(Title Image: Wales Online)
Almost every nation in the world has an official agency dedicated to collecting and publishing statistics.
Official statistics are important not only for policy-makers to keep track of how their policies are working, but also to help governments plan services, come up with new laws and determine their respective nation’s strengths and weaknesses.
The Current Situation
At present, the Office of National Statistics (ONS) – an independent UK Government department – is the designated statistics body for the whole of the United Kingdom. They’re overseen by the UK Statistics Authority.
The ONS headquarters is, as you probably all know, actually in Wales. The exact numbers of people working at the Newport HQ have been hard to pin down (as a large number of ONS workers are “field workers”, presumably working on the census), but it’s likely to be anything between 1,000-3,000. The ONS also has smaller offices at Titchfield in Hampshire (which focuses on demographics and undertakes the census) and London (focusing on media work and methodology).
According to the latest ONS business plan (pdf), the ONS is highly accurate and trusted, with 99.8% of publications being free from major errors in 2017-18 and no data breaches occurring.
In 2018-19, the total ONS budget was £283million, with £189million dedicated to day-to-day running costs and £94million provided in preparation for the 2021 census. The amount the ONS receives (vis-a-vis the census) varies from year to year, usually peaking in the census year itself.
It’s worth stating though that not all statistics are gathered and disseminated by the ONS. There are other government or non-government agencies that produce official statistics – like the UK Treasury and Office for Budget Responsibility. Other bodies and agencies produce statistics-like forecasts – such as the Institute for Fiscal Studies, WISERD, HMRC, the Welsh Revenue Authority or the Welsh Public Policy Institute.
Each of the devolved governments has their own statistics department, usually collating official statistics relating to devolved public services, such as the NHS and local government. The Welsh Government’s statistics department publishes under the Statistics for Wales and StatsWales titles – all working to UK Statistics Authority benchmarks.
In the draft 2019-20 Welsh budget (xls), the Welsh Government intends to spend around £3.6million on statistics, information and research.
However, there are notable gaps in statistics produced specifically about Wales – particularly economics and fiscal statistics – which are often based on estimates.
Eurostat, Brexit & the future of ONS Newport
At a European level, Eurostat harmonizes statistical methods across EU member states, the EFTA (Norway, Switzerland, Iceland, Liechtenstein) and EU candidate states. Eurostat is responsible for familiar models such as NUTS regional statistical areas.
As of now, it looks like the UK will withdraw completely from Eurostat after Brexit and it’s unclear what will replace it or whether the UK will simply align itself with Eurostat unilaterally. Eurostat has already started producing statistics with the UK omitted.
The policy of the UK Government since 2007 has been to gradually move as many ONS jobs and roles as possible to Newport, but this has been criticised and there’s quiet lobbying for the ONS to return to London.
In 2017, the Centre for Cities said the ONS’ out-of-town location and the weak knowledge economy of Newport suggests “no immediate local services are sustained by its presence”, while it’s struggled to fill job positions as London-based civil servants are reluctant to move to Wales. They also blamed the move to Newport for supposed “lower quality” of output, believing it made more sense for the ONS to be based in “a deeper, more highly-skilled labour market where recruitment is easier”.
The Financial Times has also criticised the physical distance between the ONS and both economics professionals and Bank of England.
In short, Wales isn’t officer class.
My personal belief is the only reason why the ONS hasn’t moved back to London is because too much time and effort has already been spent moving operations to Newport. Once it’s deemed affordable or politically desirable to do so, the ONS will probably leave Newport – independence, devo-max or not.
Statistics & Independence
First of all, it’s worth outlining what type of statistics we need from an official statistics agency:
- Demographics and the census (population, population estimates, age range)
- Health & wellbeing (disability, minority groups, general happiness/life satisfaction, lifestyles)
- Immigration and emigration
- Employment, working patterns, personal & household incomes, economic inactivity and unemployment
- Business, economics and public sector output (gross value added, imports, exports, indices of production, property, tourism figures etc.)
- Monetary supply, inflation and retail prices
- Public sector performance indicators (NHS waiting times, NHS staff, bed blocking, qualifications etc.)
- Public spending and tax (accurate, not estimates)
An independent Wales would need to designate a national statistics institute. Scotland’s Future (pdf – p317) suggested that Scotland would use the assets of the UK-wide statistical service and designate the National Records of Scotland as its statistics agency.
Wales doesn’t have an exact equivalent of the National Records of Scotland – our Archives Wales (at the National Library in Aberystwyth) is mainly focused on cultural and social archives, not statistics as such.
So the obvious choice would be to designate Statistics for Wales/StatsWales as the official statistics service, spinning it out from the Welsh Government as an independent arms-length agency.
What would be the practical implications of that?
First of all, and the most obvious one, is the ONS in Newport would almost certainly move back to London – based on the Centre of Cities study, the impact it would have on Newport would be negligible as it (supposedly) hasn’t made any noticeable impact on the city in the first place.
While all current ONS jobs will be lost, a fair proportion of them would be replicated within StatsWales (or whatever it would be called) as a result of StatsWales expanding to cover the ONS’s functions – with current ONS staff presumably having first refusal on any vacancies under a TUPE.
For want of comparison, in 2017-18 (pdf) the National Records of Scotland – which has a much broader remit than StatsWales – had annual running costs of around £26.3million and employed 430 people. You would expect, based on population, a Welsh equivalent would have annual running costs of around £16million (roughly proportional to our contribution to ONS running costs) – with top-ups for census preparations – and around 240 employees.
A Welsh statistics authority could be run by a Chief Statistician, appointed by the relevant Senedd committee. You would also assume they would be based in Newport, though perhaps in a more attractive city centre location – the present ONS site can be sold and the receipts used to help with start-up costs.
Secondly, there would be a need to replace the UK Statistics Authority as the arbiter of statistical methods and to provide oversight to StatsWales. This could take the form of a board part-appointed by the Welsh Government and Senedd, drawing on academic expertise from our universities.
Another option would be to turn the ONS (renamed) into a pan-GB statistics agency operating for Scotland, England and Wales in the same way Eurostat operates for the EU. Wales would retain control over the census and how statistics are presented, but the backroom statistical functions would remain with the ONS. Under these circumstances, the ONS might stay put in Newport, with the respective nations making a proportionate contribution to running costs and the UK Statistics Authority (renamed) having appointees from all of the nations.