The Office of National Statistics (ONS) plays a vital role in the UK by describing in cold hard numbers what is happening in our economy and society. We should not have to worry about the reliability of their figures. But that is what is being suggested now – not just by a few economic commentators – but by the Bank of England, the Office for Budget Responsibility and the Treasury Select Committee.
At the heart of the problem is the key survey used to measure what is happening to employment and within the labour market more generally – the Labour Force Survey (LFS). During the pandemic, face-to-face interviews which were used to generate the LFS data were suspended and replaced by telephone interviews. This change, and a variety of other changes which took place in the pandemic and the following recovery, affected response rates to the survey and the reliability of its results. The ONS has downgraded the status of the employment data from the LFS from its gold standard accreditation, ‘national statistic’, to ‘experimental’ to reflect these issues.
The ONS has a programme of work to address the deficiencies in the LFS but Ian Diamond, head of the ONS, admitted this week that this project – entitled Transformed Labour Force Survey – may not be complete until 2027. This admission came just a few days after the ONS published revised estimates of migration, uplifting its estimate of the level of net migration into the UK last year from 695,000 to 866,000 – a rise of over a quarter. Coupled with references in last month’s Mansion House speech by the Governor of the Bank of England to the poor state of employment statistics, this gave momentum to the story of failure and unreliability at the ONS.
If there was a widespread problem with the quality of all ONS statistics, that would be a serious worry. Fortunately, that is not the case. In my view, the ONS produces most of its economic data – such as GDP, production, overseas trade, etc – satisfactorily and it is trusted by users of this data.
However, that does not mean that all is well in the world of official statistics. I have been using official economic data for over 40 years, even before the Office of National Statistics came into existence and the relevant official body was the Central Statistical Office. In that period, I have discussed many issues with official statisticians and served on at least half a dozen of their advisory committees.
The challenge facing the ONS in producing economic statistics is that the structure of the economy is always changing. So the ability of statisticians to keep track of these changes and adapt their data collection techniques is a vital part of their job. What makes this more difficult is that these structural changes do not always take place slowly and gradually. They can be very abrupt. In the 1980s, the big challenge was to adapt GDP statistics for the big shift in the UK from a manufacturing-led economy to a situation where the services industries were the main driving force for economic growth. This required a big shift in resources by official statisticians to monitoring the services sector to measure GDP, and other key economic variables – and it took quite a while for this to be achieved.
There have been other issues which have emerged since then: problems measuring average earnings statistics in the 1990s and early 2000s, difficulty keeping track of migration since the mid-2000s, and now the problems with the Labour Force Survey. All these issues are linked to changes in the economy which the ONS has been slow and hesitant to identify and address. They have often had to be pushed to take action by other organisations – including the Treasury and Bank of England – because of the deteriorating quality of official data.
The ONS needs to be more action-oriented and responsive to the concerns of users when problems do emerge or important changes in the economy are identified. Instead, the management response has often been to initially deny there is a problem and then take a very long time to fix it. In the case of the Labour Force Survey, the current issues first emerged in 2021, so 2027 would be a long time to put in place the complete solution.
We need much quicker action from our official statisticians and their regulators to fix problems. In the case of the employment data, the ONS’s regulator, the Statistics Authority, should demand a 90-day action plan from the ONS. Failing that, the Treasury should step in and require prompt and effective improvements. We need to be confident again in UK employment data by early next year – not wait until 2027.
Click here to subscribe to our daily briefing – the best pieces from CapX and across the web.
CapX depends on the generosity of its readers. If you value what we do, please consider making a donation.
