Box 1: Progress towards gender equality in wages, where do we stand?
Gender parity in pay is proving hard to achieve. Pay differentials between men and women are a persistent form of gender inequality in the workplace and the Global Gender Gap Index 2020 finds that the progress towards closing the gender gap on this aspect has stalled. No country (including the top-ranked ones) have yet achieved gender parity in wages. Given that women continue to be less rewarded than men in the workplace, it is important to assess the extent of this difference in monetary terms and if there has been some progress looking at a longer time series, at least in some countries.
To answer these questions, it is first necessary to define how wage gaps are measured and what they include. To start, recall that wage gaps refer only to differences in pay of employees, and therefore they do not take into account differences in men’s and women’s revenues due to non-employment contracts. For instance, revenues from corporate profits or from financial assets are not considered. Since there are less women than men among entrepreneurs or investors, and non-salary revenues are higher than wages, income gaps (which include all types of revenues) tend to be larger than wage gaps. According to the estimate provided by this report, income differences are quite large: the global average of woman’s income is about $11,000 (in Purchasing Power Parity, PPPs) while the average income of a man is $21,000 (in PPPs).
The wage gap is somewhat smaller, but different measurements deliver different results. Four metrics are available for cross-country comparison:
- The OECD’s gender wage gap is defined as the difference between male and female median wages divided by male median wages. Wages are computed for full-time equivalent dependent employees and are expressed in Purchasing Power Parity terms.1
- The ILO’s gender median and mean pay gaps are two indicators computed either as the difference between the median wages of men to that of women; or as the difference between the average wages of men and women. In both cases ILO uses hourly wages.2
- The ILO’s factor-weighted gender pay gap is a corrected version of the measures above. Simply put, the raw mean and median wage differences are corrected (using econometric analysis) by four factors: education levels, age, working time (full-time versus part-time) and status (private-sector versus public-sector employment).
- The World Economic Forum’s wage equality for similar work is derived from the Executive Opinion Survey, a questionnaire answered by business leaders in over 140 countries. The respondents are asked: “In your country, for similar work, to what extent are wages for women equal to those of men?” (1 = not at all, significantly below those of men; 7 = fully, equal to those of men). These individual answers are then aggregated, and the resulting figures are converted into 0–1 scores, where 1 stands for equal pay between women and men, working in a similar position.
Each approach has advantages and disadvantages; however, it is important to be aware that to correctly measure wage gaps one needs to control for: i) incidence of part-time and hours worked: since a relatively high share of women is working part-time, wage gaps may be partially due to working fewer hours; ii) concentration in professions where salaries are higher: a relatively high share of women is working in occupations that are less well-paid then men, which of course affects the calculation of average and median wages by gender; iii) concentration in senior roles: since salaries of managers and senior professionals are higher than those in operational positions and there are fewer women in senior roles, not considering this aspect leads to over-estimating wage gaps.
Economic theory suggests that, absent frictions or distortions, wages should simply remunerate the productivity of a worker. However, this is not the case in the job market; hence, to correctly measure gender wage gaps, one should be able to disentangle productivity differentials (i.e. differences in skills), from outright gender discrimination (i.e. the part of wage differential only to a woman with a similar set of skill and same role is offered a lower salary than a man) and from other types of frictions than in turn impact wage differentials (i.e. discrimination in promoting women in senior roles).
In this respect, while OECD and ILO unweighted wage gaps have the merit to produce a quantitative measure of the monetary difference between salaries of men and women, they do not isolate the difference in pay only due to gender bias. They therefore capture overall differences in wages due to all elements that cause women to be disadvantaged in the workplace.
The ILO factor-weighted gender pay gap is a more refined estimate of differences due to gender biases, but it is still not comparing exactly wages of men and women in the same positions. In addition, this measure requires a large statistical collection effort and therefore is produced in discrete points in time in each country. As such, continuous time series are not available for most countries.
The World Economic Forum’s Executive Opinion Survey score—although it is based on perceptions and does not allow for a monetary quantification of the wage gap—compares gender wage gaps for similar roles and therefore aims at isolating the gender bias, excluding frictions in the labour market structure that lead women to be employed in different roles than men.
Figure 1.1: Wage gap between women and men, OECD countries, 2010-2019
Having clarified what each measure captures, it is possible to compare trends in wage gap measures based on statistical data (OECD gender wage gaps) and based on Executive Opinion Survey data. Time series of wage differentials based on “hard” data are only available for OECD countries. Figure 1.1 shows trends in average wage gaps in this group of countries.
According to OECD data, the differential in men’s median income and women’s median income is about 13.5%. This estimate is somewhat smaller than the ILO’s factor-weighted median gender pay gap, which is about 15%.
Looking at the trend, the average wage gap in OECD countries is closing but at a very slow rate. As shown in Figure 1.1, it was 4.5% a decade ago and is now 13.5%, and it has therefore reduced by 1 percentage point in 10 years. This direction is consistent with Executive Opinion Survey trends, which finds that in OECD countries, the wage equality for similar work has increased by approximately 2 percentage points in 10 years.3 Both trends are encouraging, but too slowly.
Further, as time series for the indicator “wage equality for similar work” are available continuously for over 100 countries, it is possible to compare wage gap trends in OECD countries and all other countries.
Figure 1.2: Wage equality between women and men, OECD and non-OECD countries, 2010–2019
Figure 1.2 shows that while in OECD countries gender equality is improving, albeit slowly, in the rest of the world, on average equality is worsening. As a consequence, the negative average trend observed in non-OECD countries over-weights progress achieved in high-income (OECD) countries. This clarifies why global progress towards closing the wage gap has stalled.
Further research is needed to fully explain why wage gaps are widening (on average) outside OECD countries, yet the analysis allows to conclude that: first, gender gap in wages are still large and women’s wages can be estimated to be about 15% lower than those of men; second, in OECD countries wage gaps are closing but the progress is too slow; third, more efforts are needed to remove all barriers that prevent women to attain similar economic opportunities as men, especially in emerging and developing countries.