top of page

Levelling Up Productivity Gaps in the UK

Ethan Ilzetzki [1]

Thursday, July 7


A majority of the CfM-CEPR panel of experts on the UK economy believes that the large productivity differences across UK regions are attributable to either agglomeration effects or sorting, self-reinforcing phenomena that make some regions become productivity powerhouses and others fall behind. 31% of the panel thinks that these productivity differences are due to fundamental factors such as education, local infrastructure, or poor transportation links across regions. These productivity gaps should be addressed with a combination of public investments, incentives for private investment, and investments in skills and education. Tax policy, devolution of fiscal power, and R&D subsidies were also mentioned as policies that would help lagging regions.


The June 2022 CfM survey asked the members of its UK panel to evaluate the reasons for productivity differences across UK regions and policies that might close these gaps.

Levelling Up Productivity Gaps in the UK

The UK has experienced unusually slow growth rates in productivity since 2010, what has been termed the UK’s productivity puzzle (see also here, here, and here) This slow productivity growth at the national level masks significant regional disparities in productivity levels, with persistent geographical clusters of high and low productivity subregions. Fifteen of the highest productivity subregions are located either in London or the Southeast region; there are additional high productivity pockets in other regions, primarily urban areas. Most of the low-productivity subregions are concentrated in rural areas, including the Yorkshire and the Humber region and other rural areas in the Midlands and Wales.

Research has cited several potential reasons for these regional differences. A review by the Industrial Strategic Council found three main explanations for these inequalities. First, there may be fundamental differences across regions. These include geography, local governance, educational quality, infrastructure, and historical patterns of economic specialization.

Second, a large body of academic research has shown evidence of agglomeration effects, with workers’ productivity enhanced by being in proximity to other productive workers. This, in turn, leads to a concentration of investment and clusters of firms in high-productivity industries, in a small number of locations. Marshall (1890, pg 225) first emphasised that “a localized industry gains a great advantage from the fact that it offers a constant market for skill”. Krugman (1991) further formalised this notion by mathematically demonstrating why firms choose to operate in close proximity to other firms and not ‘defect’ to alternate regions.

Third, there may be sorting effects, whereby high-skilled workers move to large cities to benefit from their amenities. This drives up the cost of living in these areas, inducing low-skilled workers to move out of urban areas. This leads to a bifurcation in productivity between cities and rural areas. Kerr et all (2017) substantiated this theory by analysing emigration patterns of high-skilled workers to OECD countries, attributing these movements to positive agglomeration externalities.

Finally, local transportation and digital infrastructure may play an important role. Metropolitan areas, particularly London, that have better transportation links with peripheral areas can benefit better from agglomeration effects. Public transport and road and rail links could boost productivity in lagging parts of the country by integrating them with productive metropolitan areas, so-called “borrowed” agglomeration economies. Combes et al (2008) found that low transport costs and proximity to markets have a major impact on firms’ productivity and profitability, even more so than high employment density. Good digital infrastructure, in turn, allows remote areas to be integrated with more productive regions.

The UK government has committed to reducing these regional disparities and boosting productivity under its ‘Levelling Up’ scheme. Under the broad umbrella of reducing geographic, socioeconomic and health inequalities, the government has outlined 12 missions to achieve by 2030, including increasing infrastructure and R&D investment in low-productivity regions, improving housing, education and healthcare in lagging areas, and empowering local leaders and communities by devolving power.

The UK government has allocated £4.8 billion to date to the Levelling Up Fund to improve infrastructure in deprived towns and other lagging areas, with a total of just under £1.7 billion having already been shared between 105 towns, cities and areas. The government has committed £2.6 billion to empower localities to support local businesses, focus on building ‘pride in place’ through increasing social and physical capital and increase life chances for communities, through the UK Shared Prosperity Fund. The Community Renewal Fund and the Community Ownership Funds aim at developing local skills and infrastructure at the community level.

There is still widespread debate amongst academics regarding the effectiveness of place-based policies to reduce regional disparities. Gaubert (2018) analysed how place-based policies to incentivise firm relocation do not necessarily lead to more regional equality, and instead, could even lead to more spatial inequalities as small and large cities expand at the expense of mid-sized cities. However, Garcilazo et al (2010) argued that after a certain stage in a country’s development process, the influence of the country’s ‘growth poles’ on aggregate growth will decrease due to agglomeration diseconomies. Consequently, aggregate growth will increasingly depend on lagging regions, and hence, place-based policies will be critical to ensure high aggregate growth and productivity in major developed countries.

Additionally, investment in skill development and policies dedicated to increasing labour force participation could have a positive impact on regional productivity performance.

In terms of centralization of power, the United Kingdom is below the OECD average in all dimensions, and fiscal and administrative devolution could incentivise local governments to take charge of local development and boost regional productivity.

Finally, Gai et al (2021) build a regional model of the UK and show that tax policy and labour market reforms would have a larger effect on lagging regions and are therefore effective policies to ‘level up’.

This month’s survey asks the CfM panel of experts on the UK economy to evaluate the main factors driving regional productivity disparities in the UK and which policies could best help reduce regional productivity disparities.


Question 1: What is the primary factor driving regional productivity disparities in the UK?







Twenty-three members of the panel answered this question. 44% of the panel attributes productivity differences to agglomeration effects, with an additional 9% viewing sorting of skilled workers into high productivity areas. Combined, this means that a majority of the panel believes that productivity gaps result from self-reinforcing phenomena. 22% of the panel thinks that productivity gaps are due to place fundamentals and an additional 9% opines that they are due to poor transportation and connectivity of some UK regions.


The most common view is that agglomeration effects are the main cause of productivity differences within the UK. This view is summarized by James Smith (Resolution Foundation): “Services-dominated economies, like the UK, benefit from agglomeration effects - meaning that some high value-added industries concentrate in large cities (i.e. London).” Roger Farmer (University of Warwick) supported this opinion, claiming that there is “overwhelming” evidence on the importance of increasing returns-to-scale (brought about by agglomeration effects) in ensuring high productivity, and high productivity growth.


However, most panellists state that multiple factors are in play. Morten Ravn (University College London) describes how agglomeration effects may reinforce place-based fundamentals: “Education and infrastructure differences are important elements, and these interact with agglomeration effects (sorting). It takes a well-educated workforce for companies to invest in high productivity activities, a well-educated workforce is the outcome of education, employment opportunities, and agglomeration, and without the right infrastructure, the returns on such investments may not be high.” Jagjit Chadha (National Institute of Economic and Social Research) describes several causes behind this issue, placing “a systematic failure to address shortages in capital (human, physical and financial) across the country” at the forefront. He points out that this issue was previously expounded upon in the first report of the UK Productivity Commission.


Several panellists also outlined the role of poor transport infrastructure and connectivity issues as leading factors behind regional productivity differences. Michael Wickens (Cardiff Business School & University of York) described the subpar transportation network in the North: “In York, where I live, as in the rest of Yorkshire and Humber, poor transportation and roads are a major factor. There is still no dual carriageway to Scotland from Newcastle. The same is true of roads connecting Sheffield to Manchester. And rail is even worse across the north.” Simon Wren-Lewis (University of Oxford) further highlighted the impact of connectivity differences on regional productivities using the bus network as an example, discussing how it “works well in London but is failing elsewhere.”


Question 2: Which policies could best help reduce regional productivity disparities?




Twenty-four panel members responded to this question. 38% of the panel thinks that public or (subsidized) private investment would be the most effective policy. Respondents in this category were also the most confident in their responses. 21% believe that investments in skills and education would best help reduce productivity differences. Tax policy and devolution of fiscal powers each received 13% of the responses and a single panel member suggested publicly funded R&D as the most effective policy.


The majority of the panel believes that public or subsidised investment in lagging communities is required to reduce regional productivity disparities. This view is espoused by James Smith: “More investment in transport and communications infrastructure would reduce incentives for industries to cluster.” John Van Reenen (London School of Economics and Political Science) further substantiated this viewpoint, citing Criscuolo et al (2019) as evidence for the efficacy of UK place-based investment subsidies. Roger Farmer emphasises the critical role that the government must adopt to ensure greater regional equality: “High productivity areas do not develop without infrastructure. Government has a role through the provision of road and rail networks and by investing in research universities.”

However, Simon Wren-Lewis argues that public investment will only work in conjunction with the devolution of fiscal powers. He states that “a major obstacle to improving productivity and prosperity outside London is H.M. Treasury, which has been very reluctant to fund major infrastructure projects outside the capital, and even more reluctant to allow any devolution of fiscal powers. While that continues, not much will change.” Further, Roger Farmer warns that public investment will only be effective if it is concentrated and targeted, arguing that “the UK has room for no more than two or three additional growth centres: one in the north of England, one in Scotland and one in Northern Ireland.”


Policy certainty is of critical importance if any place-based policies are to work. Jagjit Chadha stresses the need to have a “commitment to the delivery of these objectives over the long run so that firms and households believe that a significant shift [in] productivity prospects will occur and then act accordingly.” This would mean a shift in policymaking focus towards creating “a focal point around a markedly different equilibrium.”


References and Further Readings:

Combes, Pierre-Philippe, Miren Lafourcade, Jacques-François Thisse, and Jean-Claude Toutain (2008). “Long-run spatial inequality in France: Evolution and determinants”,, 5 December.

Criscuolo, Chiara, Ralf Martin, Henry G. Overman, and John Van Reenen. 2019. "Some Causal Effects of an Industrial Policy." American Economic Review, 109 (1): 48-85.

Gai, Yue, David Meenagh and A. Patrick Minford, “North and South: A Regional Model of the UK,” CEPR Discussion Paper 15635.

Garcilazo, J.E, Joaquim Oliveira Martins, and William Tompson (2010). “Why policies may need to be place-based in order to be people-centred",, 20 November.

Gardiner, Ben, Bernard Fingleton, and Ron Martin. “Regional Disparities I nLabour Productivity and the role of capital stock,” National Institute Economic Review 253 (2020): R29–R43. doi:10.1017/nie.2020.28.

Gaubert, C (2018), “Firm sorting and agglomeration”,, 14 November.

Kerr S.P, William Kerr, Çağlar Özden, Christopher Parsons (2017). “Global talent mobility and human capital agglomeration”,, 31 January.

Martin R, et al (2019), The Economic Performance of Britain’s Cities: Patterns, Processes and Policy Implications

Marshall, A. Principles of Economics, Palgrave Macmillan UK, 2013. ProQuest Ebook Central,

Krugman, Paul, (1991), Increasing Returns and Economic Geography, Journal of Political Economy, 99, issue 3, p. 483-99,



[1] I thank Suryaansh Jain for his outstanding research and editorial assistance.

bottom of page