In this article, we focus on the NUTS1 regions<\/a> of the UK, of which there are 12: the three devolved nations plus nine English regions. Productivity differences can, of course, be examined at finer levels of spatial granularity (for example, city-regions and local authorities), and recent work by Beatty and Fothergill (2019)<\/a> do just that for cities and sub-regions of the UK.<\/p>\n Data from the Office for National Statistics (ONS), covering the period to 2018, show how labour productivity across the regions and nations of the UK compares (see Figure 1). We can see that over the two decades to 2018, little changed: London pulled ahead a little, while Scotland caught up a little.<\/p>\n Recent analysis by Beatty and Fothergill (2019)<\/a> explores London\u2019s productivity advantage in these headline indicators and the extent to which different features of London and its economy could explain this advantage. These range from the effect of housing market effects (\u2018imputed rents\u2019) through to commuting and industry\/occupation effects.<\/p>\n Figure 2 summarises the effect of these different factors on London\u2019s headline productivity advantage over the UK as a whole. We can see that the commuting is a particularly important aspect, but also the composition of the labour market by industry and occupation.<\/p>\n At a broad level, we can think of four main drivers of regional productivity differences (Zymek and Jones, 2020<\/a>):<\/p>\n Workforce attributes: the skills, motivation and health of the workers that a place is able to attract or retain<\/strong><\/p>\n Research has shown the importance of human capital is in driving differences in regional economic development (Gennaioli et al, 2013<\/a>). There is also evidence that productivity and hence wage differences are driven in part by an inefficient spatial allocation and sorting of people, and that policies that restrict or frustrate movement across regions, and in particular into cities, exacerbate these inefficiencies (Hsieh and Moretti, 2015<\/a>).<\/p>\n Capital and infrastructure: the machinery, equipment and infrastructure that supports work in a given location<\/strong><\/p>\n We have too few data on regional capital stock in the UK to say too much about this as a driver of productivity differences \u2013 at least quantitatively and at an aggregate level. There are some experimental data published in 2010 for EU-27 NUTS2 regions (Derbyshire et al, 2010<\/a>) and a recent study in the National Institute Economic Review<\/em> (Gardiner et al, 2020<\/a>) that updates these data and uses them to show that variations in the capital stock are an important driver of regional labour productivity.<\/p>\n There has also been some work using data on investment (\u2018gross fixed capital formation\u2019) for Scotland as a proxy for its capital stock (Mitchell and Zymek, 2018<\/a>). Relatedly, there has been some work using firm-level data to look at the extent to which foreign direct investment can help to explain regional productivity differences (for example, Harris and Moffat, 2017<\/a>).<\/p>\n Geography and local institutions: inherent characteristics of a location that may be conducive to economic activity, such its location (coastal or inland; remote or central) or local culture<\/strong><\/p>\n The contributions of geography and local institutions are hard to quantify and assess, although areas that are more central appear to have a clear productivity advantage.<\/p>\n Research has explored phenomena like agglomeration and clusters that can enhance the productivity of businesses in an area through developing its enterprise ecosystem and firms benefiting from spillover effects (for a recent review of some if this evidence, see Cohen et al ,2019<\/a>). While not necessarily tied to the innate characteristics of an area, these effects become \u2018tied\u2019 to an area, at least for some period, and give its firms and economy an advantage.<\/p>\n Sectoral specialisation: the composition of economic activity that takes place in a given location<\/strong><\/p>\n We know that different sectors of the economy have different average levels of productivity. Where an area specialises in sectors that have high productivity (for example, those in the supply chain for the oil and gas sector in the North East of Scotland), this means that the whole region will have, on average, higher productivity than other regions.<\/p>\n The ONS has undertaken work to explore this dimension of regional productivity (ONS, 2018<\/a>). Their conclusion is that these \u2018industrial mix\u2019 differences are generally not an important factor in driving headline differences in regional labour productivity.<\/p>\n More important, the ONS analysis concludes, are differences between how productive firms within each sector are in different regions on average \u2013 the so-called firm productivity effect (see Figure 3). In essence, London and the South East have many of the most productive firms in many sectors. This drives up their overall labour productivity levels.<\/p>\n In plain terms, it is not that higher productivity regions are more dominant in particular high productivity sectors, but that some regions, like London, have more productive firms in many sectors.<\/p>\n We can see this even more clearly using ONS data on how productive firms are across regions (see Figure 4). While there are a different number of firms in each region, by plotting the distribution, we see the likelihood that a firm in each region has a given level of productivity.<\/p>\n In Figure 4, the area under the line for each region sums to 1, and the larger the gap between the horizontal axis and each line, the more firms there are in that region that have that level of labour productivity. So if we picked a business at random from each region, these plots show us where the level of labour productivity of that business would be most likely to be located (that is, around where the gap between the horizontal axis and the line for that region is largest).<\/p>\n London is the clear outlier, with a larger proportion of its firms (and hence a greater proportion of the area under the black line in Figure 4) with high levels of labour productivity.<\/p>\n Aside from that, the general shape and distribution of firm-level productivity in the different parts of the UK are fairly similar \u2013 and have not changed much over the recent decade. All parts of the UK have high productivity firms and very low productivity firms: what differs is the balance between these groups.<\/p>\n Research by the ONS to explain the differences shown in Figure 3 highlights the importance of key firm-level characteristics in explaining productivity levels. These include whether a firm trades internationally, its management practices, its ownership, its age and its size<\/a>.<\/p>\n How reliable is the evidence?\u00a0<\/strong><\/p>\n The biggest obstacle to a better understanding of regional differences in productivity is a shortage of data.<\/p>\n It was mentioned above that there is an absence of regional level capital stock data in the UK. This prevents us being able to quantify the contribution that differences in the capital stock between regions play in explaining overall differences in productivity.<\/p>\n There are other data challenges. As highlighted in Koop et al (2020)<\/a>, understanding of regional differences in productivity performance for different parts of the UK can be fundamentally altered depending on the underlying data that is used. This is clearest when we consider the case of Scotland.<\/p>\n Annual output data are produced by the ONS, and quarterly and annual output data for Scotland are produced by the Scottish government. Both of these publications are national statistics. Yet after adjusting for inflation, they tell a very different story about Scotland\u2019s output and productivity performance over the past 20 years.<\/p>\n The hours worked series used to produce the ONS and Scottish government productivity series are basically the same (bar differences in smoothing out the series), as can be seen in Figure 5. But the output series that the ONS produces (solid red line) has Scottish GDP growing faster over this period than the Scottish government series (solid pink line).<\/p>\n When Koop et al (2020)<\/a> construct a productivity series using the ONS output data for Scotland, the effect is that this series shows that labour productivity in Scotland grew more quickly than the Scottish government data suggest \u2013 indeed, faster than the UK as a whole over this period.<\/p>\n These differences in economic narrative are driven by differences in methodology for calculating economic output for Scotland between the ONS and Scottish government. This is an area where more work is needed to reconcile these differences.<\/p>\n Thinking about the impact of the pandemic on regional productivity, there are going to be a number of factors that drive differences in productivity across regions. Perhaps some of the biggest will be how the following (mostly national) factors play out:<\/p>\n Changing sectoral composition?<\/strong><\/p>\n How the sectoral composition of the economy changes \u2013 in particular, will a decline in traditional retail, hospitality and tourism (relatively low productivity sectors) boost average productivity differently across regions?<\/p>\n This Economics Observatory article on Which firms and industries have been most affected by Covid-19?<\/a> is worth a read on this, as well as articles about how hard coronavirus will hit Scotland's economy<\/a> and the Northern Irish economy<\/a>?<\/p>\n A business shake-out?<\/strong><\/p>\n Will be see a big shake-out in the business base, and to what extent will the weakest performing businesses cease trading as the UK government support packages are wound down, boosting average productivity?<\/p>\n Will we perhaps see less concentration of high productivity jobs in London and the South East given the need for social distancing and changes in methods and patterns of work, eroding the advantage that these regions currently enjoy?<\/p>\n This Economics Observatory article looks at who can work from home and how it affects their productivity: Who can work from home and how does it affect their productivity?<\/a>.<\/p>\n A labour market shake-out?<\/strong><\/p>\n How big a shake-out in the labour market will we see, and to what extent will we see reallocation of labour between firms? How extensive will unemployment be \u2013 or looked at differently, how successful has the support keeping workers and firms in employment relationships been?<\/p>\n You can read more about the labour market effects of the pandemic here: What are the effects of coronavirus on the UK and US labour markets?<\/a><\/p>\n Policy support for firms and households<\/strong><\/p>\n How will the policy response, particularly across the devolved nations, support businesses and individuals through the immediate period ahead? Will some do a better job of getting people into education and training?<\/p>\n You can read about whether labour market policies can help get people back into jobs in this Economic Observatory article<\/a>. Will some \u2018build back better\u2019 than others? Will longer lockdowns or stricter pandemic restrictions affect the speed and scale of the economic recovery?<\/p>\n The levelling up agenda<\/strong><\/p>\n Where does this pandemic and the economic response leave the UK government\u2019s \u2018levelling up\u2019 agenda?<\/p>\n There have been a number of articles on the Economics Observatory picking up on how the pandemic is likely to affect different parts of the UK, including: Which parts of the UK have been hit hardest by the Covid-19 crisis?<\/a>, How will the economic effects of coronavirus vary across areas of the UK?<\/a>, Why has coronavirus affected cities more than rural areas?<\/a>, and articles on the impact of Covid-19 on the economies of Scotland<\/a> and Northern Ireland<\/a>.<\/p>\n But to date there is little sense of whether these messages about differential spatial impacts from Covid-19 will lead to structural changes in rates of productivity across the UK\u2019s regions and nations. There has been a lot of talk about levelling up and rebalancing the UK economy. But this has been a policy goal of successive administrations for years. Time will tell if this time is different.<\/p>\n A recent VoxEU article reviewing the thoughts of some leading economists on the UK\u2019s productivity slowdown \u2013 Explaining the UK\u2019s productivity slowdown: Views of leading economists<\/a> \u2013 notes that \u2018regional investment policies could be part of the policy solution\u2019.<\/p>\n Robert Zymek and Ben Jones undertook a detailed evidence review for the Industrial Strategy Council in 2020, which is essential reading in this area: UK regional productivity differences: an evidence review<\/a><\/p>\n Gary Koop, Stuart McIntyre, James Mitchell and Aubrey Poon have combined their quarterly NUTS1 level economic output data with quarterly hours worked data from the ONS to provide a quarterly labour productivity series at the NUTS1 level in the UK from 1997: Reconciled estimates and nowcasts of regional output in the UK<\/a><\/p>\n While this article has focused on NUTS1 regions of the UK, recent work by Beatty and Fothergill (2019)<\/a> takes this down to a spatial level and considers productivity at a city and sub-region level.<\/p>\nWhat does the evidence tell us?<\/h3>\n
Figure 1: Regional labour productivity across the UK relative to the UK as a whole, 2018<\/h4>\n
![\"Figure](\"http:\/\/www.economicsobservatory.com\/wp-content\/uploads\/2020\/12\/Regional-Labour-Productivity.jpg\")
<\/p>\nFigure 2: Analysis of London\u2019s labour productivity advantage over the UK as a whole in 2017<\/h4>\n
![\"Figure](\"http:\/\/www.economicsobservatory.com\/wp-content\/uploads\/2020\/12\/London-Labour-Productivity.jpg\")
<\/p>\nSource:\u00a0Beatty and Fothergill, 2019<\/a><\/h5>\n
Figure 3: Firm productivity and industry mix effects on aggregate average productivity, 2015<\/h4>\n
![\"Figure](\"http:\/\/www.economicsobservatory.com\/wp-content\/uploads\/2020\/12\/Firm-Productivity-and-Industry-Mix-Effect.jpg\")
<\/p>\nSource: ONS<\/a><\/h5>\n
Figure 4: Distribution of firm-level productivity (gross value added per worker), 2015<\/h4>\n
![\"Figure](\"http:\/\/www.economicsobservatory.com\/wp-content\/uploads\/2020\/12\/Distribution-of-firm-level-activity1.jpg\")
<\/p>\nSource: Annual Business Survey (ABS) 2015, Office for National Statistics<\/h5>\n
Figure 5: Output, hours worked and productivity series Scotland<\/h4>\n
![\"Figure](\"http:\/\/www.economicsobservatory.com\/wp-content\/uploads\/2020\/12\/Output-hours-worked1.jpg\")
<\/p>\nSource:\u00a0from Figure 10 of Koop et al, 2020<\/a><\/h5>\n
What will drive post-Covid-19 regional productivity differences?<\/h3>\n
Where can I find out more?<\/h3>\n
Who are experts on this question?<\/h3>\n
\n
Authors: Graeme Roy<\/a>, Stuart McIntyre<\/a><\/h5>\n","protected":false},"featured_media":2966,"template":"","categories":[226,235],"tags":[],"yoast_head":"\n