UK statistics office reports error in May and June job data collection
Operational Error and Its Impact on Labour Market Data
By David Milliken
Background of the Labour Force Survey
LONDON, June 19 (Reuters) - Britain's statistics office said on Friday that an operational error in May and early June has led to reduced response rates for its main survey of unemployment and the labour market.
The ONS' Labour Force Survey suffered from a big fall in response rates during the COVID-19 pandemic, triggering concern from the Bank of England which relies on the data and a broader, critical government review of the statistics agency last year.
Recovery and Planned Overhaul
Since then, response rates for the survey had climbed back to close to pre-pandemic levels while an overhauled version of the survey is due in 2027.
Details of the Recent Error
However, on Friday the ONS' director-general for economic statistics, James Benford - who transferred to the body from the BoE as part of reforms - said mistakes had been made in data collection in May and early June.
Consequences for Data Quality
"There will be a level of reduced quality for our labour market statistics in July, with a smaller effect on the subsequent releases," he said in a statement.
Reduction in Survey Responses
Too few interviewers had been allocated to working on the LFS compared to its upcoming replacement, leading to a 19% reduction in responses in the survey waves affected, and 14% overall, the ONS said.
Unaffected Data Periods
Data released on Thursday for the three months to April, which showed a small, unexpected dip in unemployment to 4.9%, is unaffected.
Timeframe and Ongoing Effects
The main effect of the error has been to lower responses between May 3 and June 10 and some residual effects persisted until June 17, Benford said.
ONS Response and Next Steps
As a result, the ONS will need to impute or estimate more data, which was likely to lead to less of a change in the figures from previous months than might otherwise have been the case.
"We will work to quantify the impact here and will report on it transparently with the next labour market release, including any effects on bias," Benford said.
(Additional reporting by Muvija M and Sam Tabahriti; editing by William James)
