Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
The UK statistics agency has underestimated the country’s gender pay gap for the past two decades, according to a new report that raises further questions about the reliability of key economic data.
Figures released by the Office for National Statistics since 2004 have consistently under-reported the gender pay gap by a margin of about 1 percentage point, according to research in the British Journal of Industrial Relations.
The findings, published on Monday, point to flaws in the Annual Survey of Hours and Earnings (Ashe), an economic survey used by policymakers to set the national minimum wage and benchmark public sector pay.
The ONS is under scrutiny from politicians and the Bank of England over its failure to fix faulty labour market figures, which are a critical input into interest rate decisions.
The agency has been forced to delay a number of critical economic statistics over quality concerns, with retail sales data for July postponed earlier this week.
Alex Bryson, co-author of the research paper and professor at University College London, said small private companies had been “disproportionately absent” from the survey over the past 20 years and that failing to adjust for this had led the ONS to underestimate the gender wage gap.
Response rates for Ashe dropped below 50 per cent after the Covid-19 pandemic and were consistently under 70 per cent for decades before that, despite a mandatory requirement for employers to respond.
The official figure for the gender pay gap was 7 per cent in 2024, but Bryson said the findings implied the true figure could be closer to 8 per cent.
Alongside an official review of the survey’s methodology, Bryson called on the government to adequately resource the ONS and for greater enforcement of mandatory employer participation.
“The bottom line is there is no real substitute for getting employers to respond in the first place,” he added.
Professor Jonathan Wadsworth, senior research fellow at the Centre for Economic Performance at the London School of Economics, said the problems with Ashe were symptomatic of a “wider malaise” in UK statistics.
“We are going to be in this realm of uncertainty for the short to medium term, which is not the ideal environment for policymakers,” he added.
The paper also found that sample biases led to a higher estimate of average earnings, implying the share of jobs paid at or below the minimum wage was underestimated by about a fifth.
Changes to the methodology for Ashe resulted in a late upwards revision to the official estimate for average earnings at the end of last year, leading to a larger than anticipated increase in the UK’s minimum wage this April.
A government-commissioned review by former civil servant Sir Robert Devereux in June warned of “deep-seated” issues inside the ONS, with poor budgeting leaving many areas under-resourced and “a reluctance, at senior levels, to hear and act on difficult news”.
Responding to Monday’s paper, the ONS said the survey’s sampling and weighting was under review and it had improved representation of higher earners in the 2024 cycle, which will have addressed some of the bias identified in the research.
“Weighting schemes are just one aspect of the methodology of a complex survey such as Ashe, and we regularly scrutinise these methods to ensure they continue to be relevant and aligned to best practice,” it added.