Gender discrimination lowers output per capita (a lot)

José A. Tavares , Tiago V. de V. Cavalcanti, 16 October 2007

a

A

Gender discrimination is a pervasive phenomenon. All economies and all cultures discriminate between males and females. Wherever one goes, females find it more difficult than males to gain access to the labour market, political power, health and education. Gender discrimination is widespread and appears in many guises, being associated with a given country’s social, cultural, and economic characteristics, both as a consequence and a cause.

According to Hausmann et al. (2006), who present an index of gender inequality in four areas – economic and political participation, and access to health and education inputs – no country has managed to completely eliminate the gender gap. Nordic countries show the best performance in Europe, with Sweden, Norway, Finland, Iceland and Denmark all in the top ten least discriminatory countries. In Asia, the Philippines finds its place in the same admirable segment, while South Africa, at eighteenth, holds the highest position among the African countries surveyed. The United States ranks twenty-third in this ranking. Though gender discrimination tends to decrease over time, the differences across countries are quite substantial, and progress is slow and subject to setbacks. Unwarranted discrimination is first of all an ethical and political issue. Both are good reasons why it is important to analyse its economic cost.

In the labour market, women receive lower pay than men when performing the same tasks offering the same skills, experience, and educational background (see Blau and Kahn (2007)). This so-called “gender wage-gap” discourages female labour force participation, with a direct negative effect on output.1 The lower rate of participation of females is a universal empirical phenomenon. We believe this lower participation is likely to increase fertility and discourage investment in the education of the offspring, namely girls (see Klasen (1999)). Female-to-male wage earnings ratios can range from about 75% in the United States to 41% in Ireland, around 19% in Iran and only 21% in Saudi Arabia.

Discrimination is economically inefficient since it prevents the equalisation of marginal rates of substitution in production. This inefficiency shows up directly in terms of lower wages for women, but it also lowers an economy’s total output. There is a wide literature on the individual cost of gender discrimination in the labour market. Available, “back of the envelope” estimates of the aggregate output cost of gender discrimination point to very significant amounts. For instance, the United Nations estimates that, only in the Asian region, about $47 billion of yearly output are lost every year from a lack of female participation in labour markets, as reported in United Nations (2007).

What has been missing so far is an integrated theoretical and empirical study that assesses the relative impact on aggregate output of the different consequences of gender discrimination on individual behaviour. Our recent paper, “The Output Cost of Gender Discrimination: A Model-Based Macroeconomic Estimate”, aims to provide a relevant contribution in all aspects highlighted above.

We start from a benchmark growth model where families decide their levels of saving, fertility and labour-market participation by women and by men. Couples value both consumption and children, which use up, respectively, financial resources and individual time. Since there is wage discrimination, the model’s outcome is that women participate less in the labour market than their male counterparts, in accordance with empirical findings. We then chose the various parameters (including the gender wage-gap and female labour force participation) to line-up a model on the US economy-base case.

With this model in hand, it is possible to compute the output cost of an increase in gender discrimination – the factor by which women are underpaid relative to men. We find that a 50% increase in the gender wage-gap leads to a 25% drop in per capita income. This decrease in output stems from two distinct effects – a direct decrease in output due to women working fewer hours and an indirect effect due to an increase in fertility. The actual country data shows that countries with higher gender inequality – as measured, for instance, by the Global Gender Gap Report Index – do indeed display lower levels of female labour market participation and higher fertility.

We have compiled independent estimates of the female-to-male earnings ratio for a wide cross-section of countries – rich and developing – and construct a new economy, similar in every way to the US economy, except for the degree of gender discrimination. As an illustration, we use the gender wage-gap implied by Egyptian data as an input to the model economy that, in all other respects, displays the characteristics of the US. By comparing the level of output per capita predicted by our model with the actual output per capita of Egypt, we are able to assess how much of the difference between US and Egyptian income per capita is due to gender discrimination.

We find that, for several countries, a very large fraction of the difference in output per capita relative to the US is due to gender inequality. For countries such as Ireland and Saudi Arabia, wage discrimination actually explains all of the output difference with the US. In other words, were it not for gender discrimination, citizens of Ireland and of Saudi Arabia would be as rich or richer than US citizens. In the case of Egypt, gender discrimination explains about 65% of the difference in output per capita relative to the US. If we impose a constant fertility rate, discrimination explains about 37% of the difference. Here as well as in other economies, the indirect fertility channel accounts for almost half of the total decrease in output per capita. In the table below we present some examples of country income per capita relative to the US in the data and in the model, first with exogenous fertility and then with endogenous fertility. Our analysis suggests that eliminating gender inequality, however difficult it may be, would have a significant positive impact in output that is difficult to match by any other policy option.


Output per capita and Gender Discrimination Model vs. Data

Countries
Data,
United Nations (2005)
Model with Exogenous Fertility
Model with Endogenous Fertility
US
100
100
100
Greece
40.69
85.07
72.01
Ireland
72.60
82.12
66.47
Saudi Arabia
46.05
58.68
32.28
Iran
17.59
70.74
47.19
Egypt
13.20
67.58
42.68
India
7.69
79.84
62.29


Source: Author´s computations.

Remember, this is a first model-based assessment of the aggregate cost of discrimination. There are important aspects of discrimination that we ignore, most notably the role of education. We believe that considering discrimination in education, with its long-run and cross-generational consequences, would substantially increase the aggregate cost of discrimination. Nevertheless, this basic model and our estimates point to costs of gender discrimination which are very substantial and cannot be ignored easily. Decreasing gender-based discrimination would help “rich” nations address the problem of population aging in the near future, by shoring up pension contributions. In developing countries, decreasing gender-based discrimination should feature pre-eminently in any serious macroeconomic policy aimed at increasing output in the long-run.

References

Blau, F., and L. Kahn (2007). “The Gender Pay Gap: Have Woman Gone as Far as They Can? Forthcoming Academy of Management Perspective.

Cavalcanti, T., and J. Tavares (2007), “The Output Cost of Gender Discrimination: A Model-Based Macroeconomic Estimate”, CEPR Working Paper DP6477.

Hausmann, R., L. D. Tyson, and S. Zahidi (2006), “The Global Gender Gap Report 2006”, World Economic Forum, Geneva, Switzerland.

Klasen, S. (1999). “Does Gender Inequality Reduce Growth and Development? Evidence from Cross-Country Regressions,” Policy Research Report on Gender and Development, Working Paper Series # 7.

United Nations (2007), “Economic and Social Survey of Asia and the Pacific 2007”, United Nations Economic and Social Commission for Asia and the Pacific.

Footnotes

1 The gender gap has been the subject of several Vox columns.

Topics: Labour markets
Tags: Discrimination, female labour force, gender inequality

Lecturer in Economics at University of Cambridge
Associate Professor at the Faculty of Economics, Universidade Nova de Lisboa and CEPR Research Affiliate