Pretty much everyone thinks it’s a good idea to have more economists (code for analytical capability) at the ONS but opinion divides when there’s discussion as to what they should be doing. There is a need to have people who can acquire and probe exiting data assets to make them sweat in the spirit of the Bean Review. In contrast, there is no need for the ONS to have any more descriptive writing and (sometimes dumbed-down) publications that serve some unspecified need. That would be a wasted opportunity. There is a risk that the hiring of large numbers of economists in a hurry, mostly in their early careers, as opposed to curious souls with experience, will lead the ONS down the wrong path. Meanwhile economists outside government need to start making the case for better statistics.
The Bean Review (which is now gospel!) discusses economists at the ONS. In the recommendations section, there is reference to “greater economic and analytic expertise” to be “embedded within ONS” (para 1.43). It says: “Shedding light on hard-to-explain puzzles often requires digging down below the surface into the underlying data. Being adept at this is key to understanding the shortcomings and limitations in the data, and also identifying new trends. An enhanced capability within ONS to interrogate the underlying microdata would better support the production of economic statistics …….“(para 1.42). That in turn leads to: “Recommended Action 15: Increase the economic expertise within ONS and implement a smart and effective system for quality assurance and sense-checking across the production of all economic statistics.”
Bean almost puts pressure on ONS to recruit economists now and fast! It says: “It seems plausible that the increase in the number of economists across government departments represents greater appetite for economic analysis for policy and operational reasons. The number of economists at ONS has increased to over 40 in 2015 which represents a very high rate of increase, but from a very low base. However, ONS still has far fewer than many other departments such as HM Treasury or even the Ministry of Justice. It is also a long way behind other analytical institutions like the Bank of England or the OECD. It seems clear that ONS is not only much less dominant as a centre of statistical expertise than it once was, but that it is also lagging as a centre of economic expertise.” (para 4.105)
The Report went on to say: “Since the Interim Report, ONS has announced plans to increase the number of professional economists, including by embedding more of them within statistical production teams. This will bring together staff trained in economics with those with backgrounds in statistics and other fields to create multi-disciplinary teams for statistical production. This is a very welcome first step in augmenting the capability of production teams and encouraging them to be more self-critical; it may also yield some quick returns in moving to a smarter approach to quality assurance and a reduction in the frequency of unnecessary errors.”
Progress to date
Although not entirely clear, “Economics at the ONS: Increasing Openness, Improving Capability“, published in December 2015, is probably the plan referred to in the Bean Review. It is mostly about planned work not the recruitment of economists per se, and says only this “ONS intends to recruit increased numbers of economists by a multiple of the existing complement over the next few years”.
A little more insight into economists at the ONS was revealed in the response to my FoI request last November (see the table below). Most notable perhaps was the drop in grade 7 and above from the recent peak of 15 in 2010 to just 8 in 2013. That shows how economic statistics and related analysis had been run down by UKSA under the previous National Statistician. It was in part quite probably a consequence of relocation to Newport.
It is perhaps worrying to see that the ONS is recruiting mainly at the entry level. The response said that the ONS had “nineteen affiliated members of the GES and 23 sandwich students accessed through a Government Economic Service (GES) selection process”. It appears that affiliated members of the GES “would be automatically converted to permanent roles if the individuals concerned can pass the entrance examinations to become full members of the GES”. We wait to see if recruits that are not yet up to GES standards, and are probably mostly in their first jobs, will be able to add the sort of curiosity and innovation that Bean envisaged. Recruiting a lot of economists in a hurry, to an organisation and town with limited track record in the domain will be a real challenge if it is to be done successfully.
The ONS is currently advertising for (4) grade 7 economists and that ad looks more Bean-consistent.
It seems to me that there are some publications that could be cut or trimmed back and it has to be hoped that more “analysis” will not lead to more of these sorts of publications. Relative to the deeply curious probing of data sources that is required, much of the descriptive writing produced by ONS is of medium to low grade. To give three examples:
- This week the ONS published: “The UK’s trade and investment relationship with Africa“. My immediate reaction was to wonder why. There’s nothing inherently wrong with a “short analysis piece on the UK’s trade and investment relationship with Africa in the last 20 years.” But is it what users need and where the ONS attention should be focussed? It would be more in the spirit of the Bean recommendations if staff were delving deep into the many pressing and fundamental issues set out in “UK trade development plan: 2016“, the latest ONS work plan on trade. There are other “analytical” pieces in the ONS release calendar that could be done without. It is for those outside the ONS to write stories based on published data.
- There is much to be excited about in the ONS Business Plan (2016-20) but on the question of publication it does not say much. Under dissemination (page 6), it does say: “We will engage with users and external experts to ensure economic statistics and the ‘economic story’ are easily accessible. We will continue to produce the Economic Review as a platform for ONS to provide analysis and commentary on the UK economy, increasingly drawing out and pulling together the economic story from individual releases.” I find it hard to know what to make of the Economic Review. The latest edition is nearly 30 pages long and its purpose is far from clear. It’s an eclectic mix. Some is tending to the banal. It’s odd to find some of it in a monthly publication. (For example, consider the start of section 7 on trade: “Trade remains an important element of UK economic growth; its performance has been varied during the last six years …….”.) It doesn’t provide a rounded view of the economy as the ONS decided back in 2011 (at the same time the website was changed) to stop publishing non-ONS data. Without a rounded view using all data, there will always be a weakness in the description of any domain or topic. Anyone wanting such an analysis would go elsewhere, perhaps to the Bank of England’s quarterly inflation report or its narrative on monetary policy.
- Some press releases are unreadably long. Looking at some of those just published, that on public sector finances was 38 pages of text, labour market 37 pages and retail sales was 19 pages – and those were monthly releases. There is a debate to be had but something needs to change.
Finally I’d make a plea to encourage economists out of government to be active in asking for improvements in statistics. It’s been hard to engage with the ONS at times but the new structures look promising and certainly point to a willingness of the statisticians to help users. It was great to see the Royal Economic Society co-sponsoring the “Bean Review of Economic Statistics: Making it Happen” meeting at the Royal Statistical Society earlier this month. While there are notable exceptions and economists made some great responses to the Bean Review, there is often an incentive for practising economists not to complain about data. Conclusions of reports always need to be striking to catch the readers’ eye and saying that the data on which the report is based has notable weaknesses or is deficient in some way hardly helps! The more users that can share their views with the ONS (and the UKSA board) the better for the long health of the country’s statistical system – and resulting analysis and policy.