Normally, free marketeers and those who are especially worried about the efficiency costs of taxation are on opposite camps from those social activists that believe that government intervention is needed to achieve a host of social goals.
Here is a policy proposal that should make the two camps agree: reduce income taxes on women and increase, by less, income taxes on men in a way that holds total tax revenue constant. This policy would simultaneously reduce overall tax distortions and increase women’s participation in the labour force, thus achieving the goals of quotas and affirmative action but in a more efficient way. While quotas impose quantitative constraints that prevent agents from equalizing costs and benefits at the margin, gender-based taxation changes relative prices but lets agents free to optimize at the margin. For those who believe that women should not face discrimination gender-based taxation should be attractive; it is “fair” to compensate women for the fact they bear the brunt of maternity and early child care costs and this harms their career prospects. Gender based taxation offers a form of compensation that helps redress these inequalities in a less distortionary, more transparent and simpler way that affirmative action quotas.
So, how can one achieve this “miracle” solution of reducing the tax distortion while maintaining tax revenue? The answer is well known to any graduate student in public finance. The supply of labour of women is more elastic than that of men, thus it is more responsive to their after-tax wage. Research shows that reduction in taxes increases labour participation of women substantially, but that men’s labour supply is more rigid so a tax rise does not reduce their labour supply much, if at all.1 Ergo, a tax cut on women, teamed with a tax rate hike on men of a smaller magnitude would maintain total revenue and lower tax distortions overall. This is simply an application of the general principle of public finance – the Ramsey rule – that that goods with more elastic supply should be taxed at a lower rate. Every study we know in many different countries finds that the elasticity of women’s labour supply is much higher than that of men. In fact the latter is often found to be close to zero, while that of women is often close to 1.
Gender based taxation is justified not only by gender differences in labour supply elasticities but also by gender differences in the earnings distribution “hazards”. At any given income level Y, the distortions caused by the marginal tax rate t(Y) are proportional to the fraction of subjects who earn Y divided by the fraction of subjects who earn more than Y (which is the hazard of the earning distribution). This because those who earn Y are distorted at the margin by the tax rate t(Y), while, for those who earn more than Y, t(Y) is an infra-marginal tax rate that generates revenues for the government without inducing distortions. Since women have typically higher distributional hazards than men at all earning levels, gender based taxation reduces distortions through this mechanism, independently of elasticities.
Our “back of the envelope” computations available in our paper “Gender based taxation” suggest that the difference in tax rates across genders based upon different labour responses could be quite large. Much more computational work is needed to quantify the welfare gains of this idea, taking into account the complexities of actual tax structures, in particular progressivity, the elasticities of labour supply at different income levels and the heterogeneity of agents. But the initial calculations suggest that this may be really a case of governments leaving many 100 euro bills on the table. Even more so for countries which impose joint family taxation (e.g. US and France), a fiscal system that de facto implies higher tax rates for women than for men.
Even if the optimal tax argument for gender-based taxation is airtight, one needs to worry about whether such a policy conflicts with other social goals. We don’t believe it does. In fact, social activists should favour it. Increasing labour participation of women is an explicit goal of the venerated Lisbon agenda. Even though it does not tell us how, this Agenda sets a target for women participation which is very ambitious for many countries, especially in Southern Europe. Preoccupations with labour-market gender discrimination underlie many policies that involve gender “quotas” or affirmative action. Spain has introduced quotas for women in many areas, Italy is considering the same, and affirmative action is routinely applied in the US. Gender-based taxation may be a more efficient way of achieving the same goals. Since women would obtain higher after tax wages for given pre tax wages it would become cheaper for employers to hire women and to promote them to higher levels of firms’ hierarchies. In other words, discrimination would become more costly. We suspect that gender-based taxation raises more eyebrows than quotas or affirmative action simply because it makes more transparent what are the benefits and costs of such policies geared toward favouring women in the labour force.
Often those who care about women’s work emphasize the policy of supporting it with public child care facilities and other similar services. This kind of policy is typical of a paternalistic European mentality, particularly in the left: instead of giving more resources to the citizens, let’s tax them even more (child care subsidies are not manna from heaven, they require distortionary taxation somewhere) and tell them how to spend their money: public child care facilities, say, rather than privately hired nannies! A higher take home salary for women would allow them to buy more child care at market prices, and, since child care facilities employ mostly women, they would benefit on their cost side as well. Moreover, a subsidy to child care targets women who have children in child bearing age, while discrimination and women labour participation are more general problems than that. Not all countries may want to subsidize fertility directly, and if one wants to be picky one may note that public childcare facilities discriminate against couples who cannot have children. In an event gender-based taxation is not meant to increase fertility, and in principle it can be coupled with policies subsidizing fertility.
In the long run gender-based taxation may contribute to change the traditional division of labour within the family which currently sees men working more in the market and women more at home. If and when that may happen (and many social activists consider that a desirable goal per se), the elasticities of male and female labour supply will become more similar than they are today. At that point one may need to reconsider the gender differences in tax rates, precisely as the basic principles of optimal taxation suggest. But this may take decades to occur, as the recent literature on the economic effects of culture shows.
As for any tax reform this one would have redistributive consequences: there would be winners and losers. But differently than in other welfare improving reforms that are not undertaken because winners cannot compensate losers, in the case of gender-based taxation a large amount of these compensations would take place within households as long as resources are shared. As a result, married men with working spouses might gain as well, so that gender-based taxation has a good chance of making a majority better off.
We are sure that at this point many of those who are still reading are thinking: ok very well but it is simply unfair to fiscally treat men and women differently. We do not believe so. There is nothing more hypocritical than to invoke equal treatment in some areas (taxation) for those who are not treated equally in many other areas (labour market, sometimes family allocation of tasks, child bearing costs, etc). An even deeper philosophical issue is whether women should be compensated because their participation in the labour force is negatively affected for biological reasons, like the potential of bearing children or menstrual cycles that Ichino and Moretti (2007) have found being responsible for two third of the gender gap in absenteeism and 11-13% of the gender gap in wages and promotions.
Bottom line: the idea of gender-based taxation was laying in some relative obscure closet of optimal tax theory as something true in theory and not applicable in practice. We think that this “crazy” idea is actually surprisingly sensible, the more so the more one thinks about it without prejudice.
1 See for example: Blundell R. and T. Macurdy (1999) ”Labor Supply: A Review of Aletrnative Approaches” in O. Ashenfelter and D. Card (eds) Handbook of Labor Economics, Vol 3 North Holland Amstredam