A few centuries back, Europe was under a feudal system. A self-appointed caste of aristocrats was supported - thanks to the labour of the laymen. The system was based on the latter having fewer civil rights than the former; they had to pay high (mostly in kind) taxes, so that the nobility did not have to work for a living. Predictably, at some point such a system was deemed unfair. The people of the Enlightenment abolished the privileges of the nobility and wrote constitutions stating that all citizens were equal before the law.
In recent papers and Vox columns, Alberto Alesina, Andrea Ichino, and co-authors propose to overturn these constitutions and to restore some fiscal privileges based on a birth characteristic: sex.1 That is, they propose that women pay less taxes than men, everything else equal.
This proposal has long been associated with a fringe of radical feminism, so it surprises me to see it coming out of mainstream economics and the academic establishment. In fact, it is becoming so mainstream that Spain’s allegedly conservative Partido Popular has a tax break for female workers in its platform for the next election. Given that the ruling Socialist Party is unlikely to oppose such a “progressive” measure, the Spaniards will have gender-biased taxation whether they like it or not. This will probably remind them of the good old days of Franco.
So why is it that mainstream politicians and academics alike are now endorsing the fringe radical feminist political agenda? Let us start with the economics.
The economic argument
A central argument is based on the so-called “Ramsey taxation principle”, which states that a good whose demand (or supply) is more elastic – i.e. responsive to prices – than another one should be taxed at a lower rate. The reason is that the more responsive the equilibrium quantity of the good is to prices, the more the tax distorts the allocation of resources. Alesina et al. then elaborate on this by applying it to people and argue that (as is shown by econometric studies) women’s labour supply is more elastic to wages than that of men. So if we were to increase taxes for men and reduce taxes for women, the increase in female labour supply would be larger than the fall in male labour supply and GDP would go up. Now, this is not a totally convincing argument because increasing GDP is not a goal in itself. The Soviet concentration camps increased GDP but that does not mean we should have them.2 But instead of looking at GDP you can prove that some utilitarian social welfare measure (i.e. some abstract quantity which supposedly adds welfare across individuals with different identities and preferences) goes up if we allow for “gender-based” taxation. That is, according to this criterion we increase the welfare of women by more than we reduce the welfare of men.
Now, it turns out that if I am maximising any welfare criterion, I can always do better by discriminating than by not discriminating. This is because non-discrimination is a special case of discrimination, where all groups are treated equally. If different groups have different economic behaviour, then to maximise my welfare function I need to discriminate as much as possible, and I will treat each group differently. So we should have different taxes depending on sex, age, race, marital status, city of residence, state of health, and so on.3
At the “optimum”, some of these taxes would sound “right” to those with a “progressive” mind—like the one favouring women proposed by Alesina et al. Others would sound horribly wrong to these same people; in fact the very gender-tax proposal could be implemented as a reduction in transfers to women at the bottom of the distribution of income, since their greater labour supply elasticity implies these transfers are more distortionary than for men.
In other words, we do not learn much by cherry-picking the policy experiment that we analyse in a model to fit a pre-determined political agenda. And if “society” (up to now) wrote down constitutions saying citizens are equal before the law, while our utilitarian social welfare functions say that some should be more equal than others, it is unlikely that these welfare functions represent the social preferences implicit in those constitutions.
Winners and losers
While some dubious measure of aggregate welfare would go up, gender-biased taxation would clearly create losers. The welfare of single men would fall, as would that of married couples where the man has substantially more skills than the woman. It is not possible to make everybody better-off by means of a transfer from gainers to losers. For example, a single man who faces a higher marginal tax rate on his labour income would have to be compensated by a higher lump-sum transfer equal, at the margin, to the increase in the taxes he pays. But it makes no economic sense to increase somebody’s taxes in a distortionary way and compensate them with a transfer. In other words, the compensatory scheme would have to be financed by higher distortionary taxes on women which would eliminate all their gains.
It remains true that there is a case for taxing marginal hours at the household level at a lower rate than infra-marginal ones. But this can be achieved by a gender-neutral reduction in the tax rate on the secondary earner’s hours (which could well also apply to the primary earner’s marginal hours such as overtime). And since such a scheme can make the household better-off by supplying more hours without reducing the total taxes they pay to the government, one can actually leave the household free to choose between that and a more traditional tax schedule.
Such a gender-neutral scheme would be all the more appealing, relative to the gender-discrimination gender-based tax proposed, since the view that men’s labour supply is less elastic than women’s is quickly becoming obsolete. It is associated with a traditional household where women are secondary earners. Only a minority of people are in such a situation. In many countries, for example the US or the UK, more than 50% of adults are single. Furthermore, women earn more than men in a large minority of married couples (30% in Canada). The rapid deterioration of men’s educational achievements relative to women suggests that this figure is going to increase. And, if men do provide lower effort in the university system (perhaps because they are less likely than in the past to have to provide for a family), this suggests that their behaviour can indeed be quite elastic.
Furthermore, Alesina et al. simply assume that the labour supply response to their proposed tax will be identical to what would be obtained if the same change in income was due to some exogenous change in wages. I think that despite being disguised as a tax break for women with no mention of the fact that men will have to pay more taxes to finance it, the Alesina et al. proposal will nevertheless be perceived as unfair by many men who will retreat from the labour market.
For example, there is much talk in France of the so-called “suburbs” and of the riots which regularly happen in these neighbourhoods. Many “progressive” analysts contend that the “youth” who perpetrate those riots do so because they are disenfranchised and have no prospects in the French labour market. Indeed, the youth unemployment rate in such areas is as high as 40%. Another obvious fact is that most of the rioters are male, and it is also well known that women from the same ethnic backgrounds are better integrated in French society and have far better educational achievements. Furthermore, a lot of the unskilled jobs that will be created in the future and that the “youth” could get are overwhelmingly feminised, in areas such as health care for the elderly and child care. Do we believe we will improve the “suburbs” if we tell the male “youths” that should they exit illicit activities and enter the formal labour market, they will have to pay more taxes than their sisters?
Abolishing equality before the law is wrong and dangerous
All these are practical objections which suggest that the supposed gains from gender-biased taxation may turn out to be much smaller than the authors think. But my opinion is that such a proposal is both wrong and dangerous not because of these practical shortcomings but because it is about abolishing equality before the law. The authors do not seem to realise that they are contributing to the decay of our democratic institutions.
The Spanish proposal
Mr. Rajoy, the candidate of the Spanish Partido Popular, did not mention the Ramsey taxation principle. Instead, he said that “our” goal was to achieve “equality” between men and women; ergo, women should pay lower taxes than men until “equality” is achieved. When that goal is reached, the gender-biased tax is supposed to be removed.
By “equality”, Mr. Rajoy cannot possibly mean equality in rights, since he precisely proposes to abolish that. Instead, he is pursuing equality of “outcome”, but even such a phrasing is fraudulent, since a man would be poorer than a woman with the same skills doing the same job at the same firm; because of the tax, women would in fact be paid more than men for equal work. So Mr. Rajoy and other members of the elite who are proposing gender-biased taxes are not interested in equality of outcomes but rather in equality of statistics. They want to equalise the average wage and the average rate of labour market participation between men and women regardless of what people want.
Let us pretend for a minute that politicians are sincerely interested in such a goal. The question is: “How much of our civil rights are they willing to abolish to reach these Stalin-style planning objectives?” Anti-discrimination laws constraining the hiring and pay policies of private firms already exist. Now we are being told that this is not enough and that taxes should discriminate by gender. Given that there always is some statistic which differs between two groups, there will always be some reason for the government to undermine constitutional rights under the pretence of fixing such imbalances.
But it would be very naïve to believe that if ever “equality” is achieved, the discriminatory policies would go away. Who would seriously think that the tax break for women proposed by the PP would be abolished the day they work as much and are on average paid the same as men? The statistics that are supposed to be equalised are carefully picked in accordance with the political agenda of organised interest; instead of the “equalising” policy being removed when the official objective is attained, the focus shifts on some other statistic that justifies maintaining it or introducing new similar ones. A telling example is how the debate on the achievement of women in US universities started focusing exclusively on their under-representation in the hard sciences when they became a majority in virtually all other fields. And we still have not heard proponents of “equality of outcomes” advocating policies that discriminate in favour of men in areas where women are doing better such as education or life expectancy.
Equality before the law as safeguard of liberal democracy
So, what game is being played here? “Equality before the law” was written in constitutions by the people of the Enlightenment not only because they genuinely believed in it, but as a safeguard to prevent democracy from degenerating into tyranny. Absent of individual rights, a majority can impose arbitrary harm to a minority. The constitution defines the rights that the individual has that cannot be overturned by majority rule. Now, given that men and women have many stakes in common, a men vs. women split could seem unlikely. But the decline of marriage and the fall in fertility have reduced these common interests; and the rise of “political correctness” has led us to a situation where men are the only minority that can be played against for electoral gains. So I believe we are in fact entering a dirty – and most dangerous – round of divisive identity politics.
1 See Vox columns 9/1/08, and 8/6/07, and research papers CEPR Discussion Paper 6591 December 2007, and here.
2 See the Vox column on the Gulag as a worker-discipline device by Marcus Miller and Jennifer Smith.
3 For one exploration of this line of thinking, see N. Gregory Mankiw and Matthew Weinzierl, “The Optimal Taxation of Height: A Case Study of Utilitarian Income Redistribution,” December 2007