Fresh ONS labour data error raises questions over trust in UK economic statistics
The Office for National Statistics has admitted a fresh error affecting the collection of UK labour market data, in another setback for the agency as it works to rebuild confidence in some of the country’s most important economic figures.
The issue relates to the Labour Force Survey, the household survey used to produce headline estimates for employment, unemployment and economic inactivity. According to reports, telephone interviewers were mistakenly allocated away from the main Labour Force Survey and towards its planned replacement, the Transformed Labour Force Survey, during May and early June.
The mistake meant that around 1,200 telephone interviews were missed. The ONS has said the problem will not affect the most recent labour market release, which covered the three months to April 2026, but it will reduce the quality of the next release in July.
For most businesses, a survey error may sound like a technical problem. In reality, it matters because labour market data feeds directly into decisions on interest rates, public spending, welfare policy, wage negotiations and business planning.
The Bank of England watches employment, wage growth and unemployment closely when deciding whether inflationary pressure is building or easing. Employers use the same data to understand hiring conditions, staff shortages and pay pressures. Government departments rely on it when designing skills, benefits and employment policy.
That makes confidence in the numbers essential.
What went wrong?
The ONS has said the problem was operational rather than methodological. In simple terms, too many telephone interviewers were directed towards the future replacement survey and too few remained on the existing Labour Force Survey.
The affected period ran from 3 May to 10 June, with some residual effects continuing until 17 June. The ONS said response levels were reduced by 19% in the affected survey waves and by 14% overall.
Because fewer people were interviewed than intended, the ONS will need to rely more heavily on imputation and estimation. That means missing responses will be filled in using statistical methods.
This is not unusual in survey work. All large surveys involve some degree of estimation. But the concern is that higher levels of estimation can reduce the ability of the data to capture real changes, particularly over short periods.
In practice, this means July’s labour market data may be less responsive to genuine movement in unemployment, employment or inactivity than usual. Sharp changes may appear more muted, and policymakers may need to place greater weight on other sources, such as HMRC payroll data, vacancy figures and claimant count data.
Why the timing is awkward
The error comes at a difficult moment.
The UK labour market is already sending mixed signals. The latest ONS release showed unemployment at 4.9% in the three months to April, down from the previous quarter but still higher than a year earlier. Vacancies fell to 707,000, their lowest level since early 2021. Payrolled employees also fell over the year, according to HMRC data.
At the same time, wage growth remained stronger than many economists had expected. Regular pay increased by 3.4% and total pay, including bonuses, rose by 4.4% in the three months to April.
That combination creates a difficult picture for policymakers. Falling vacancies and weaker payroll numbers suggest employers are becoming more cautious. But wage growth still matters for inflation, especially if pay rises continue to feed through into prices.
For the Bank of England, this makes reliable data particularly important. If the labour market is weakening, that may support the case for lower interest rates. If wages remain too strong, policymakers may be more cautious.
A temporary reduction in the quality of labour market data therefore does not just create an internal ONS problem. It adds uncertainty to economic decision-making at a time when the economy is already facing pressure from weak growth, public finance constraints and global instability.
A wider problem of confidence
The latest error is especially damaging because it follows several years of scrutiny over the quality of UK labour market statistics.
The Labour Force Survey has been under pressure since response rates fell sharply after the pandemic. Lower response rates make survey data more difficult to interpret because the sample may become less representative of the wider population.
The ONS has been working to improve the existing survey while also developing the Transformed Labour Force Survey, which is intended to modernise data collection and improve reliability. But running the old and new systems alongside each other is costly and complicated.
The organisation has said it is managing more than 594 field interviewers and most of its 139 telephone interviewers across the two surveys, covering samples of more than 118,000 households per month. That scale helps explain the operational challenge, but it also underlines why strong systems and controls are essential.
There is a balanced point here. The ONS has identified the error, disclosed it publicly and explained that the impact will be temporary. That transparency is important. Mistakes in complex statistical systems can happen, and early disclosure is better than allowing users to rely on data without warning.
However, the fact that the error happened at all will raise questions about management, resourcing and oversight. If official statistics are to command confidence, users need to believe not only that errors will be corrected, but that preventable errors are being reduced.
Why businesses should care
For businesses, the issue is not just about national statistics. It is about the reliability of the information used to make decisions.
A retailer considering whether to expand its workforce needs to understand consumer confidence, wages and employment conditions. A manufacturer setting pay budgets needs a clear view of labour shortages and wage pressure. A recruitment business depends on vacancy trends. A lender or investor watches employment data for signs of household stress.
If labour market data becomes harder to interpret, businesses may need to be more cautious and draw on a wider range of indicators.
That means looking beyond the headline unemployment rate. Payroll data, vacancy trends, sector-specific recruitment surveys, wage settlements, consumer spending, insolvency data and internal business intelligence may all become more important.
In other words, the ONS error is a reminder that even official data should not be read in isolation. Good management relies on triangulation, especially when the economy is uncertain.
The Bank of England’s dilemma
The Bank of England has previously expressed concern about weaknesses in UK labour market statistics because the data plays a central role in assessing inflationary pressure.
If unemployment is rising and vacancies are falling, that suggests the labour market is loosening. If wages are still rising strongly, that suggests inflationary pressure may persist. The balance between those two signals matters for interest rates.
A data quality warning makes that balance harder to judge.
It does not mean the Bank of England will ignore the figures. Nor does it mean July’s numbers will be useless. But it does mean policymakers are likely to read them with more caution and compare them closely with other sources.
That could make interest rate decisions more cautious, especially if other economic indicators are also mixed.
The case for patience
It would be easy to treat the latest ONS error as evidence of a broken system. That would be too simplistic.
The ONS is managing a difficult transition from an old survey model to a new one, while trying to improve response rates and maintain regular publication of high-profile data. The agency has also been more open about data quality limitations than many users were used to seeing.
There is value in that openness. A statistical agency that admits uncertainty is not necessarily weaker than one that hides it. In many ways, transparency is part of restoring trust.
But transparency alone is not enough. Businesses, markets and policymakers need reliable data, consistent processes and confidence that quality issues are being reduced over time.
The ONS now needs to show that it has learned from the error, strengthened its operational controls and maintained momentum in improving labour market statistics.
A technical mistake with real-world consequences
This episode may have started with an internal allocation error, but its significance is broader.
The UK economy is at a point where every signal matters. Employers are cautious, vacancies are falling, wage growth remains important for inflation, and interest rate expectations are finely balanced. In that environment, uncertainty over labour market data has consequences.
The immediate impact may be temporary. July’s figures will carry a health warning, and later releases should be less affected. But the wider issue is confidence.
Official statistics are part of the economic infrastructure of the country. They guide policy, shape market expectations and influence business decisions. When that infrastructure is weakened, even briefly, it makes decision-making harder.
For the ONS, the priority is clear: fix the process, explain the impact and rebuild trust.
For businesses, the lesson is equally clear: use official data, but do not rely on one number alone. In uncertain conditions, the best decisions are made by reading several signals together.
