Do women and minorities face glass ceilings in employment?

Should economics be able to explain divorce rates amongst married couples? Normally, that is something that would be left to the techniques of psychology and anthropology. But surprisingly enough, Gary Becker won the Nobel Prize in Economics in 1992 and one of his most celebrated papers was precisely in explaining divorce.

What enables economists to venture into such unlikely terrain? It is the legitimizing power of their tools of trade: And in this case, it was the legitimizing power of carefully examined data. Scrutinizing data for the hidden stories is an important pastime of economists.

Scrutinizing some of the employment data in Sri Lanka, in the public and private sectors also seems to reveal a hidden story: This time it is not about divorce but about discrimination.

Are women discriminated in employment?

On this question, the data is somewhat worrying and begs a question: Is there a hidden story of gender discrimination?

Women constitute half the population in Sri Lanka but only 34 percent of the labour force. Women are also less likely to be employed than men. Unemployment amongst women is more than twice as high as unemployment amongst men. As a result, when looking at total employment in the economy, men account for 67 percent and women for 33 percent: A ratio of two to one in favour of men. That itself is not conclusive on the state of discrimination.

Women are likely to be opting disproportionately to work as homemakers but statisticians don’t categorise that as employment and the related preferences could perhaps explain why women may be finding it harder to be employed.

The worrying part comes when we examine the representation of women in top rungs of organisations, where they are already employed.

Verité Research tediously gathered data on the gender and ethnicity of the top rungs of the public sector in 2012 and the other data referred to in this article comes from published sources and the Central Bank reports.

Women in public sector

The employment ratio of two to one in favour of men increases dramatically in the top two tiers of the public sector. In Tier 1 (the top tier) of the public sector it becomes six to one. In Tier 2 (second highest tier) of the public sector it is still five to one.

Adjusting for the fact that employment ratios are two to one in favour of men, here is another way to make sense of this data: Men in the public sector are three times as likely to climb to the top as women in the public sector. That is a very serious disparity in promotional prospects and such large disparities do beg the question of discrimination.

Women in private sector

But the very top of the private sector does significantly worse than the public sector. Looking only at the top 100 private companies (by revenue) and just focusing on the position of CEO, it turns out that only 5 percent of them are women. That is, given the two to one starting disparity in employment, a man in the top 100 companies is about 10 times more likely to end up as its CEO as is a woman.

It is not possible to dismiss these kinds of large disparities as statistical noise. The data requires an explanation and the possibility of gender-based discrimination needs serious scrutiny.

Are ethnic minorities suffering the same fate as women in employment?

The question raised by the data on the employment of ethnic minorities is very similar to the question raised by the data on the employment of women.

Starting with the population ratios as a base, very approximately (and rounding things off to whole numbers) the ratio of the Sinhala to Tamil population is about four to one and the Sinhala to Muslim population is about seven to one (using approximate percentages of 73, 18 and 7 for Sinhala, Tamil and Muslim respectively).

While on the question of gender discrimination, the available data was more incriminating of the private sector, on the question of ethnic discrimination in employment, the available data is more incriminating of the public sector.

Ethnic minorities in public and private sectors

Tamil representation: Figure 2 shows that with regard to ethnic representation, the Tamil representation in the university admissions and in the CEOs of the top 100 private firms is just about the same as in the population (about 18 percent). But when it comes to Tier 2 of the public sector this falls to under 6 percent (creating a 16 to one ratio in favour of Tier 2 being Sinhalese over Tamil), and in Tier 1 (the top tier) it is 3 percent (creating an almost 32 to one ratio in favour of the top Tier being Sinhalese over Tamil).

Taking into account the larger population representation of the Sinhalese, the disparity can be explained in this way. Sinhala workers in the public sector are four times as likely to climb to Tier 2 and eight times as likely to climb to Tier 1 as their Tamil counterparts.

Muslim representation: Figure 2 shows that with regard to Muslim representation, the ratio entering universities is in fact higher than the ratio in the population, in relation to Sinhalese. A 10 to one population ratio between Sinhalese and Muslims reduces to an eight to one ratio in university admissions. But this falls to a ratio of 16 to one against the Muslims, when looking at the CEOs of the top 100 private firms. In short, the data poses somewhat of a challenge to the standard stereotyping of the Muslim population as being more successful in business than in education.

In the top rungs of the public sector too, the Muslim representation is lower than in university admissions and in the population. Both in Tier 1 and Tier 2 of the public sector the ratio of Sinhala to Muslim workers is over 30 to one. It means that a Sinhala worker in the public sector is three times more likely to climb to the top two tiers than their Muslim counterparts.

Looking under numbers

Do the numbers unearth a hidden story of discrimination – do women and ethnic minorities in Sri Lanka encounter a glass ceiling that prevents them climbing the ladder of employment in the private sector and the public sector, in relation to men and Sinhala workers? The available data makes a strong case for being concerned. The issue clearly requires further discussion and analysis.

(Verité Research provides strategic analysis and advice for governments and the private sector in Asia)