after various forms of transformation processes had been performed. This implied that a parametric test, like the
The gender pay gap was calculated at several points in the guaranteed package distribution within each salary band to provide a detailed picture of the status and the trend of the gap. The points were p10, p25, p50, p75 and p95 (This refers to the percentiles in a pay range, i.e. 10th percentile, 25th percentile, 50th percentile, 75th percentile and 95th percentile). The salary bands contained six categories: A, B, C, D, E and F. The Peterson rule was applied in merging sub-grades of A to form band A and sub-grades of bands B to F to form lower and upper sub-bands. The pay brackets or salary bands were established based on the number and complexity of decisions associated with the job. The respondents were allocated to the band that fitted their level of responsibility. Band A corresponds to employees with basic skills, band B to semi-skilled, band C to skilled, band D to middle management, band E to senior management and band F to top management (Steyn & Jackson,
The analysis was conducted on the pay information of the 217 902 respondents with complete information. This consisted of 75 373 (34.59%) women and 142 529 (65.41%) men. When disaggregated by race, it consisted of 141 465 (64.92%) black people, 44 181 (20.28%) white people, 21 140 (9.70%) people of mixed race and 11 116 (5.10%) Asian or Indian people (
Population frequency by gender and race.
Race | Gender |
Total | |
---|---|---|---|
Female | Male | ||
Black African | 45 483 | 95 982 | 141 465 |
White | 16 178 | 28 003 | 44 181 |
Mixed race | 8931 | 12 209 | 21 140 |
Indian | 4781 | 6335 | 11 116 |
Population frequency by salary band.
Paterson salary band | Freq. | Per cent | Cum. |
---|---|---|---|
A | 12 483 | 5.73 | 5.73 |
BL | 65 043 | 29.85 | 35.58 |
BU | 25 741 | 11.81 | 47.39 |
CL | 54 578 | 25.05 | 72.44 |
CU | 26 692 | 12.25 | 84.69 |
DL | 23 993 | 11.01 | 95.70 |
DU | 5942 | 2.73 | 98.43 |
EL | 2476 | 1.14 | 99.56 |
EU | 670 | 0.31 | 99.87 |
FL | 228 | 0.10 | 99.97 |
FU | 56 | 0.03 | 100.00 |
Freq., frequency; Cum., cumulative.
A, basic skill level; BL, semi-skilled level; BU, semi-skilled supervisory level; CL, skilled level; CU, skilled level supervisory; DL, middle or tactical management level; DU, middle or tactical supervisory level; EL, senior management level; EU, senior management supervisory level; FL, top management; FU, top management supervisory level.
Population frequency by salary band and gender.
Paterson salary band | Gender |
Total | |
---|---|---|---|
Female | Male | ||
A | 3669 | 8814 | 12 483 |
BL | 17 270 | 47 773 | 65 043 |
BU | 9 413 | 16 328 | 25 741 |
CL | 23 512 | 31 066 | 54 578 |
CU | 9790 | 16 902 | 26 692 |
DL | 8955 | 15 038 | 23 993 |
DU | 1879 | 4063 | 5942 |
EL | 684 | 1792 | 2476 |
EU | 157 | 513 | 670 |
FL | 42 | 186 | 228 |
FU | 2 | 54 | 56 |
A, basic skill level; BL, semi-skilled level; BU, semi-skilled supervisory level CL, skilled level; CU, skilled level supervisory; DL, middle or tactical management level; DU, middle or tactical supervisory level; EL, senior management level; EU, senior management supervisory level; FL, top management; FU, top management supervisory level.
The guaranteed package in the study sample ranged from R550.00 to R7 690 888.00. Men’s pay was consistently higher than that of women in band A, sub-bands B-lower and B-upper (except at the 75th and 95th percentiles for both sub-bands), sub-bands C-lower and C-upper, sub-bands D-lower and D-upper, sub-bands E-lower and E-upper (except at the 25th percentile) and sub-bands F-lower and F-upper.
Results of the rank-sum test indicated that the gender pay gap is statistically significant with a
Assessment of guaranteed package trend per salary band.
Salary band | Gender | Mean | p10 | p25 | p50 | p75 | p95 | |
---|---|---|---|---|---|---|---|---|
A | Female | 94 911.6 | 43 004.0 | 61 556.0 | 103 998.0 | 122 781.0 | 133 752.0 | < 0.0001 |
Male | 104 184.9 | 47 814.0 | 81 968.0 | 113 039.5 | 132 042.0 | 141 555.0 | ||
B-lower | Female | 134 903.0 | 54 000.0 | 81 888.0 | 132 685.0 | 171 014.0 | 260 027.0 | < 0.0001 |
Male | 144 054.9 | 76 220.0 | 124 782.0 | 139 643.0 | 168 625.0 | 234 762.0 | ||
B-upper | Female | 206 535.2 | 123 420.0 | 157 797.0 | 195 596.0 | 254 824.0 | 312 983.0 | < 0.0001 |
Male | 214 043.7 | 144 412.0 | 177 736.0 | 202 514.0 | 251 829.0 | 310 123.0 | ||
C-lower | Female | 292 408.9 | 164 851.0 | 229 051.0 | 300 000.0 | 352 766.0 | 445 200.0 | < 0.0001 |
Male | 324 748.5 | 190 000.0 | 260 641.0 | 329 092.0 | 384 791.0 | 488 301.0 | ||
C-upper | Female | 436 404.1 | 291 865.0 | 377 582.0 | 425 265.0 | 503 770.0 | 633 248.0 | < 0.0001 |
Male | 476 040.4 | 338 542.0 | 396 825.0 | 460 845.5 | 557 771.0 | 678 162.0 | ||
D-lower | Female | 626 769.6 | 390 156.0 | 477 000.0 | 618 030.0 | 751 579.0 | 954 000.0 | < 0.0001 |
Male | 690 669.3 | 432 000.0 | 524 160.0 | 680 883.5 | 828 863.0 | 1 048 244.0 | ||
D-upper | Female | 941 562.6 | 642 037.0 | 767 634.0 | 942 004.0 | 1 084 458.0 | 1 360 000.0 | < 0.0001 |
Male | 1 025 837.0 | 708 000.0 | 861 553.0 | 1 022 314.0 | 1 179 870.0 | 1 440 358.0 | ||
E-lower | Female | 1 283 874.0 | 866 838.0 | 1 046 208.0 | 1 270 383.0 | 1 474 791.0 | 1 813 135.0 | < 0.0001 |
Male | 1 372 581.0 | 946 099.0 | 1 133 239.0 | 1 341 645.0 | 1 566 815.0 | 2 000 004.0 | ||
E-upper | Female | 1 804 075.0 | 1 195 498.0 | 1 521 900.0 | 1 791 000.0 | 2 100 000.0 | 2 515 261.0 | 0.1227 |
Male | 1 888 222.0 | 1 260 000.0 | 1 520 402.0 | 1 836 400.0 | 2 193 800.0 | 2 845 515.0 | ||
F-lower | Female | 2 298 798.0 | 1 474 280.0 | 1 808 603.0 | 2 351 606.0 | 2 815 338.0 | 3 382 656.0 | 0.3131 |
Male | 2 427 518.0 | 1 600 000.0 | 1 980 601.0 | 2 400 565.0 | 2 841 652.0 | 3 710 000.0 | ||
F-upper | Female | 2 937 805.0 | 2 370 375.0 | 2 370 375.0 | 2 937 805.0 | 3 505 234.0 | 3 505 234.0 | 0.0851 |
Male | 4 240 381.0 | 3 031 657.0 | 3 392 604.0 | 4 087 615.0 | 4 888 798.0 | 7 057 320.0 |
The number of women to men decreased in bands B-lower, C-upper and from D-upper to F-upper (see
Trend of ratio of women to men by salary band.
Furthermore, there was increased representation of men from D-lower to F-upper, which corresponded to respondents in middle management, senior management and top management, respectively (Twenty-first-(21st)-Century-Pay-Solutions,
The chi-square test showed statistically significant association between
Regression model of annual income, gender, race, job family and industry.
The model predicted that men’s guaranteed package was higher than that of women by an average of R15 125.33. The gender reference point was
Also, the predicted average earning of respondents in a job category relative to
The model further predicted guaranteed package by industry using
This section is presented based on the research questions:
What is the gender pay gap in salary bands or among employees performing similar work?
The investigation of equal pay for work of equal value and representation of women relative to men in salary bands revealed that there was an increase in the proportion of women relative to males in semi-skilled jobs, skilled jobs and the lower band of middle management jobs, while there was an increase in the proportion of men relative to women in the upper band of middle- to top management jobs. These findings support the observation of Gayle, Golan and Miller (
The analysis of the gender pay gap within the salary bands was done from the point of view that respondents within a salary band performed similar duty or tasks. The statistically significant higher income of men compared to women in salary bands A, B, C, D and E-lower indicated that respondents who performed similar work were not similarly compensated. Among other factors, the significant difference in their income could be explained by discrimination and by occupational segregation.
The pay gap at the bottom and top of the pay distribution, corresponding to the unskilled and management levels, respectively, suggest the presence of a sticky floor and glass ceiling. The phenomena of the sticky floor and glass ceiling are associated with the gender pay gap at the bottom and senior levels, respectively (Addabbo & Favaro,
In the present study, men and women in salary band F and sub-band E-upper were fairly compensated for similar work despite the pay gap. This could be explained by the presence of women with a high-level of education, which is known to narrow the gender pay gap (Aláez-Aller et al.,
What is the effect of gender, race, job family and industry on the gender pay gap in the study population?
The statistically significant association of income with gender and race could be explained by a persistence of discrimination against women, occupational segregation and women’s relatively lower accumulated human capital development. The proportion of men in the study sample is not consistent with the country’s demographic of 51.3% women and 48.7% men reported in the 2011 census. The under-representation of women in the study population, compared to their representation in South Africa’s population further point to presence of discrimination (Folke & Rickne,
Furthermore, the confirmed sustained under-representation of women in employment is a source of inequality. According to Steyn (
Although females now have unrestricted access to education and human capacity development, their under-representation may be a result of the fact that currently the talent pool from which to select women to meet government’s target of 50% representation of women in employment, especially at management level, is too small. This view is supported by the findings of the research published by Kahn and Motsoeneng (
The study sample was similar to South Africa’s population in term of ethnicity representation, as it consisted of black people in the majority and Asian people the least. It differed in the number of white people, which was twice that of people of mixed race. According to the 2011 census, South Africa’s population consisted of 79.2% black people, 8.9% white people, 8.9% mixed race people and 2.5% Asian people (Stats,
The statistically significant association of income with job family and industries suggests the presence of occupational segregation in the study sample. Although there is a statistically significant association between income and gender, race, job family and industry, the model could not explain 56.0% of the variance in annual income. This may be because of variables such as education, work experience and marital status missing in the model (Steyn,
Although the sample size was big, the study findings should be generalised with caution because the extent to which the study sample represents South African population is not certain. Moreover, bias because of missing values could not be completely ruled out.
There is a need to conduct a survey with sample size powered for findings generalisable to the South African population. Application of mixed methods should be useful for gaining a comprehensive understanding of findings. The status of the gender pay gap in job family and industry needs to be studied in detail.
The government empowerment programme for women seems to be yielding promising results. Women are represented in all job categories including in senior to top management positions. This observation supports the findings of the study by Kahn and Motsoeneng (
As the gender pay gap persists widely, all existing policies and legislation need to be reviewed and revised where necessary. Implementation programmes for the legislation need to be developed and actively monitored for compliance. According to Kahn and Motsoeneng (
Government intervention to empower women seems to be yielding promising results. Women are represented in all salary bands or job categories. However, the gender pay gap is present in all salary bands and is statistically significant except at salary bands E and F, corresponding to the senior management and top management levels, where the pay gap is converging to similar pay for similar work. Furthermore, the number of women is smaller compared to their male counterparts in all salary bands, most especially in senior management and top management positions. The annual income is dependent on gender, race, job family and industry. The gender pay gap can be attributed to discrimination, occupational segregation and educational or skills level. There is need to review the existing policies and legislation. The implementation of the policies and legislation also needs to be reviewed and possibly revised. There is need to consider developing a new policy to cater for the dual roles of women as a caregiver and a wage worker. Finally, a broad-based approach that caters for the needs of both men and women, with relevance to race, is needed to ensure the effectiveness of interventions in order to close the gender pay gap.
The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this paper.
A.A. was responsible for the project design and analysis. M.H.R.B. made intellectual and conceptual contributions. A.A. and M.H.R.B. co-wrote the manuscript.