In the last budget, chancellor George Osborne gave Sir Charlie Bean the task of undertaking a review of the UK’s economic statistics. The proposal was triggered by concerns about the difficulties in measuring productivity accurately in a modern economy and an independent review was thus commissioned by the chancellor of the quality, delivery and governance of the UK's economic statistics.
The call for evidence in the review had just finished and Sir Charlie attended a meeting at the RSS last month (September 2015) to discuss his remit and listen to what others had to say about the current status of economic statistics. Sir Charlie was asked to compile the review on account of his 14 years on the Monetary Policy Committee at the Bank of England where he was also deputy governor.
He began by outlining what he would be assessing in his review. The major structure of the national accounts was installed in the 1930s and there are questions as to whether this is giving an accurate measure of the modern financial, services and digital industries. Following this, he says he will examine the capability of the Office for National Statistics (ONS) to deliver modern economic statistics and to what extent they should make better use of admin or external data sources.
Finally he will be looking at the governance arrangements for the production of these statistics. The inclusion of governance as an element of the review has raised concerns about how it will affect the current independence of official statistics. When questioned on where this will lead, Sir Charlie reassured those present that independence was a red line, both for him personally and in the review’s terms of reference.
Sir Charlie finished his short overview by saying he was currently in ‘listening mode’. So to widen the discussion, two other speakers were invited to give their view on where economic statistics stand today. Diane Coyle, economics professor at the University of Manchester spoke first, followed by David Caplan, a statistician who was formerly the deputy director for the UK's national accounts.
Diane began her overview by showing the audience a statistical bulletin from the 1880s. Within the bulletin was page after page of agricultural statistics, yet only a few pages were devoted to industry, despite the industrial revolution being well established by that time. Her point was that economic statistics have always lagged behind the actual changes taking place in the wider economy.
With this in mind, she asked what effect the dramatic falls in the price of technology like computer chips and genome sequencing is having on economic output. New technologies such as 3D printing or graphene will have a major economic impact in the years to come, but the effects of the growing availability and affordability of products like this are difficult to track in our current statistics. This begs the questions of how innovation is benefiting society outside of the old statistical calculations of GDP.
Beyond this, Diane asked if the current categorisation of jobs and industries was keeping pace with what was happening in the digital and short-term 'gig' economies. These new industries have created a host of new job titles and employment arrangements that need to be carefully considered if the ONS is to accurately measure them in the future.
David Caplan then began his talk by briefly running through the long line of economic reviews that preceded Bean (which he has previously written about on these pages). He also described the diverse and dynamic needs that users of economic statistics can have. This problem is compounded for statisticians by economists having fundamental disagreements on what kind of data is important to their work.
Another problem is fact that the EU is the major driver behind many of the statistics that the ONS produce. These are compiled so that cross country comparisons can be made in Europe, but it introduces the dilemma of when national needs should supersede Eurostat’s needs.
He also described the caution needed when using proxy or admin data to compile statistics, as these can be susceptible to system changes and could end up biasing the data or playing havoc with the time series.
David finished by reiterating that the needs of users has to be prioritised. He said the ONS largely has a producer’s culture, and while this is changing, those end users need to kept in mind at all times.
The floor was then opened up for questions and comments. These included how the recent reviews into price statistics will relate to the Bean review, how to counter trivial movements in economic statistics getting blown out of proportion in the media, and the need for an independent system for the production of statistics.