A new equality bill will shortly be put before parliament. Harriet Harman, the equality minister, indicated that redressing the gender pay gap would be at the bill’s heart. In a curious phrase she spoke of “empowering the resentful” – not the deserving, or the poor, but the resentful.

In April last year, men working full-time earned 17% more per hour than women working full-time. This is seen as evidence of a systematic bias against women. Employers are held to blame. In a poll earlier this year, 88% said stronger government action is necessary. Are they right? Should “the resentful” be empowered?

Let’s put this in perspective. The pay gap has fallen, is likely to fall faster and could even go into reverse. Women outperform men at university, just as girls do better than boys at school. Female employment is concentrated in the expanding service sector. Women are choosing to have children later, have fewer of them and return to the labour market faster. These are all factors likely to improve their relative earning power. Excellent.

What accounts for the gender pay gap? Not discrimination. For one thing, you find differences within male and female populations that employer prejudice can’t explain. As an example, although married men earn more than married women, single women earn the same or, as they get older, more than single men. There are differences between ethnic groups. Black Caribbean women earn slightly more per hour than black Caribbean men, while Bangladeshi women earn a quarter more than Bangladeshi men. Or consider sexual orientation: gay men earn more per hour on average than “straight” men, while lesbians earn more than heterosexual females. How does that fit the view that labour markets are riddled with discrimination? These pay differentials arise partly from differences in the jobs people do. Few Bangladeshi women work: those who do are well educated and so have jobs where they earn more than the typical male, a third of whom work in restaurants. Gay men are relatively highly educated and concentrated in a narrow range of well-paying jobs.

The work men do often carries disadvantages which higher pay compensates for. Men are 1½ times more likely to be made redundant than women and 2½ times more likely to suffer a serious injury at work. Males are more likely to work outside and to work long and unsocial hours. They commute further.

Career preferences differ: 83% of all undergraduates studying subjects allied to medicine are women; so are three-quarters of vets and teacher trainees. But women are only 15% of engineering students and 22% of computing students. When it comes to first employment, therefore, choices are different. Of women’s top 25 preferred graduate jobs, 12 are in the public or voluntary sectors, where high pay is unusual, as against just four of men’s top 25. Women typically look for a lower first salary than men expect. They are less likely to attempt to negotiate over pay, less likely to seek promotion, less likely to change jobs in pursuit of higher pay. These choices are difficult for governments to influence.

Aren’t we worse than other countries? True, the UK pay gap is above the European Union average. But this misleads. We have a large gap because lots of women work. Italy has a gender pay gap half ours, but its overtaxed and overregulated labour market creates few opportunities for women to work: only the more highly skilled (and thus better paid) can find jobs.

It isn’t appreciated what a crude statistical artefact the gender pay gap is. For instance, after the Berlin Wall came down, the gap in Eastern Germany fell by 10 percentage points. Hurrah. But this was because women’s employment, at one of the world’s highest levels in the GDR, collapsed as its economy imploded. Perhaps Harman would have us be like Bahrain? That country has a pay gap of about 40% – in favour of women. Very few women, only the educated members of elite families, are in paid work.

Few policies advocated to reduce the pay gap – compulsory pay audits, use of government procurement, enhanced flexible working, big childcare subsidies – seem likely to succeed. Most have been tried elsewhere: nowhere has pay been equalised as a result. They are, however, costly to implement.

Even if they succeeded in reducing the gap, this would be at the expense of greater inequality between households. Over the past 15 years there has been an increasing polarisation between “work-rich” (two-earner) and “work-poor” (no-earner) households. Increased female pay would increase the financial advantage of two-earner households and raise conventional measures of inequality and relative deprivation. Another group of “disempowered resentful” would be created.

If the new bill sparks real debate – perhaps even questioning the appropriateness of hallowed legislation such as the Equal Pay Act – it could open up questions about the possibility and desirability of complete pay equality. Alas, I fear it will merely be further legislative flimflam. We should feel resentful about that.

Professor J R Shackleton’s Should We Mind the Gap? Gender Pay Differentials and Public Policy is published by the Institute of Economic Affairs

Times Recommends

• The ghost of Robin Cook haunts Chilcot’s feast
• The truth shouldn’t need oaths
• Violence: let’s separate the men from the boys