Monthly Archives

May 2019

Are Britain’s Regional Divides Large or Small? A Response to Chris Giles

By | Uncategorised

Philip McCann1
University of Sheffield
16 May 2019

There are major problems with Chris Giles’ piece “Britain’s Regional Divide is Smaller than it First Appears”, Financial Times 9 May 2019.2

Chris Giles states that the evidence on regional inequality is much more complicated than often presented and he proceeds to argue that the UK regional inequality measures are somewhat distorted or even illusory due to “…the unique way the UK defines its administrative boundaries for international comparisons…”. He then compares London, Reading, Milton Keynes and the Wirral – and on this basis suggests that the gap between the highest and lowest urban areas in the UK is largely typical of other countries. He also compares Camden and The City of London with Torbay to argue that household income is a better and more equal measure of regional differences than GDP per capita. On this measure UK interregional inequality is much lower than for GDP per capita.

Yet, many aspects of Chris Giles’ argument are simply not correct. There is nothing ‘unique’ about how the UK defines its administrative borders; all countries display such idiosyncrasies. This is why it is essential to ensure comparability as far as possible when comparing different countries – exactly what the OECD data are designed to facilitate within individual classification or definitional schemes.

Chris Giles also jumps around between different spatial classification types, namely OECD-TL2 areas, OECD-TL3 areas, and OECD Functional Urban Area definitions. These spatial units cannot be compared in such a manner.

Finally, regional variations in household income will always be lower than regional variations in GDP per capita, so this observation tells nothing new. Household income and per capita GDP are measures designed to tell us different things. While Chris Giles may prefer the former most economists working in the fields or urban and regional economics and economic geography consider the latter to be a much better measure of regional productivity and prosperity, the proper focus for regional policy.

Once we compare like-with-like across 30 countries and for 28 different measures of interregional inequality, amongst the advanced economies which the UK considers as its peer group, the UK emerges as the most unbalanced and unequal country across the largest range of indicators, including disposable income.3 Indeed, apart from five countries at much earlier stages of development, namely Chile, Mexico, Turkey, Bulgaria and Romania, the only OECD or EU country which is more interregionally unbalanced than the UK is Slovakia. As such, the argument that UK interregional inequalities are typical of other countries is simply not correct.

In order to understand the degree to which the UK is interregionally unequal three further observations are very useful.

(a) Immediately after reunification in 1990 Germany was, not surprisingly, the most interregionally unequal advanced OECD economy. However, over three decades Germany has become more interregionally equal, while the UK has become more interregionally unequal. In terms of the ratio of the top 10% over the bottom 10% of GDP per worker levels, the UK became more interregionally unequal than Germany in 2008 and the same is true in terms of the top 20% over the bottom 20%. In terms of the top 10% over the bottom 75% of GDP per worker values, the UK became more interregionally unequal than Germany in 2004. The result of these trends is that today almost half of the UK population live in regions whose productivity (the best proxy for economic prosperity) is no better than the poorer parts of the former East Germany.

(b) Close to half of the UK population today live in regions which are poorer than the poorest US states of West Virginia and Mississippi (McCann 2016) – places which are often the focus of UK journalists aiming to articulate why people vote for President Trump.

(c) Out of the 179 NUTS3 areas within the UK, 159 areas have productivity levels below the UK average.4

Chris Giles has also tweeted5 on the subject “It’s rarely wise to rail against the conventional wisdom on regional divides in Britain. But look at some facts people…”.

This argument is also incorrect.

The longstanding conventional wisdom in the UK, and especially in many of the central institutions and think-tanks, has been that UK interregional inequality is not really important or significant in any way. The recognition that Chris Giles refers to on the part of some politicians and some economists actually represents a very recent, and still very incomplete, shift in thinking. Indeed, the conventional wisdom in many high-level circles is still that UK interregional inequality is still not a major issue. In this context the arguments put forward by Sir Angus Deaton this last week at the launch of new IFS enquiry into inequality were especially pertinent, highlighting the fact that many different indicators of UK inequality are evident interregionally.6

Chris Giles further contributes to the confusion by arguing that “There is no doubt that place has become a serious UK faultline, dividing north from south, urban from rural and cities from towns”. Yet, a careful reading of the OECD, Eurostat and ONS data7 shows that much of this recent conventional wisdom is also incorrect in that within the UK there are few, if any, major gaps in prosperity between urban and rural areas, or between cities and towns. Contrary to some of today’s popular narratives, the UK gaps between cities and towns and urban and rural areas are actually small by international standards.

In other words, the real geographical inequalities and imbalances within the UK are between regions; they are not between cities and towns or between urban areas and rural areas, and by international standards UK interregional inequality is very high. Indeed, the same IFS report referred to by Chris Giles also points out that household income today in most of the Midlands, Northern regions and Wales is equivalent to household income in the South East during the 1990s. In the case of GDP per capita the differences are even greater. I would suggest that many people are well aware that their communities are literally decades behind.

[1] I would like to thank Sir Paul Collier (University Oxford), Professor David Bailey (University of Birmingham), Professor Raquel Ortega-Argilés (University of Birmingham), Dr Paolo Veneri (OECD, Paris) and Ben Gardiner (Cambridge Econometrics) who all provided constructive feedback on earlier drafts of this short paper. The views advocated here reflect my own understanding of the issues.
[2] See:
[3] “Perceptions of Regional Inequality and the Geography of Discontent”, 2018, Discussion Paper, Productivity Insights Network PIN-D1, See:
UK2070 Commission Discussion Paper
A variant of this paper is forthcoming in Regional Studies. See:
[4] See:
[7] “Understanding Spatial Labour Productivity in the UK”, UK Office for National Statistics, 03 May 2019, See:

The value of adopting a systems approach to the productivity puzzle

By | Uncategorised

Is policy to blame for lagging productivity in the UK?
Since the 2007 financial crisis, productivity growth in the UK has been persistently weak confounding policy makers and prompting a flurry of research around what has become known as the productivity puzzle. There are many theories about the roots of this poor performance but one constant is the perception that public policy ought to be able to intervene to improve outcomes. While most acknowledge that there is likely no silver bullet there is emerging consensus that the evolution of productivity policy in the UK has not (yet) yielded predictable or positive impacts on productivity performance. We argue that the siloed nature of productivity policy may be hindering the development of an effective productivity programme and that adopting a systems approach to policy may provide new insights into the productivity puzzle.

It’s complicated: systems and productivity policy
We observe that even when a productivity policy in the UK was explicitly defined it was conceptualised as a bundle of separate policy areas rather than as a coherent policy programme. Even though there have been efforts to join-up policies and work cross-departmentally the policies affecting productivity have largely been enacted in disconnected silos.

The policy model depicted in Fig 1 (from the Industrial Strategy) illustrates this kind of thinking, which assumes that policies created in silos will aggregate to produce positive outcomes. Yet problems rarely conform to these assumptions. Policy spaces are messy, complex, and, sometimes even chaotic. There can be relationships between phenomena that appear stable but are subject to sudden shifts, or relationships that seem like they should function a certain way but do not. In fact, the ‘working whole’ is not reducible and cannot be described by the attributes of its parts alone.

Systems approaches focus on mapping and explaining systems or processes characterised by complexity – in other words systems that lack order and stability and universal laws. Fig 2 reimagines, in a vastly simplified form, the productivity landscape as seen through a systems lens. The core tenets of systems thinking are that outcomes are driven by intersections and interdependencies between elements within a system and that these relationships are (a) imperfectly recognised in the state of the art of research and policy and that (b) these relationships are not easily knowable.

What can systems approaches offer policy?
Here we highlight two of several high-level implications of adopting a systems approach to productivity policy.

First, bundling disparate policies is not enough. Rather, policies should be conceptualised to uncover, understand, and act on the interdependencies inherent in the system. This is not an easy task but will certainly remain elusive as long as policy silos dominate. Here there is an opportunity to build bridges between policymakers and researchers to better map and model the policy landscape.

Secondly, systems approaches encourage the adoption of experimental policy practices. These are interventions designed to engage with core stakeholders in targeted programmes and are explicitly designed to process feedback to better understand policy effects and uncover missing links. These policies emphasise experimental policy designs that incorporate rigorous and collaborative evaluation processes as well as the potential for frequent course corrections in response to findings. While experimental processes have been adopted in some areas of policy at present they tend to be small-scale and disconnected projects.

The Productivity Insights Network aims to contribute to the development of more effective policy by adopting a systems lens to mapping the systems and subsystems that affect productivity. The project’s mandate specifically aims to uncover intersections and interdependencies that have been overlooked or underplayed in an effort to better understand the dimensions of the productivity puzzle.

Dr Jen Nelles, City University of New York
Professor Tim Vorley, University of Sheffield

Read the full report here.