No sane person these days believes in the labour theory of value. Even John Roemer, an economist sympathetic to Marxism, has described it as “not a useful empirical theory.”

But there’s one place where this silly idea is still believed – in the Equal Pay Act, the law requiring that women get equal pay for work of equal value.

The Equal Opportunities Commission summarizes the law thus:

The EPA describes equal value in terms of the demands of jobs. It gives ‘effort skill and decision’ as examples of particular demands…What matters are the demands made on employees by their work, rather than the value of the work to the employer.

This is absurd. What matters to an employer is the value of that employee to him – or more precisely, her marginal product. Output matters, not inputs.

Think of it this way. A recurring, if under-emphasized, theme of studies of the labour market is that observable skills account for only a small proportion of the variation in wages. R-squareds on Mincer equations are often less than 25 per cent; see for example, this paper or this one. Many apparently identically qualified people have unequal wages.

How should we react to this?

The neoclassical economist says such differences reflect differences in unobservable skills, or other differences that yield different marginal products. From this perspective, the Equal Pay Act is just grossly inefficient managerialism. It’s the pretence that experts can second-guess markets.

However, you don’t have to believe this to believe the Equal Pay Act is silly. There are, at least, four problems with it.

1. It’s arbitrary. Let’s say differences in wages are due to luck. An unlucky woman then has a claim for compensation under the EPA, but an unlucky man doesn’t. I suspect this is the case for some of the high-profile cases brought by women in the City. Salaries there vary enormously from individual to individual, often without any obvious link to merit.

2. Women are not the only victims of discrimination. Wages differ according to height (pdf), looks and whether one has a tidy home. But the EPA doesn’t protect short ugly untidy people from discrimination.

3. The Act imposes a mindless equality. If a woman proves her work is more valuable than a man’s, she’s not entitled to higher pay than him. The EOC says: “Tribunals cannot award higher pay for work of greater value.”

4. The Act doesn’t ask how discrimination is possible.

Consider a tribunal’s recent decision that nurses were discriminated against because they were paid less than medical technicians, despite having comparable skills. Why, if this is the case, would anyone want to become a nurse rather than a technician? It’s not as if the pay gap has only just emerged.

One obvious answer is that there are non-wage advantages to being a nurse. You get job satisfaction. You get respect; nurses used to be called angels, whereas lab technicians rarely are. And nurses can have lots of choice about the type of work they do and where they do it. These compensating advantages might explain the pay gap between nurses and technicians. In one sense, they must do so – why else would someone choose to train for a low-paid job when they could train for a higher-paid one?

Feminists might reply that women’s preference for lower-paid but more rewarding work is the result of women being indoctrinated into gender roles at a young age.

But is this really true? Lots of men have this preference – me, for one; I left the City to take a lower-paid but more rewarding job in journalism. And even if the feminists are right, so what? Are people really entitled to compensation because they would be unhappy if their preferences were different?

Feminists, of course, have another reply - that women are, obviously, paid less then men. But this alone is not proof of discrimination.

One of the best recent investigations of this question is this huge pdf, published by the EOC. It found that women who work full-time earn 18 per cent less than male full-timers.

However, much of this gap is due to factors other than discrimination. Men have more education than women (1.4 percentage points of the gap); they have more experience (3.5 percentage points); they work in larger firms; and so on.

The report found that only seven percentage points of the 18 per cent pay gap was a pure gender effect. But even this is not proof of discrimination. It might just reflect differences in preferences and motivation. It is, the authors say: “the upper limit on the component of gender discrimination that is direct discrimination.”

That said, there might also be indirect discrimination. Women who have dropped out of work to care for children subsequently have lower earnings, even controlling for the direct loss of work experience; there seems to be a wage penalty for childcare. But it’s easy to imagine that this penalty might be due to lower marginal products, rather than to employers’ ripping off women.

But let’s say I’m wrong here, and that women – and women alone – are systematically paid less than their marginal product.

If this is the case, it can only be the result of monopoly or monopsony power. In a textbook competitive market any employer who tried to pay a woman less than her marginal product would find her lured away to a higher-paying firm.

True systematic discrimination, therefore – as distinct from either justifiable or random wage differences – can only be due to market imperfections. As Gary Becker has said, competition is a friend of minorities.

In this sense, feminists should support a free market economy. So why don’t they?