Benefits of decentralisation?

Published: 27 April 2015

David Comerford discusses the latest IFS analysis of Scottish Government finances under so-called full fiscal autonomy. How did the SNP respond to these figures?

The IFS has published an analysis of Scottish Government finances under so-called full fiscal autonomy, as called for in the SNP manifesto. The Scottish Labour party have highlighted the figures from this analysis as they try to claw back some ground in the opinion polls before the coming general election. And the figures certainly are not comfortable reading for the SNP, with a projected fiscal balance for Scotland in 2019-20 of -4.6% of GDP compared with a whole UK figure of +0.3%. Nicola Sturgeon has responded by describing the IFS figures as "academic projections for a status quo situation" which could be improved with extra powers for the Scottish parliament. So the argument from the SNP seems to be an acceptance that the public finances in Scotland are worse than the UK public finances, but an expectation that more powers would lead to an improvement in these finances.

A CEBR report from 2012 showed that the status quo public finances relative to the UK average were very negative for North East, and North West England, Yorkshire, the East and West Midlands, Wales and Northern Ireland; broadly neutral for East and South West England, and for Scotland[1]; and very positive for London and the South East of England. This pattern could be used to justify the relatively centralised nature of the UK, and to argue against moves towards decentralisation or federalism: decentralisation is unaffordable for certain regions because they are not productive enough to fund their spending internally and the loss of fiscal transfers would cause a further fall in living standards for these already underperforming regions. As for the SNP in Scotland, any proposals to move towards a more decentralised UK must be accompanied by an expectation that decentralisation will change the pattern of economic activity to boost output in peripheral regions. Is there any evidence that autonomous regions are likely to do better or worse than peripheral regions within a centralised state?

This is a hard question to answer because so many things differ across regions and across countries that it is very difficult to definitely make the claim that differences in the autonomy of regions causes differences in their productivity and in the living standards of their inhabitants. Feld & Schnellenbach (2010) summarise the literature on this question and conclude that there is no evidence of any effects of decentralisation upon growth rates, but that “the evidence hints at a positive effect on overall productivity. …   the effect of federalism to a large extent is a level effect: a switch from a unitary to a federal regime would help to realize efficiency gains and increase total factor productivity, but there would be no lasting effect on the steady-state growth rate.

One way to see these hints of a positive effect, and at the same time appreciate the difficulties of assessing whether this something real or spurious, is to look at the variation in income across regions in Europe. The results of doing this can be described in terms of what happens across UK, France and Germany (which together account for 36 out of the 84 European regions considered). These three large countries have similar average per capita incomes, but while decentralised Germany has relatively low regional inequality, centralised Britain and France have much poorer peripheral regions with large capital cities that are among the richest places in the world. We thus see a positive correlation between the level of decentralisation and regional productivity and living standards. But does the decentralisation cause the better regional performance? That is a much harder claim to make. Indeed causation could even run the other way with wealthier regions demanding greater autonomy rather than greater autonomy making the regions wealthier. There are also concerns with the data: e.g. the fiscal proxies for regional autonomy make Germany appear highly decentralised, but there is also a large degree of equalisation in Germany. So depending upon your belief about what particular aspect of decentralisation might cause enhanced regional productivity, the true level of such decentralisation in Germany, and across the data, may differ from the fiscal proxies usually used.

The SNP can point to the correlation between productivity and decentralisation to support their pitch for full fiscal autonomy. Social scientists however will be more sceptical in advance about whether any productivity boost can be expected. What is needed to provide hard evidence on the link between productivity and decentralisation is a natural experiment. Perhaps, with Smith, EVEL, Devo-Manc, and a record number of SNP MPs, the UK will implement a large programme of decentralisation after the election; if there is then a step change in the trends relating to regional inequality then perhaps solid inferences can be drawn. Ex-post, it may be good policy or bad policy to implement such a programme of decentralisation – we don’t know yet given the current state of knowledge, but one unambiguously positive outcome will be its contribution to science!


[1] Scotland’s deterioration since then is largely explained by oil revenues.


Coronavirus and the Constitution

Scotland, the last in line to relax lockdown

Leadership, learning and knowledge: lessons from COVID-19

Why have the UK's governments diverged on easing lockdown?