The Gap That Will Not Close
The UK median gender pay gap has hovered between 13% and 15% for roughly a decade. Six years of mandatory gender pay gap reporting have not shifted the number in any meaningful direction. Three decades of equal pay litigation have not shifted it either. The gap is not closing because the legal framework was built to address a version of the problem that — for most of the modern economy — is no longer the dominant one.
In October 2023, Claudia Goldin won the Nobel Prize in Economics for explaining why. Her answer has uncomfortable implications for anyone who advises on equal pay claims under the Equality Act 2010.
Greedy Jobs
Goldin's central finding — set out most accessibly in her 2021 book Career and Family and developed across decades of historical labour market analysis — is that pay in many high-earning sectors does not scale linearly with hours worked. Work twice the hours and you earn more than twice the salary. She calls these "greedy jobs": roles in finance, law, consulting, and senior management where sustained availability is rewarded disproportionately.
The mechanism is not complicated. Clients and employers value continuity, responsiveness, and the ability to be present when the work demands it — evenings, weekends, and at short notice. A solicitor who takes a call on a Sunday, or a banker who turns around a model overnight, commands a premium that two part-time workers splitting the same total hours cannot replicate. The work is not fungible. A relationship with a client is not divisible by two.
The result is that pay follows a curve, not a line. The steep part of the curve rewards those who can commit to long, inflexible hours. The shallow part accommodates everyone else. The curve is not created by discriminatory intent. It is created by the structure of client demand and, in many sectors, by the economics of human capital that is difficult to substitute.
At lower hours, the two curves are close together — the greedy-job premium barely registers. At full availability, the curves diverge sharply. That divergence is the pay gap Goldin identifies: not a product of discrimination, but of how the market prices sustained, inflexible commitment. The woman who reduces her hours after having children moves left on the chart. Her linear-pay colleague in pharmacy loses income proportionately. Her greedy-job colleague in law or finance loses it exponentially.
Practitioners will recognise this immediately. The self-employed Bar is arguably the worst case. Earnings are entirely availability-based: they track willingness to accept returns at short notice, travel to distant tribunals at unsocial hours, and remain continuously available during long trials. The cab-rank rule means you cannot curate a caseload around school hours. Cashflow is lumpy and unpredictable — fees arrive months or sometimes years after the work is done — so you need a constant pipeline of new instructions just to stay solvent. There is no statutory maternity pay beyond the basic allowance, and the habitual delay in payment means a career break creates an income void that extends well beyond the break itself. Chambers rent is typically tapered — expenses come from the first slice of your gross fees — which means that a barrister working part-time pays a higher effective rate than a full-time colleague generating the same hourly income. The financial architecture of self-employment at the Bar penalises reduced availability at every turn.
The Bar Council's own data bears this out. Its 2015 report, Snapshot: The Experience of Self-Employed Women at the Bar, found that women make up 50% of those called to the Bar but only 35% of the practising Bar, 12% of self-employed KCs, and just 6% of self-employed women with more than 22 years' call — compared to 3,346 men at that level. The Biennial Survey found that 57% of women at the Bar with children were primary carers, compared to 4% of fathers. The employed Bar was 46% female; the self-employed Bar, 33%. And the Momentum Measures research concluded that "on current trends the practising Bar will not ever achieve gender balance." The pipeline starts at parity and bleeds women at every stage — with the sharpest drop after parenthood. As one participant in the Bar Council's focus groups put it: "The job is simply not compatible with motherhood."
And because barristers are self-employed, the entire equal pay framework is irrelevant to them. There is no employer. There is no comparator. There is no s.66 equality clause to imply. A woman at the Bar who earns less than her male colleague of equivalent call because she cannot sustain the same availability after having children has no equal pay claim at all — not a weak one, but literally none. The framework does not apply to her. The Bar is Goldin's thesis in its most concentrated form: a profession where the greedy-job premium is extreme, the child penalty is acute, the structural penalties for part-time work are baked into the financial model, and the legal protections are entirely unavailable.
The Child Penalty
Before parenthood, Goldin's data shows that the earnings gap between men and women with equivalent qualifications is relatively narrow — in some sectors, negligible. After parenthood, it opens sharply. In most economies, it does not close again.
She calls this the "child penalty." Women who take on primary childcare responsibilities — and they overwhelmingly do — cannot consistently meet the demands of greedy jobs. They move to part-time arrangements, take career breaks, or shift into roles with more temporal flexibility. Those roles sit on the shallow part of the pay curve. The gap is not caused by an employer paying a mother less than a father for the same work. It is caused by the mother no longer doing the same work — or no longer doing it on the same terms.
The Comparator Problem
Sections 64 to 80 of the Equality Act 2010 contain the equal pay framework. The architecture requires a comparator: a real, named individual of the opposite sex, employed on like work, work rated as equivalent, or work of equal value, at the same establishment or one with common terms (s.64, s.65, s.79). Where such a comparator exists and is paid more, a sex equality clause is implied into the claimant's contract (s.66).
This framework was built for a specific mischief — the same work, remunerated differently on grounds of sex. It is effective against that mischief. What it cannot easily reach is the structural pay architecture Goldin describes, because the Goldin-type gap does not arise from a woman doing job X being paid less than a man doing job X. It arises from women being less likely to be doing job X at all after parenthood — or from the pay structure of job X being designed around a temporal commitment that motherhood makes difficult to sustain.
Consider a concrete example. A female associate at a City firm returns from maternity leave and moves to an 80% schedule. She is paid 80% of her full-time male colleague's base salary — strictly pro rata. But he earns a bonus of 60% of base; she earns 15%. The bonus rewards sustained availability: client pitches at short notice, late closings, weekend drafting sessions. On paper, the pay structure is sex-neutral. In practice, she earns roughly half his total compensation for the same seniority and the same hourly rate. She has no viable like-work claim because her terms are different. She may have an equal-value claim, but the forensic exercise of establishing equal value across different working patterns, with different bonus metrics, is formidable.
The Material Factor Defence
Assuming a claimant can identify a comparator and establish like work or equal value, the employer's defence is the material factor under s.69: the pay difference is due to a factor which is not the difference of sex, and which is a material difference between the claimant's case and the comparator's.
If the factor does not put women at a particular disadvantage relative to men, that is the end of the analysis. The employer need not justify it further. In Rainey v Greater Glasgow Health Board [1987] IRLR 26, the House of Lords confirmed that recruiting from the private market at higher rates is a legitimate material factor, provided it is not a sham for sex discrimination. In Strathclyde Regional Council v Wallace [1998] ICR 205, the House of Lords held that where a material factor is genuinely sex-neutral and not indirectly discriminatory, there is no obligation to show objective justification.
But if the factor does put women at a particular disadvantage, the analysis changes. The employer must then show that the factor is a proportionate means of achieving a legitimate aim (s.69(2)). That is the indirect discrimination test, transplanted into the equal pay setting. And it is the point at which Goldin's research becomes directly relevant.
An employer whose pay structure rewards sustained, inflexible availability — where bonuses, promotions, and partnership tracks are calibrated to total hours and on-demand responsiveness — is operating a system that puts women at a particular disadvantage, because women disproportionately bear the childcare burden that makes those hours unsustainable. The doctrinal tools for challenging this have existed since Bilka-Kaufhaus GmbH v Weber Von Hartz [1986] ECR 1607, where the ECJ held that excluding part-time workers from an occupational pension scheme required objective justification because part-time workers were overwhelmingly female. Bilka is 40 years old. Goldin's contribution is to show, with granular economic data, just how pervasive and persistent the mechanism it identified really is.
The difficulty is getting to that analysis. The s.69(2) proportionality question only arises if the claimant first clears the hurdles of identifying a comparator, establishing like work or equal value, and demonstrating that the material factor relied on by the employer has a disparate impact. Each of those steps is contested, expensive, and slow. In the greedy-jobs context — where the pay differential is built into the structure of compensation rather than into a simple hourly rate — clearing those hurdles requires expert evidence, statistical analysis, and a tribunal willing to look past formal equality of terms to substantive inequality of outcomes.
The Blocked Route
A practitioner reading this might ask: why not bypass equal pay entirely and bring a standalone indirect sex discrimination claim under s.19? The PCP — requiring sustained availability or penalising reduced hours — puts women at a particular disadvantage. The analysis seems straightforward.
It is blocked. Section 70 of the Equality Act 2010 provides that a sex discrimination claim relating to a contractual term that is (or could be) the subject of an equal pay claim must proceed under the equal pay framework, not under s.19. The pay term is channelled. You cannot route around the comparator requirement by reframing a pay complaint as indirect discrimination.
The Part-Time Workers (Prevention of Less Favourable Treatment) Regulations 2000 offer a parallel route, but a narrow one. A part-time worker can claim less favourable treatment compared to a comparable full-time worker doing the same or broadly similar work for the same employer. That covers pro-rata pay, access to training, and other terms. It does not, however, reach the non-linear premium that is the core of Goldin's thesis — the fact that the full-time worker earns disproportionately more, not just proportionately more. Pro-rata equality is exactly the formal equality that masks the structural gap.
The Statistical Route
The most promising doctrinal avenue is the statistical approach from Enderby v Frenchay Health Authority (Case C-127/92; [1994] ICR 112). The ECJ held that where statistics show a significant pay difference between two groups doing work of equal value — one predominantly female, the other predominantly male — the burden shifts to the employer to explain the difference by reference to objective, sex-neutral factors.
Enderby is the closest the equal pay framework comes to engaging with Goldin's structural analysis. It allows aggregate patterns to do evidential work without requiring proof of discriminatory intent. If an employer's workforce data shows that women cluster in the lower-paid roles and men in the higher-paid ones, and if an equal value assessment can bridge the two groups, the employer must justify the gap — and "the market demands this level of availability" is not self-evidently proportionate if the availability requirement is what produces the gendered distribution in the first place.
But Enderby was developed in the context of collective bargaining — separate negotiating structures producing separate pay scales for speech therapists and pharmacists — not individual hours-based differentials within a single profession. Extending its logic to the greedy-jobs context is analytically defensible but practically demanding. The equal value exercise alone can take years.
Where This Leaves Us
Goldin's research is a diagnostic, not a cause of action. It explains why the gender pay gap is so persistent and so resistant to litigation: the mechanism that sustains it — the non-linear premium for unbroken availability — is not the mechanism the equal pay framework was designed to catch.
The framework catches explicit pay discrimination within workplaces. Goldin's data suggests that this is now a relatively small component of the overall gap. The larger component is structural: how certain jobs are designed, how hours and availability are remunerated, and who can sustain the commitment those structures demand after becoming a parent. The gender pay gap reporting regime tells us what Goldin's framework predicts — consistent median gaps, often wider than mean ones, with women concentrated in lower-paid roles — but imposes no obligation to close the gap it reveals.
For practitioners, the framework matters in two ways. First, when resisting a material factor defence in a higher-earning sector: if an employer relies on an availability-based or hours-based pay differential, and the workforce data shows a significant gender split between those who achieve the premium and those who do not, there is a serious argument that the factor carries a disparate impact and requires objective justification under s.69(2). Bilka established the principle. Goldin supplies the economic evidence to support it.
Second — and this is the harder point — as a reality check on the limits of case-by-case adjudication. The gap Goldin identified will not yield to individual claims brought by individual claimants against named comparators. The architecture of the law was built for a different problem. Whether it needs to be rebuilt is a question for Parliament, not tribunals. For now, the honest answer is that UK employment law can see the gap clearly enough. It just lacks the tools to close it.