Why are there regional disparities?
This blog was written by our Research Director Tony Venables, Professor of Economics at The Unviersity of Manchester.
There are many descriptions of regional disparities in the UK and other countries, and many policy prescriptions offered. But there are few systematic attempts to analyse why such disparities exist. How do they originate, why are they persistent, and why has policy been largely ineffective in dealing with them?
How do disparities emerge?
The largest regional disparities within countries are found in the developing world, in particular in those that are doing well. Economic development is never a uniform process, affecting all places in a country at the same rate. It starts in some places – generally large urban areas – and is manifest in emerging regional gaps in income levels and in other socio-economic and health indicators. But part of the process of development is that prosperity spreads, albeit slowly and unevenly. Whilst measures of regional per capita income disparities are much higher in emerging markets than they are in advanced economies, measures of regional convergence area also greater (and increasing), indicating that emerging markets are seeing gaps narrowing.1
In advanced economies disparities are similarly driven by the success of booming regions – in particular major world cities – but also by negative shocks in other regions, generally associated with technical change and international competition. This shows up clearly in the experience of the UK and the US. The UK experienced dramatic structural change during the in 1970s and 1980s, much of this deindustrialization hitting sectors that were regionally concentrated.2
A good measure of these negative shocks is the fall in the male employment rate in affected places. UK areas that experienced the largest falls in the 1970s had been ‘average performers’ in the 1960s, but by the 1990s they were amongst the local authority districts with the worst socio-economic deprivation indicators. The effect has been extraordinarily persistent, with these places now making up 2/3rds of the most deprived districts, more than 40 years after the negative shocks occurred.3
The US picture is somewhat similar, as not-working rates (measures of unemployment plus those who have withdrawn from looking for work) have risen and remained high, particularly in parts of the far west and ‘a great swath of Middle America that runs from Louisiana up to Michigan’.4
Why are disparities persistent?
The persistent effects of negative shocks in the UK are broadly consistent with the experience of other advanced economies. Measures of regional convergence averaged across these OECD countries fell steadily through the 1970s, and have indicated barely any convergence at all since 1990.
What are the economic mechanisms that should drive convergence, and why have they failed? In broad terms there are two mechanisms; one is that people move to places where there are jobs, and the other is that jobs move to places where there are people.
People moving out of lagging regions has been a key part of economic development; the rural population of China has fallen by 300 million people in the last 30 years and some places have become completely depopulated. In the US high degrees of labour mobility were the major part of adjustment to shocks until the 1990s, but mobility has declined dramatically since then, to levels as low as those seen in Europe.
Not only is labour mobility slow to operate, it’s also problematic, deeply so in high income countries. It is generally the younger and those with greater earnings potential who move, leaving ageing populations and skill shortages behind. Thriving cities will and should continue to draw in population, but it’s a mechanism that tends to exacerbate rather than solve the regional disparity problem, at least in the short- to medium-run.
The second mechanism is that jobs should move into lagging regions to fill employment gaps. This is the outcome most of us want to see, and that we might expect as part of an automatic adjustment process. But there are several reasons why it doesn’t work well.
The first is that, particularly in advanced economies, regional wage gaps are often not large enough to encourage firms to invest. If a region suffers a negative shock the biggest impact is on prices of housing and land, not on wages; it’s the ‘immobile factor’ that takes the hit, as economic principles suggest. Lower land and office rents are an inducement for firms to move, but are generally only a small share of firms’ costs so have little traction.
Second, firms in many sectors are deeply embedded in local business ecosystems. Their productivity is high in the place where they already have customers, suppliers, a suitably skilled labour force and institutions that support innovation and supply knowledge. These are all complementary factors that are mutually reinforcing, creating agglomeration economies and clusters of activity. There is little incentive for a firm to move out of such a cluster and thereby forego the productivity benefits it delivers.
The strength of these agglomeration forces varies across sectors. They are very high in garments and apparel (Dhaka has an estimated 4 million garment workers) and – surprisingly – in zips and buttons (the city of Qiaotou produces 60% of the world’s buttons and 200,000 km of zippers per year). In advanced economies they currently appear strongest in high-skill activities such as finance, business services, media, and high-technology and R&D intensive activities. These are the sectors that cluster most and that are ‘sticky’, difficult to move.
The flip side of this is that less-skill intensive sectors seem to be easier to move (e.g. call centres, warehousing, back-office operations, food processing). This suggests that a region that has suffered a negative shock may see employment recover – albeit slowly – but biased towards relatively low-skill and low-wage activities. The shock gradually transforms from unemployment to persistent disparities in the skill mix of places, and consequent earnings. This seems to have occurred in the UK’s post-industrial regions.
Why hasn’t policy worked?
A place that loses its traditional competitive advantage is subject to the adjustment mechanisms outlined above. Replacing jobs is slow, and the replacements are likely to be relatively low skill. There is out-migration of some of the young and higher skilled, this making the place even less attractive as a destination for high-skill firms.
Land and house prices fall, and so too may public sector revenues, leading to an aging capital stock, low maintenance levels, loss of amenity, and possible deterioration of public services. This what economists term a ‘low-level equilibrium’. Essentially the geographical stickiness of many firms and sectors means that the market outcome is consistent with persistent disparities.
This stickiness makes policy difficult. If firms prefer to invest in places that have related businesses and a pool of appropriately skilled labour (as well as infrastructure and amenities) then there is a first-mover (or coordination-failure) problem. Which firm is going to be the first to enter an as yet non-existent cluster? Scale, or its lack, is an inherent part of the challenge. A pool of skilled labour or cluster of related firms necessarily has to reach some substantial scale, which in turn means that relatively few places can attract such a cluster.
So why hasn’t policy worked? Three conditions are necessary – if not sufficient – for successful policy. The first is to understand and be able to act on local circumstances. Numerous factors determine whether or not a place is attractive to the key decision takers, firms, workers, and households. Local information is needed to figure out what are the specific critical factors and hence set the policy priorities. And local institutions need to be empowered to act on these priorities.
The second is to understand the complementarities that link economic activities in a place. Some of these are at the local level – business links within a town – and others across wider geographical units, such as city hinterlands and the supply of skills in a travel-to-work area. Decision takers need to be aware of them, and empowered to foster their development. Some of these are externalities, and institutions need to be able to secure a degree of coordination and ensure that these spillover effects are internalized in decision taking.
The third is to understand scale. Some projects are inherently large and lumpy – such as high-speed railways or major science centres – and have to be concentrated in few places. Some places are higher priority for intervention than others, some because of high levels of deprivation, others because they are better placed to drive growth and offer better prospects of success. This ‘lumpiness’ means that, for some items of expenditure, policy must pick spatial winners rather than smearing resources across places.
These conditions are not met by traditional instruments of top-down regional policy such as fiscal subsidies and tax breaks. Tax incentives may induce a firm to invest, but often fail to catalyse local development. They may even be counter-productive as they simply accelerate the process of filling places with relatively mobile and low-skill jobs in warehousing, back-offices, and the like.
The three conditions require a difficult distribution of levels of authority. Local knowledge and empowerment, regional coordination and cooperation, and national decisions on ‘lumpy’ investments make for a complex policy environment. In the UK we are far from having established a policy framework that can achieve this.
1Floerkemeier, H., N. Spatafora and A.J. Venables (2021) ‘Regional Growth and Inclusiveness’, IMF WP/21/38
2Not the first negative shock, as many of these places were also hard hit in the interwar period.
3Rice, P.G. and A.J. Venables (2021) ‘The persistent consequences of adverse shocks; how the 1970s shaped UK regional inequality’, Oxford Review of Economic Policy, 36.3.
4Austin, B., E. Glaeser and L. Summers, ‘Jobs for the heartland; place-based policies in 21st-century America’, Brookings Papers on Economic on Economic Activity, spring 2018.