2. How COVID-19 could accelerate local labour market transitions

COVID-19 is a tsunami on top of broader economic, social and demographic shifts already reshaping local labour markets. Digitalisation and automation; globalisation; climate change and the green transition; and demographic changes (population ageing, migration, urbanisation) are changing the nature and location of jobs, as well as the composition and skills of the workforce. In most cases, COVID-19 will reinforce these trends, accelerating the need for a rapid policy response. However, for some others, it could re-orient them in new directions (see Table 2.1).

Even before COVID-19, many felt anxiety about the future, as the pace and scale of these changes was disconcerting. Researchers from the McKinsey Global Institute estimate that the changes that will result from megatrends such as urbanisation, technological change, and population ageing are 10 times faster and 300 times the scale compared to the changes of the Industrial Revolution (Dobbs, Manyika and Woetzel, 2015[1]), and 61% of the general population in 28 countries think the pace of change in technology is too fast (Edelman, 2020[2]). Already, 13% percent of all workers in the United States are employed in types of jobs that did not exist in 1970 (Autor and Salomons, 2019[3]). Job stability has decreased in the majority of OECD countries over the past two decades (once accounting for ageing of the workforce), with the less-educated, including both older and younger workers, particularly affected (OECD, 2019[4]). Skills imbalances have also been widening over the last decade, with growing shortages of high-level cognitive and soft skills, and increasing surpluses of routine and physical skills (OECD, 2018[5]). Yet, many of these changes also have positives. People are living longer and healthier lives. Technological change is making workplaces safer and more productive, and creating complementary new jobs.

While these transitions are almost universal, they are not uniform across places. Some of these changes, such as population ageing, will affect almost all communities, although they will be more pronounced in some places than others. Other changes will shift the geography of jobs and skills, potentially deepening the feeling that there are “winners” and “losers” in tomorrow’s economy. National aggregates can overlook these difficult transitions for communities and the people that live there, as the people that lose jobs may not be in the right location or have the right skills for the new jobs created.

Automation and digitalisation, globalisation, and climate change and the green transition will create and destroy jobs, but not necessarily in the same places or for the same set of skills.1 Already, jobs have become more geographically concentrated in most OECD countries in recent years. Pre-COVID-19, estimates from the McKinsey Global Institute suggested that most net job growth in the United States and Europe through 2030 will be concentrated in a few urban areas, further deepening existing divides. Just 25 cities, high-growth hubs, and their peripheries are predicted to account for 60% of job growth in the United States. While jobs in healthcare will be added nationwide, job growth in other occupations, such as STEM, creative fields, and business and legal professionals, will be more geographically concentrated. In Europe, 48 megacities and superstar hubs are predicted to capture 50% of job growth (Smit et al., 2020[6]; Lund et al., 2019[7]).

COVID-19 could transform these ongoing shifts into abrupt changes, making the transitional period even more difficult for some people and places. While involuntary job losses typically account for a relatively small share of overall labour market churn (see Box 2.1), the number has been rising as a result of COVID-19 containment measures and the more general downturn. There could also be an increase in the number of mass layoffs at the firm and sector level, as industries such as transportation, retail, tourism and hospitality struggle to remain viable. The story is not all negative, however, as COVID-19 could also open up new opportunities for job creation outside of traditional high-growth, urban centres.

Technological changes, from industrial robots to artificial intelligence and ongoing digitalisation, are reshaping labour markets and the geography of jobs. They are replacing specific work tasks or entire jobs, shifting the occupation structural of the labour market and the skills in demand (see Box 2.2). They are also boosting labour productivity and leading to the creation of new jobs that are complementary to these technologies. They are also creating new opportunities to decentralise jobs, production and public services, thanks to the rise of telecommuting, new production technologies, and e-services.

Almost half of jobs across the OECD are expected to change as a result of automation: 32% could see significant changes, while an additional 14% are at a high risk of automation all together (Nedelkoska and Quintini, 2018[18]).2 (“High risk of automation” refers to a 70% or above risk of automation, while “significant risk of change” reflects a risk of automation between 50% and 70%.) Manufacturing and agriculture have the highest share of jobs at risk on average. Comparatively, only a few service sectors – e.g. postal and courier services, land transport and food services – face relatively high risks. In contrast, the sectors with the lowest relative risks are predominantly service sectors, including many knowledge-intensive services. Across sectors, the occupations at the highest risk tend to be those that do not require specific skills or training – food preparation assistants, assemblers, labourers, refuse workers, cleaners and helpers – followed by occupations that require at least some training and include interacting with machines, mainly in the manufacturing sector (machine operators, drivers and mobile plant operators, workers in the processing industry, skilled agricultural workers, metal and machine workers etc.). Generally, the risk of automation decreases as the skill level of jobs increases (Nedelkoska and Quintini, 2018[18]).

COVID-19 will likely accelerate automation, as firms turn to labour replacing technologies to respond to sanitary requirements and labour shortages resulting from containment measures (Field and Murphy, 2020[19]). Longer-term social distancing requirements, as well as broader shifts in business and risk management strategies, could further the uptake of automation. Already in the February/March 2020 EY Global Capital Confidence Barometer, 36% of high level executives across the world said they were accelerating investment in automation as a result of COVID-19, and a further 41% said they were currently re-evaluating their strategies in this area (EY, 2020[20]). For example, automation is anticipated to rapidly increase in the retail sector, impacting warehouse and delivery operations (e.g. use of drones and robots in fulfilment centres); e-commerce (e.g. customer marketing, order tracking); as well as brick and mortar locations (e.g. robot cleaners, automatic check-out) (Sillitoe, 2020[21]). Some of these changes could occur relatively quickly, while other more capital-intensive investments may take several years.

Downturns in general have also been shown to accelerate automation (Muro, Maxim and Whiton, 2020[22]). In previous recessions, employers have shed less-skilled workers and replaced them with technology and complementary higher-skilled workers, increasing labour productivity. Over the past three decades, 88% of job losses in routine occupations in the United States took place following a recession, and these jobs were unlikely to be recovered post-recession (Jaimovich and Siu, 2020[23]). Similarly, other research has found that firms in metro areas the hardest hit by the Great Recession tended to replace workers who performed automatable and routine tasks with a mix of technology and higher-skilled workers (Hershbein and Kahn, 2018[24]). While similar trends have been found in Canada (Blit, 2020[25]), different patterns have been found in other countries (Graetz and Michaels, 2017[26]). This suggests that the relationship between technology and jobless recoveries deserves further study in different national contexts (Jaimovich and Siu, 2020[23]).

Across countries, Scandinavian countries tend to have a lower share of jobs at risk, while higher shares can be found in some Eastern and Southern European countries (e.g. Slovak Republic, Lithuania, Greece, Spain). While both sectoral and occupational structures contribute to national differences in automation risk, most of the differences across countries results from the fact that countries have very different occupational mixes within sectors. Even within occupations, the types of tasks performed vary across countries, changing automation risks (Nedelkoska and Quintini, 2018[18]).

In addition to national differences, regional differences within countries in the share of jobs at high risk of automation can reach 10 percentage points (Slovak Republic) or be as low as 1 percentage point (Norway). Across OECD regions, the share of jobs at high risk of automation reaches nearly 40% in some regions (for example, West Slovakia) but can be as low as around 4% in others (the region around Oslo). However, even these figures underestimate how automation will vary across communities, as the differences between communities within regions can be stark, as described in Box 2.3 for Canada.

Regions already struggling with other labour market challenges tend to have a higher share of jobs at risk. Regions that have a highly-educated workforce and a strong tradable services sector, and that are more urbanised, have fewer jobs at high risk of automation. Regions that have low productivity growth and high unemployment tend to have higher shares of jobs at risk (OECD, 2018[27]). Additionally, some places face the risk of a double hit from both accelerated automaton and direct COVID-19-related job losses (see Box 2.4).

The number of local jobs that are at risk can be daunting. For example, in the Basque Country (Spain), over 200 000 jobs are at high risk of automation, with a large share in the industrial sector (OECD, forthcoming[29]). Dealing with a challenge of this scale will require training at a large scale to help workers transition to other jobs within or across sectors, as well as efforts to transition entire local sectors to higher value-added production and services. Additionally, as described in Box 2.4, some regions risk facing a double hit from automation and COVID-19.

Both local and national factors will determine how these trends play out, as a high share of jobs at risk is not destiny. While many communities that previously specialized in traditional manufacturing activities have struggled with such large-scale structural changes in the past, others have bucked the trend. Akron, Ohio; Albany, New York, and Pittsburg, Pennsylvania in the United States, as well as Dresden, Germany and Eindhoven, Netherlands have been highlighted as success stories. These older industrial cities managed to transition from traditional manufacturing to centres of advanced industrial production (e.g. in polymers, nanotechnologies, and semiconductors) and innovation, in part due to their collaborative and multidisciplinary approach to innovation (van Agtmael and Bakker, 2016[30]). OECD research on places undergoing industrial transition similarly suggests the importance of fostering “high-road competitiveness” strategies, built around innovation-led growth and that broadly share the benefits of this growth across people and places (OECD, 2019[31]).

Broader, national institutional settings will also play an important role in how these trends play out in different communities. In looking at the adoption of industrial robots, local labour markets in different countries appear to have responded differently. In the United States, the local uptake of industrial robots has led to declines in the employment rates and wages within the local commuting zone (Acemoglu and Restrepo, 2017[32]). However, in studying the adoption of industrial robots in local labour markets in Germany, local job losses in manufacturing were offset by gains in the business service sectors. Manufacturing job losses were not due to displacement of incumbent workers, who tended to take on new roles within their organisation, but rather fewer manufacturing jobs for new labour market entrants (Dauth et al., 2018[33]). While further study is needed, mitigating factors could include different labour market institutions, regulations and traditions across countries.

Some forces will contribute to an increasing concentration of jobs in urban areas, but other forces – especially in relation to COVID-19 – could push jobs to deconcentrate. Generally, cities that already had a highly-skilled labour force disproportionately benefited from past waves of technological change. They have been able to attract more high-skill jobs and workers (OECD, 2019[17]; OECD, 2018[27]), as well as reap the spillover effects in terms of creating other local jobs (e.g. as a result of increased demand for local services such as restaurants, hairdressers, etc.). Although there is debate on the scale of these effects, evidence generally shows that new high-tech, skilled or tradeable sector jobs have larger spillovers for local economies than other types of jobs (Moretti, 2012[40]).3

While similar patterns could repeat themselves this time around, new technologies and digitalisation could also create opportunities to decentralise some jobs. For example, some argue that 3D printing could help to decentralise some elements of production, giving more opportunities to places outside of traditional high growth centres (see Box 2.5). Evidence also suggests varying urban/rural patterns for different types of new jobs emerging from changing technology, shifting tastes, and rising incomes (Autor and Salomons, 2019[3]). Jobs related to producing, installing, maintaining, and deploying new technologies (i.e. “frontier” work) and providing in-person services for affluent consumers (i.e. “wealth work”) have concentrated in denser, urban labour markets in the United States. In contrast, last-mile jobs (carrying out nearly-automated tasks that retain only a residual set of human components) are somewhat less prevalent in urban than in non-urban areas, as many do not require face-to-face interactions.

Additionally, if COVID-19 indeed sparks a longer-term trend towards increased teleworking, there could be a further dispersion of jobs. Many employers had to quickly invest in cloud technologies and other digital tools to pivot to teleworking during strict confinement periods. This rapid pivoting could open up the door for a broader adoption of teleworking over the long term. Many large firms, particularly in the tech sector, have already announced plans to significantly expand teleworking over the long term, or even permanently (Sandler, 2020[41]). Cities have a larger share of jobs amenable to teleworking compared to smaller towns, villages and rural areas (see Chapter 1), suggesting that a move to more remote working could impact jobs traditionally performed in cities the most.

A major impact of technological change has been a declining share of routine, middle-skill jobs, i.e. job polarisation, across most OECD countries, sectors, and regions (OECD, 2017[42]). Middle-skill jobs, such as clerical and production jobs, typically entail routine manual or cognitive tasks and are considered easier to automate given the current state of technological developments. On the other hand, low-skill jobs tend to involve non-routine manual tasks, for example requiring manual dexterity. High-skill jobs, such as managerial and professional occupations, are considered to be complemented, rather than substituted, by new technologies (Autor, Levy and Murnane, 2003[43]). Polarisation has previously been documented by others in the United States (Autor, Katz and Kearney, 2006[44]) and Europe (Goos, Manning and Salomons, 2009[45]). In most cases, this loss of middle-skill jobs has been accompanied by an increase in high-skill jobs. A large share of this polarisation is the result of in-sector shifts (see Box 2.6).

Almost all OECD regions have seen a decrease in the share of middle-skill jobs since 2000. In over a quarter of regions, the share of middle-skill jobs decreased by 10 percentage points or more. Only Chile, the Czech Republic, Denmark, Korea, and the United States have one or more regions where the share of middle-skill jobs increased. In most of these regions, the increase has been relatively small (1 to 2 percentage points or less). While polarisation is generally accompanied by upskilling, in about one out of six regions, it was accompanied by downskilling (i.e. growth in the share of low-skill jobs outpaced growth in high-skill jobs). Over half of countries with available data have at least one region in this situation (see Annex Table 2.A.1).

Particularly large regional variations in the scale of polarisation can be found in countries hard hit by the 2008 crisis (Greece, Portugal, Italy, Spain), Eastern European countries (Poland and Romania), as well as in the United States (see Figure 2.5). For example, in Italy, the decrease in the share of middle-skill jobs ranged from 13 percentage points in Marche to 0.2 percentage points in Calabria. In Israel, Slovenia, Greece, France, and Finland, the capital region experienced the smallest percent point change in share of middle-skill jobs, while in Austria, Belgium and Poland, the capital region experienced the largest change. Regions with high initial shares of employment in middle-skill jobs tended to experience the strongest polarisation over the time period considered.4 This suggests that polarisation is a pervasive phenomenon that all places will have to contend with eventually, if they have not already.

Recent studies within countries also find that job polarisation has been more pronounced in large, urban areas, even if the initial share of middle-skill jobs is relatively low.5 Research in France highlights the important distinction between different types of urban areas. Nearly all French cities saw the share of middle-skill jobs decline, but large cities saw a sharper contraction of middle-skill jobs and a shift towards high-skill jobs, i.e. they “upskilled”, whereas smaller cities shifted towards low-skill jobs, i.e. they “downskilled” (Davis, Mengus and Michalski, 2020[50]). As regions typically combine cities of different sizes there is no clear relationship between the level of urbanisation at the TL2 level and polarisation over time in this analysis. Having a larger share of the population living in urban areas is indeed positively related with the shift towards more high-skill employment, and relatively fewer middle-skill jobs in 2018.

Polarisation, at least in European OECD countries, is predominantly linked to changing labour market opportunities for new labour market entrants, including declining opportunities for those without a tertiary degree compared to previous cohorts (OECD, 2020[51]). Recent OECD research suggests that new labour market entrants are now less likely to hold a middle-skill job relative to a low- or high-skill job than in the past, and workers without a tertiary degree are now more likely to be employed in low-skill occupations. The share of women without a tertiary degree in low-skilled jobs in particular has grown. However, those middle-skill workers who do lose their jobs may find it difficult to transition to a comparative job, particularly in places where job losses are part of larger structural transitions or local labour market shocks (such as the closure of large manufacturing plants). Additionally, the transition from middle-skill to higher-skill occupations requires up-skilling in terms of both cognitive and task-based skills, making it easier for workers in the middle of the skill distribution to move to lower-skill occupations rather than high-skill occupations as a result of a job loss (Bechichi et al., 2018[52]).

Polarisation could also have implications for local labour market resilience, particularly in light of COVID-19. Emerging evidence on the geography of COVID-19 related job losses suggests that low-skill service workers are particularly vulnerable to job losses in wealthy urban areas, i.e. places with more polarised labour markets. In such places, a higher share of low-skill jobs are dependent on the discretionary spending of local high-income earners, which has been slower to recover than spending of other income groups (Chetty et al., 2020[53]). Additionally, the rising shares of women without a tertiary degree working in low-skill jobs (OECD, 2020[51]) may be linked to the disproportionate share of COVID-19-related job losses they have experienced.

Globalisation has brought many benefits – productivity improvements, technological and innovation diffusion, opening up of new consumer markets, and lowered costs of goods and services. At the regional level, specialisation in tradeable sectors has been an important factor helping lagging places catch up and bounce back from the shock of the global financial crisis (OECD, 2016[54]; 2018[55]).

Yet, globalisation has also generated considerable public anxiety, and trade growth was stagnating even pre-COVID-19. Compared to the early 2000s, when global trade was rising at more than twice the pace as output, it has risen only marginally faster in recent years (OECD, 2020[56]), and global value chains stopped expanding about a decade ago (OECD, 2020[57]). A public backlash against globalisation has sparked a re-emergence of protectionist policies and trade tensions in recent years. (OECD, 2019[58]). While the overall scale is still limited, reshoring, or the relocalisation of manufacturing back to developed countries has been of growing importance even pre-COVID-19 due to factors such as declining cost advantage of emerging economies and the need for production to be close to markets and innovation (De Backer et al., 2016[59]). Actors in a number of countries have launched economic development strategies to help reshore some aspects of production (for example, Reshore UK). Additionally, some argue that new production technologies, such as 3D printing and automation, as well as shifting consumer demands, may further contribute to deglobalisation (Livesey, 2018[60]).

COVID-19 upended global trade in the short term, and could lead to further stagnation over the longer term. Trade decreased by over 15% in the first of 2020 (OECD, 2020[61]). Going forward, European countries are expected to face particularly sharp declines, reflecting strong cross-border relationships, the importance of tourism in some economies, and the vulnerability of commodity-exporting economies to the drop in demand (OECD, 2020[56]).

Disruptions in supply chains are causing firms and governments to reassess the risks associated with complex global supply chains more generally. Shortages in essential medical equipment or pharmaceuticals produced abroad in particular have brought these questions squarely into the public debate. This could lead to a longer-term shift of re-shoring of strategic production activities, notably in relation to priority goods in health care. Shorter food production chains may also be promoted. Already in April, for example, Japan announced a stimulus package that includes USD 2 billion to support firms in shifting production back to Japan.6

A reduction in global trade could hurt more globalised regions acutely in the short term (see Chapter 1) but also offer new opportunities. Should reshoring become a significant trend over the longer term, local economies may be able to diversify economic activity and restore some middle-skill jobs, although some argue that the impact on jobs will primarily benefit high-skilled workers (De Backer et al., 2016[59]). Re-shoring will be easier in places that are already more diversified or with strong local skills bases in related sectors or occupations. Additionally, should global trade rebound, more globalised regions may be able to bounce back more quickly, a trend seen in the last crisis.

Transition costs associated with previous trade shocks – positive or negative – have indeed been geographically concentrated in the places with high shares of jobs in the most affected sectors. In particular, the entry of China into global markets had more persistent, localised impacts than anticipated. In the United States, sectors exposed to trade competition from the entry of China into global markets tended to be geographically concentrated. Losses were not isolated to manufacturing employment directly exposed to this competition, and these local labour markets experienced longer-term and more persistent negative impacts in terms of unemployment, labour force participation rates, and wages than economists traditionally predicted (Autor, Dorn and Hanson, 2016[62]). Similar impacts for reductions in manufacturing employment have been found for Norway and Spain, although spillovers to the local labour markets more generally varied (Donoso, Martín and Minondo, 2015[63]; Balsvik, Jensen and Salvanes, 2015[64]). In Germany, the impact of Chinese import competition was attenuated by increased trade with Eastern European countries following the fall of the Iron Curtain, albeit with different impacts for import- and export-competing regions (Dauth, Findeisen and Suedekum, 2014[65]).

A few recent studies provide some insights as to how local economies may fare when globalisation retreats, rather than advances. In modelling the geographic impacts of tariff hikes put in place by the United States and its trade partners in 2018, researchers found that the Great Lakes region of the Midwest and the industrial areas of the Northeast benefitted the most from tariff protection, while the rural regions of the Midwestern plains and Mountain West faced higher tariff retaliation in the short-run (Fajgelbaum et al., 2019[66]). Some experts have suggested that Brexit will have a longer-term negative impact on the places already struggling, potentially further exacerbating existing geographic divides in the United Kingdom (Carter and Swinney, 2019[67]).

While governments have made important commitments to addressing climate change, concrete actions to invest in renewable energy and reduce our economic dependence on fossil fuels remain well below the scale needed. Accordingly, the bulk of the green transition is still on the horizon. Fully scaling up the transition to a greener economy will result in job destruction in “brown” sectors and job creation in “green sectors”, as well as macro-level impacts resulting from changes in demand patterns, GDP, etc. However, most macroeconomic models predict minimal net employment changes as a result of the green transition. The OECD estimates, for example, that job churn as a result of climate action across sectors (summing up the creation and the destruction of jobs) will only be 1.5% of total employment by 2050 (OECD, 2017[68]).

How COVID-19 will impact the green transition over the longer-term remains an open question.7 In the short term, COVID-19 containment measures and the more general downturn are expected to reduce global CO2 emissions by 8% in 2020 compared to 2019 (IEA, 2020[69]). However, this drop is only temporary, reflecting the drastic slowdown of economic activity and travel, and emissions have typically rebounded following other recession-based dips. Looking forward, COVID-19 could generate new political will to tackle this type of collective, global crisis. Behavioural changes and shifts in consumer preferences could also be longer-lasting, such as a reduction in business travel, international tourism, or daily commuting. The unprecedented stimulus packages many governments are now rolling out could also accelerate investment in green infrastructure, if properly targeted.

However, COVID-19 could also create new tensions between preserving jobs at all costs, and transitioning carbon-intensive sectors to greener production methods. For example, governments may face strong pressures to bailout struggling carbon-intensive industries, such as airlines or carmakers. These bailouts present an opportunity to make support contingent on reducing emissions, but this is not a given. Additionally, while green stimulus packages can help reorient economic development and deliver growth over the long term, some evidence suggests that they may be less efficient at creating jobs in the short term (Popp et al., 2020[70]), although further research is needed. This may make them less politically appealing than other types of stimuluses that create jobs more immediately.

Even if net employment impacts are predicted to be minimal, the green transition implies significant adjustment and transition costs at the local level, particularly for natural resource regions. Material-intensive or extractive sectors tend to cluster around specific places with natural resources or enabling infrastructure, and thus are highly geographically concentrated. In Canada, for example, nearly one-third of Alberta’s GDP and 6% of jobs are tied to the fossil fuel industry, compared to 8% and 1% respectively nationally, without even taking into account indirect or induced employment (Mertins-Kirkwood, 2018[71]).

The transition to a more circular economy implies a shift away from employment in materials-intensive activities towards service-driven activities, and from industry and primary production sectors to secondary production and services sectors. Accordingly, jobs in the circular economy are less reliant on the natural resources, and firms are more mobile to locate where they can find the right type of workers. Thus, while there is a good chance that places that specialise in material-intensive or extractive sectors will lose jobs in this transition, their ability to reap the benefits of the complementary job creation is less assured (Laubinger, Lanzi and Chateau, 2020[72]). While this research refers to countries and international regions, the same arguments can also apply within countries.

Some circular economy activities will be more geographically dispersed than others. For example, low-skill recycling and repair jobs will be needed across territories, and therefore are not likely to concentrate specifically in urban or rural areas. Others, however, rely on sufficient demand or density (e.g. specialised repair jobs, sharing economy), and are more likely to cluster in urban areas (Laubinger, Lanzi and Chateau, 2020[72]).

A study of green jobs in the United States over the period of 2006-2014 found that they were more geographically concentrated than comparable non-green jobs, but that there was some catching up effects over this time period. The places with the highest shares of green employment were wealthier and more high-tech, and were more likely to host public R&D laboratories, have more green patents per capita, and a higher-than-average share of employment in high-tech manufacturing and knowledge-intensive services (Vona, Marin and Consoli, 2018[73]). In the Netherlands, urban areas were found to have the highest density of circular jobs per square kilometre, with urban peripheries concentrating core circular jobs linked with traditional manufacturing, and city centres concentrating enabling circular jobs linked more to knowledge-intensive activities and services (Circle Economy and EHERO, n.d.[74]).

Climate change itself will also impact regional economies differently due to changes in tourism patterns, the location of agricultural production, and demand for energy. Extreme events such as hurricanes, flooding, and droughts will also have extremely localised impacts. In Australia, the drought beginning in 2017 is projected to decrease farm GDP by 30% by 2020, with the Murray–Darling Basin, which accounts for around one-third of the total value of Australia's agricultural production, severely impacted (Reserve Bank of Australia, 2020[75]). Projections for Europe suggest large parts of Southern Europe – which are dependent on tourism and agriculture – as well as the Alps (tourism) and South Eastern Europe (agriculture) are particularly sensitive to the economic changes as a result of climate change, as well as some parts of Scandinavia due to changing energy demands (Greiving, Fleischhauer and Lindner, 2013[76]). In the United States, evidence suggests that the economic impacts of climate change could further entrench existing geographic divides: the poorest 10% of counties are estimated to face economic losses 9.5 times larger than the richest 10% of counties (Hsiang et al., 2017[77]).

The low-skilled will face particular challenges in adapting to this new world of work. They are more likely to be employed in jobs vulnerable to automation and face increasing competition from middle-skill workers who have been displaced from traditional middle-skill jobs. Some research also suggests that low-skill workers will be most impacted by decarbonisation policies (Chateau, Bibas and Lanzi, 2018[78]), and as discussed in Chapter 1, they are particularly vulnerable to COVID-19-related job losses.

Relying on labour mobility to help counterbalance the shifting geography of jobs will likely not be sufficient. For one, most moves are not actually for job-related reasons. On average across OECD countries with data available, only 9% of residential moves were for job-related reasons. This compares to 41% for housing-related reasons and 34% for family-related reasons (Causa and Pichelmann, 2020[79]).8 Higher-skilled individuals and households tend to be more geographically mobile (Eurostat, 2017[80]; Causa and Pichelmann, 2020[79]). In Europe, people living in cities are also more mobile on average (Eurostat, 2017[80]), and in the United States, when people do move, they tend to move to similar communities, rather than to megacities or high-growth hubs (Lund et al., 2019[7]).

Geographic mobility also varies considerably across OECD countries. This could be the result of both cultural factors, as well as the institutional framework (housing policies, occupational licensing, other labour market regulations, etc.). While international comparisons of internal mobility are difficult to construct,9 evidence suggests that domestic residential mobility it is relatively high in Nordic countries, Australia and the United States, and relatively low in southern and eastern European countries (Causa and Pichelmann, 2020[79]; Caldera Sánchez and Andrews, 2011[81]) (although international migration is relatively more important in the latter). These findings have been generally confirmed by other research, which also finds relatively high rates of mobility in Korea and Canada as well (which were not included in the other studies) (Bell et al., 2015[82]). Even in some traditionally mobile countries, mobility is declining, including the United States (US Census Bureau, 2017[83]) and Australia (Charles-Edwards et al., 2018[84]). In the United States, the share of people moving annually has almost halved since data first started being collected over fifty years ago.

While labour mobility cannot be the only solution, varying trends over time and across countries do suggest that there is room for policy interventions. Addressing occupational licensing restrictions and housing market rigidities can help reduce existing barriers to mobility. However, further study is needed to understand the degree to which labour mobility can offset employment challenges, as some research suggests that even greater geographic mobility would only marginally reduce unemployment rates (Marinescu and Rathelot, 2018[85]).

Changes to the size and composition of local labour forces will be just as important as the demand side factors reshaping the future of work. Population ageing and shrinking, as well as mobility within and across countries, will have significant impacts on local labour markets. In fact, some experts suggest that demographic shifts will be even more important than technological changes in reshaping labour markets.10

COVID-19 is unlikely to have a significant impact on broader demographic changes such as population ageing and shrinking national labour forces, but could affect mobility patterns within and across countries. However, it remains to be seen if some of the short-term changes sparked by COVID-19 will persist, such as decreases in international migration and movement out of more urban areas.

Longevity increases and declines in birth rates have led to a general trend of population ageing in OECD countries. By 2050 it is estimated that over half of OECD countries will have a smaller working age population than in 2010. As shown in Figure 2.6, Lithuania, Latvia, Japan, Greece, Korea, Poland, Portugal, and Spain are projected to have the biggest relative decreases.

Almost 30% of TL2 regions with available data have already seen the size of their labour force decrease in the past decade (see Figure 2.7). This is despite the general trend of increasing labour force participation rates, and likely reflects both differences in local age profiles as well as inter-regional migration patterns. In Japan, Greece and Finland, more than three quarters of regions had a shrinking labour force, and in seven additional countries, between 50% and 75% of regions have seen their labour force shrink (Czech Republic, Denmark, Poland, Portugal, Slovenia, Spain and the United States).11

Natural population decline (i.e. due to declining birth rates) is further compounded by internal mobility and migration, particularly movement from rural to urban areas. Although the pace of urbanisation is decelerating in most OECD countries, people continue to move from more rural to urban areas in many countries. In almost 80% of OECD countries, the share of national populations living in urban regions has increased between 2008 and 2018. Conversely, rural and intermediate regions saw declining shares in most places (Figure 2.8).

However, population shrinkage does not strictly follow urban/rural lines. A significant share of cities are also losing population. In five countries, including Chile, Greece, Mexico, Poland, the United States, the share of people living in urban regions decreased between 2008 and 2018. Almost one in four cities (23%) with more than 50 000 inhabitants in the OECD has shrunk in population since 2000. Smaller cities (i.e. less than 250 000 residents) account for the bulk of cities losing population (OECD, 2019[87]). In contrast, in the United States, the three biggest metropolitan areas – New York City, Los Angeles and Chicago – have all registered population declines in recent years (US Census Bureau, 2019[88]).

COVID-19 could slow down or even reverse some of these trends. As discussed in Box 2.7, factors such as a rise in teleworking and the changing value of urban amenities could shift patterns in urbanisation, The movement of workers away from the largest metropolitan areas could open up opportunities for more rural communities or smaller metropolitan areas to attract residents, or lead to more growth in suburban areas. Indeed, some communities have already been pursuing teleworkers: pre-COVID-19, Vermont (US) and Tulsa, Oklahoma (US) launched programmes to offer financial incentives to attract teleworkers.12 However, any predictions about how COVID-19 could impact demographic trends is purely speculative at this point, as much remains to be seen about how the crisis and its impacts will unfold.

Even if COVID-19 decelerates urbanisation, many places will be facing labour shortages in the coming years. A vicious cycle can set in at the local level, with employers relocating their operations because they cannot find the local workers they need, more people relocating as economic opportunities decline, etc. Already in Japan, where the working age population has been shrinking for years, more than 80% of employers surveyed in a 2017 poll expect labour shortages will restrict the number of services they can provide (OECD, 2019[92]). Loss of local employers and jobs, in turn, could lead to a decrease in tax revenues to invest in infrastructure and services, in turn leading to more out-migration. Thinner labour markets, characterised by fewer workers and employers, are also thought to be have lower quality worker-employee matches and result in longer spells of unemployment (Moretti, 2011[93]).

A number of strategies can help offset such shortages: activing those currently out of the labour force, attracting new domestic or international residents, and adopting labour saving technologies. As discussed in Chapter 1, at least some of the current inactive population could and would like to work if the right supports and labour market opportunities were available to them. Addressing the barriers that prevent them from doing so, such as access to childcare or mental health care, could help close these gaps. Attracting international migrants to help offset the decline of native-born populations is another approach. As discussed later in this chapter, migrants tend to cluster in urban areas, but there are opportunities for them to help revitalise more remote areas (Galera et al., 2018[94]). Boosting productivity will also be essential. Some research already suggests that the uptake of new automation technologies, such as industrial robots, has helped offset some of the pressure of an ageing labour force in the United States (Acemoglu and Restrepo, 2018[95]).

As the average age of the workforce and the length of working lives increases, so does the risk that skills will become outdated in the face of new technologies or other labour market changes. Already, results from the OECD Survey of Adult Skills (PIAAC) show that one-third of 55 to 65 year olds have no computer experience or fail core ICT tests. Only one in ten older workers have medium to good skills related to problem solving in technology-rich environments (OECD, 2019[96]). While all places will have to find ways to support older works in keeping their skills up-to-date and relevant, it will be particularly urgent for those places where population ageing is most pronounced. In the EU, 62% of city residents have basic or above digital skills, compared to 55% of people in towns and suburbs and 48% in rural areas (Eurostat, 2020[97]).

Older workers, particularly blue-collar workers, may face specific challenges in bouncing back in terms of time to re-employment and earnings loss after being displaced from their jobs (OECD, 2018[98]), (Quintini and Venn, 2013[11]) (see Box 2.1). As their experience often has declining relevance, a significant mismatch between their skills and the new types of jobs available is a contributing factor. Older workers are also less geographically mobile than their younger counterparts on average, and thus may be less likely to move to pursue a job opportunity (Causa and Pichelmann, 2020[79]). They may also face age discrimination in hiring. Regardless of the reasons, they are particularly vulnerable to negative repercussions of job losses discussed above, and less likely to be able to benefit from the new jobs created because of skills and/or geographic mismatches.

While older workers face more risks adapting to these longer-term structural changes, younger workers have been more impacted in the short term by COVID-19-related job losses and reduction in working hours. As discussed in Chapter 1, extended periods of unemployment at a young age can leave “scarring effects” in terms of employment and wages over the longer-term. Thus, expanding access to learning and training will be important across all ages.

Life-long learning can help workers adjust to these transitions, but there are important regional differences in the rate of participation in training. The regional difference in the share of adults participating in training is above 10 percentage points in a number of countries, including Australia, Colombia, Denmark, Italy, Slovak Republic, Sweden, and Switzerland. However, further study is needed to see if these variations are just artefacts of different skill composition or age profiles of local labour forces, or rather reflect other regional differences in access to training.

Rising levels of education are a general trend across the OECD: in 2000, 22% of adults in the OECD had completed tertiary education, while in 2018, 37% had. In several countries, the share of adults with a tertiary education doubled over this period (e.g. Czech Republic, Ireland, Italy, Korea, Luxembourg, Poland, Portugal, Slovak Republic, Slovenia, and Turkey) (OECD, 2020[99]).

This trend holds true across virtually all regions, but some places have benefited more than others. In 2018, the share of tertiary-educated adults in the best performing region was almost double (1.9) that of the worst performing region on average within OECD countries with data available. Additionally, as discussed in Box 2.8, some places struggle to put high-skilled workers to good use. Employment rates for tertiary educated adults can vary by as much as 10 percentage points across regions within countries. Even for those who are employed, the OECD estimates that over one-fifth of workers are overqualified for their jobs, a rate that can exceed 30% in some regions (see Box 2.8).

Urban areas, and in particular capital regions, tend to have a more highly-educated population. In almost all countries with available data (25 out of 27), capital regions had the most highly educated population in their respective countries. In most countries, places that already had relatively high shares of tertiary-educated adults in 2008 saw a greater increase over the following decade, resulting in an increasing gap between the best and worst performing regions, with Norway, Belgium and Latvia as the only exceptions (Figure 2.11). In contrast, regional differences in the share of the population with at least an upper secondary education have generally declined over the past 15 years, an education level reached by almost 79% of the adult population in the OECD (OECD, 2018[100]). This suggests that the geographic concentration of skills is most pronounced at the highest skills level.

A number of factors may contribute to these regional disparities. For one, urban areas tend to attract young people, students and the highly skilled because of education and employment opportunities as well as the amenities. In most OECD countries, almost all within country youth migration (95%) is directed towards metropolitan regions (OECD, 2020[101]). OECD PISA data suggests that the quality of initial education may vary across urban and rural areas.13 In almost two-thirds of OECD countries with available data, urban students outperform rural students in science, although these differences disappear once socio-economic conditions are taken into account, suggesting that socio-economic factors play a bigger role than any inherent urban/rural divides. Urban and rural students may also have different educational aspirations. On average across the OECD, only 30% of students in rural schools expect to complete at least a university degree, compared to nearly half of the students in urban schools. Unlike gaps in performance, these gaps in aspirations do not disappear once socio-economics are taken into account (Echazarra and Radinger, 2019[102]).

For a number of OECD countries, emigration, i.e. people moving abroad, is also a significant factor in shaping local skills supply. Nine OECD countries had 10% or more of their population living abroad in 2015/16 (Figure 2.14). For three-quarters of OECD countries, the highly skilled are more likely to emigrate. In some countries, high rates of emigration are accompanied by high rates of immigration (e.g. New Zealand, Luxembourg) while in others, population flows are more unidirectional (e.g. Portugal, Lithuania, Mexico). In Europe, the share of high-skill EU movers in the total EU employed population tripled between 2004 and 2016. Poland, Slovakia, Estonia, Bulgaria, Croatia, Latvia, Portugal, Lithuania and Romania were the main outgoing countries (i.e. high proportion of citizens living in another EU country and low proportion of other EU citizens living there) (ICF, 2018[107]). However, the countries facing the biggest “brain drain” are small, developing countries outside of the OECD (OECD/AFD, 2019[108]).

As discussed in Box 2.7, COVID-19 could decelerate the concentration of skilled workers in cities. Indeed, there is reason to think that high-skilled workers may be the group most likely to move as a result of COVID-19. They are generally more geographically mobile, and are more likely to hold jobs compatible with teleworking. Additionally, should universities expand online learning more permanently, young people may be less likely to move to urban centres for higher education.

Approximately one in ten OECD residents are foreign-born, and in 2018, 5.3 million new permanent immigrants arrived in OECD countries. Since 2000, the immigrant population has increased across OECD countries, with only a few exceptions (Estonia, Israel, Lithuania, Latvia and Poland) (OECD, 2019[109]). The number of tertiary-educated immigrants in OECD countries more than doubled between 2000/01 and 2015/16 (OECD/Bertelsmann Stiftung, 2019[110]) and in 2015/16, there were more tertiary-educated immigrants in OECD countries than low-educated immigrants, a reversal of the figures from 2000/01 (OECD/AFD, 2019[108]). While migration flows can vary over time as a result of changes in the political and economic context, prior to COVID-19, there was no reason to think migration flows would slow or reduce over the long term.

However, COVID-19 has disrupted international migration, at least in the short term (OECD, 2020[111]). As part of containment measures, countries put unprecedented restrictions on the international movement of people, with border closures a common feature of national responses during the strictest lockdown measures. While strict measures are easing in some places, this immediate response to COVID-19 could reduce the openness of many communities to international visitors and migrants, which could translate into a desire over the longer-term to be less open. Additionally, international moves may become less desirable for students, for example if universities move to online classes over the longer-term, as well as to highly-skilled workers. How these dynamics will play out over the longer-term, and the impact for local communities, remains to be seen.

Approximately two-thirds of migrants live in metropolitan regions (OECD, 2018[112]). Between 2005 and 2015, areas with larger existing migrant communities also experienced the greatest increases in the population share of migrants (Diaz Ramirez et al., 2018[113]). Highly skilled migrants are also more likely to settle in regions with a more highly skilled native-born population.

Accordingly, cities have historically concentrated the challenges and opportunities associated with migration. Integrating migrants can present important challenges related to language, skills recognition and mismatches, etc, but also opportunities to invigorate local labour markets with new talents and skills. However, the experience and success of integrating migrants and their families in cities can vary considerable across and within countries, and even within cities across neighbourhoods (Crul and Mollenkopf, 2012[114]). Additionally, for many places, immigration has been important in offsetting declines in native-born populations, particularly as the age profile of immigrants tends to skew younger. Should migration patterns change, cities will feel the impact most strongly.

As a result of COVID-19, the jobs of tomorrow may be coming sooner than anticipated. The task at hand hasn’t so much radically changed, but rather became more pressing in terms of scale and scope: how to ensure the short-term livelihoods of people, firms, and entire communities, while still keeping an eye on these broader transitions. Sacrificing longer-term economic resiliency for short-term gains would be a mistake, but this is a hard argument to make when unemployment is spiking, and finding a job – any job – is a first order priority for many. The following chapter – Chapter 3 – provides recommendations for how national and local actors alike can respond to short-term needs in relation to the COVID-19 economy, but in a way that builds longer-term local resilience.

References

[95] Acemoglu, D. and P. Restrepo (2018), Demographics and Automation, National Bureau of Economic Research, Cambridge, MA, http://dx.doi.org/10.3386/w24421.

[32] Acemoglu, D. and P. Restrepo (2017), Robots and Jobs: Evidence from US Labor Markets, National Bureau of Economic Research, Cambridge, MA, http://dx.doi.org/10.3386/w23285.

[105] Adalet McGowan, M. and D. Andrews (2015), “Labour Market Mismatch and Labour Productivity: Evidence from PIAAC Data”, OECD Economics Department Working Papers, No. 1209, OECD Publishing, Paris, https://dx.doi.org/10.1787/5js1pzx1r2kb-en.

[120] Autor, D. (2019), “Work of the Past, Work of the Future”, AEA Papers and Proceedings, Vol. 109, pp. 1-32, http://dx.doi.org/10.1257/pandp.20191110.

[62] Autor, D., D. Dorn and G. Hanson (2016), “The China Shock: Learning from Labor-Market Adjustment to Large Changes in Trade”, Annual Review of Economics, http://dx.doi.org/10.1146/annurev-economics-080315-015041.

[44] Autor, D., L. Katz and M. Kearney (2006), The Polarization of the U.S. Labor Market, National Bureau of Economic Research, Cambridge, MA, http://dx.doi.org/10.3386/w11986.

[43] Autor, D., F. Levy and R. Murnane (2003), “The Skill Content of Recent Technological Change: An Empirical Exploration”, The Quarterly Journal of Economics, Vol. 118/4, pp. 1279-1333, http://dx.doi.org/10.1162/003355303322552801.

[3] Autor, D. and A. Salomons (2019), “New Frontiers: The Evolving Content and Geography of New Work in the 20th Century”, http://conference.nber.org/conf_papers/f129906.pdf (accessed on 7 March 2020).

[64] Balsvik, R., S. Jensen and K. Salvanes (2015), “Made in China, sold in Norway: Local labor market effects of an import shock”, Journal of Public Economics, Vol. 127, pp. 137-144, http://dx.doi.org/10.1016/j.jpubeco.2014.08.006.

[116] Bartik, T. and N. Sotherland (2019), Local Job Multipliers in the United States: Variation with Local Characteristics and with High-Tech Shocks, W.E. Upjohn Institute, http://dx.doi.org/10.17848/wp19-301.

[52] Bechichi, N. et al. (2018), “Moving between jobs: An analysis of occupation distances and skill needs”, OECD Science, Technology and Industry Policy Papers, No. 52, OECD Publishing, Paris, https://dx.doi.org/10.1787/d35017ee-en.

[82] Bell, M. et al. (2015), “Internal Migration and Development: Comparing Migration Intensities Around the World”, Population and Development Review, Vol. 41/1, pp. 33-58, http://dx.doi.org/10.1111/j.1728-4457.2015.00025.x.

[35] Berman, B. (2012), “3-D printing: The new industrial revolution”, Business Horizons, Vol. 55/2, pp. 155-162, http://dx.doi.org/10.1016/j.bushor.2011.11.003.

[36] Birtchnell, T. and J. Urry (2013), “3D, SF and the future”, Futures, Vol. 50, pp. 25-34, http://dx.doi.org/10.1016/j.futures.2013.03.005.

[25] Blit, J. (2020), “Automation and Reallocation: Will COVID-19 Usher in the Future of Work?”, Canadian Public Policy, Vol. 46/S2, pp. S192-S202, http://dx.doi.org/10.3138/cpp.2020-065.

[38] Bonnín Roca, J. et al. (2017), Getting Past the Hype About 3-D Printing, https://sloanreview.mit.edu/article/getting-past-the-hype-about-3-d-printing/ (accessed on 26 March 2020).

[81] Caldera Sánchez, A. and D. Andrews (2011), “Residential Mobility and Public Policy in OECD Countries”, OECD Journal: Economic Studies, Vol. 2011/1, https://dx.doi.org/10.1787/eco_studies-2011-5kg0vswqt240.

[67] Carter, A. and P. Swinney (2019), “Brexit and the Future of the UK’s Unbalanced Economic Geography”, The Political Quarterly, Vol. 90/S2, pp. 72-83, http://dx.doi.org/10.1111/1467-923X.12649.

[79] Causa, O. and J. Pichelmann (2020), “Should I stay or should I go? Housing and residential mobility across OECD countries”, OECD Economics Department Working Papers, No. 1626, OECD Publishing, Paris, https://dx.doi.org/10.1787/d91329c2-en.

[84] Charles-Edwards, E. et al. (2018), Population Shift: Understanding Internal Migration in Australia, Census of Population and Housing: Reflecting Australia - Stories from the Census, 2016, https://www.abs.gov.au/ausstats/[email protected]/Lookup/by%20Subject/2071.0~2016~Main%20Features~Population%20Shift:%20Understanding%20Internal%20Migration%20in%20Australia~69 (accessed on 27 March 2020).

[78] Chateau, J., R. Bibas and E. Lanzi (2018), “Impacts of Green Growth Policies on Labour Markets and Wage Income Distribution: A General Equilibrium Application to Climate and Energy Policies”, OECD Environment Working Papers, No. 137, OECD Publishing, Paris, https://dx.doi.org/10.1787/ea3696f4-en.

[34] Chernoff, A. and C. Warman (2020), “COVID-19 and Implications for Automation”, NBER Working Paper, No. 27249, http://www.nber.org/papers/w27249 (accessed on 12 August 2020).

[53] Chetty, R. et al. (2020), How Did COVID-19 and Stabilization Policies Affect Spending and Employment? A New Real-Time Economic Tracker Based on Private Sector Data, Opportunity Insights, https://opportunityinsights.org/wp-content/uploads/2020/06/tracker-summary.pdf (accessed on 6 July 2020).

[74] Circle Economy and EHERO (n.d.), Circular Jobs: Understanding Employment in the Circular Economy in the Netherlands, https://www.circle-economy.com/resources/circular-jobs-understanding-employment-in-the-circular-economy-in-the-netherlands (accessed on 16 March 2020).

[47] Cirillo, V. (2018), “Job polarization in European industries”, International Labour Review, Vol. 157/1, pp. 39-63, http://dx.doi.org/10.1111/ilr.12033.

[114] Crul, M. and J. Mollenkopf (eds.) (2012), The Changing Face of World Cities: Young Adult Children of Immigrants in Europe and the United States, Russel Sage Foundation.

[65] Dauth, W., S. Findeisen and J. Suedekum (2014), “The rise of the east and the far east: German labor markets and trade integration”, Journal of the European Economic Association, Vol. 12/6, pp. 1643-1675, http://dx.doi.org/10.1111/jeea.12092.

[33] Dauth, W. et al. (2018), Adjusting to Robots: Worker-Level Evidence, Federal Reserve Bank of Minneapolis, http://dx.doi.org/10.21034/iwp.13.

[50] Davis, D., E. Mengus and T. Michalski (2020), “Labor Market Polarization and the Great Divergence: Theory and Evidence”, NBER Working Paper No. 26955, National Bureau of Economic Research, Cambridge, MA, http://dx.doi.org/10.3386/w26955.

[59] De Backer, K. et al. (2016), “Reshoring: Myth or Reality?”, OECD Science, Technology and Industry Policy Papers, No. 27, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jm56frbm38s-en.

[113] Diaz Ramirez, M. et al. (2018), “The integration of migrants in OECD regions: A first assessment”, OECD Regional Development Working Papers, No. 2018/01, OECD Publishing, Paris, https://dx.doi.org/10.1787/fb089d9a-en.

[1] Dobbs, R., J. Manyika and J. Woetzel (2015), No ordinary disruption : the four global forces breaking all the trends, Public Affairs Books.

[63] Donoso, V., V. Martín and A. Minondo (2015), “Do Differences in the Exposure to Chinese Imports Lead to Differences in Local Labour Market Outcomes? An Analysis for Spanish Provinces”, Regional Studies, http://dx.doi.org/10.1080/00343404.2013.879982.

[102] Echazarra, A. and T. Radinger (2019), “Does attending a rural school make a difference in how and what you learn?”, PISA in Focus, No. 94, OECD Publishing, Paris, https://dx.doi.org/10.1787/d076ecc3-en.

[2] Edelman (2020), Edelman Trust Barometer 2020 Global Report.

[97] Eurostat (2020), Urban and rural living in the EU, https://ec.europa.eu/eurostat/en/web/products-eurostat-news/-/EDN-20200207-1 (accessed on 15 September 2020).

[80] Eurostat (2017), People in the EU - statistics on geographic mobility, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=People_in_the_EU_-_statistics_on_geographic_mobility#General_overview (accessed on 10 March 2020).

[91] Eurostat (2016), Urban Europe: Statistics on Cities, Towns and Suburbs 2016 edition, Publications office of the European Union, Luxembourg, http://dx.doi.org/10.2785/91120.

[20] EY (2020), Global Capital Confidence Barometer: How do you find clarity in the midst of a crisis?, http://ey.com/CCB (accessed on 15 June 2020).

[66] Fajgelbaum, P. et al. (2019), “The return to protectionism”, NBER Working Paper Series, No. 25638, http://www.nber.org/papers/w25638.

[19] Field, H. and M. Murphy (2020), COVID-19 will herald an automation boom - Protocol, Protocol, https://www.protocol.com/automation-boom-caused-by-coronavirus (accessed on 27 April 2020).

[94] Galera, G. et al. (2018), “Integration of Migrants, Refugees and Asylum Seekers in Remote Areas with Declining Populations”, OECD Local Economic and Employment Development (LEED) Papers, No. 2018/03, OECD Publishing, Paris, https://dx.doi.org/10.1787/84043b2a-en.

[117] Goos, M., J. Konings and M. Vandeweyer (2018), “Local High-Tech Job Multipliers in Europe”, CEPR Discussion Papers.

[48] Goos, M. and A. Manning (2007), “Lousy and Lovely Jobs: The Rising Polarization of Work in Britain”, Review of Economics and Statistics, Vol. 89/1, pp. 118-133, http://dx.doi.org/10.1162/rest.89.1.118.

[45] Goos, M., A. Manning and A. Salomons (2009), “Job Polarization in Europe”, American Economic Review, Vol. 99/2, pp. 58-63, http://dx.doi.org/10.1257/aer.99.2.58.

[26] Graetz, G. and G. Michaels (2017), Is modern technology responsible for jobless recoveries?, American Economic Association, http://dx.doi.org/10.1257/aer.p20171100.

[76] Greiving, S., M. Fleischhauer and C. Lindner (2013), ESPON CLIMATE-Climate Change and Territorial Effects on Regions and Local Economies, ESPON & IRPUD, TU Dortmund, https://www.espon.eu/sites/default/files/attachments/Final%20Report%20Main%20Report.pdf.

[24] Hershbein, B. and L. Kahn (2018), “Do Recessions Accelerate Routine-Biased Technological Change? Evidence from Vacancy Postings”, American Economic Review, Vol. 108/7, pp. 1732-1772, https://doi.org/10.1257/aer.20161570.

[77] Hsiang, S. et al. (2017), “Estimating economic damage from climate change in the United States”, Science, Vol. 356/6345, pp. 1362-1369, http://dx.doi.org/10.1126/science.aal4369.

[107] ICF (2018), Study on the movement of skilled labour, Publications Office of the EU, Luxembourg, https://op.europa.eu/en/publication-detail/-/publication/05079c0e-fc2e-11e8-a96d-01aa75ed71a1/language-en/format-PDF/source-114336449.

[69] IEA (2020), Global Energy Review 2020, IEA, Paris, https://www.iea.org/reports/global-energy-review-2020 (accessed on 18 August 2020).

[23] Jaimovich, N. and H. Siu (2020), “Job polarization and jobless recoveries”, Review of Economics and Statistics, Vol. 102/1, http://dx.doi.org/10.1162/rest_a_00875.

[15] Lamb, C. and V. Vu (2019), “Skills demand in a digital economy”, Brookfield Institute for Innovation + Entrepreneurship, https://brookfieldinstitute.ca/commentary/skills-demand-in-a-digital-economy/ (accessed on 18 September 2020).

[72] Laubinger, F., E. Lanzi and J. Chateau (2020), “Labour market consequences of a transition to a circular economy:  A review paper”, OECD Environment Working Papers, No. 162, OECD Publishing, Paris, https://dx.doi.org/10.1787/e57a300a-en.

[118] Lee, N. and S. Clarke (2019), “Do low-skilled workers gain from high-tech employment growth? High-technology multipliers, employment and wages in Britain”, Research Policy, Vol. 48/9, p. 103803, http://dx.doi.org/10.1016/j.respol.2019.05.012.

[60] Livesey, F. (2018), “Unpacking the possibilities of deglobalisation”, Cambridge Journal of Regions, Economy and Society, Vol. 11/1, pp. 177-187, http://dx.doi.org/10.1093/cjres/rsx030.

[7] Lund, S. et al. (2019), Future of work in America, McKinsey Global Institute, https://www.mckinsey.com/featured-insights/future-of-work/the-future-of-work-in-america-people-and-places-today-and-tomorrow (accessed on 10 March 2020).

[37] Maghazei, O. and T. Netland (2019), “Drones in manufacturing: exploring opportunities for research and practice”, Journal of Manufacturing Technology Management, Vol. ahead-of-print/ahead-of-print, http://dx.doi.org/10.1108/jmtm-03-2019-0099.

[85] Marinescu, I. and R. Rathelot (2018), “Mismatch Unemployment and the Geography of Job Search”, American Economic Journal: Macroeconomics, Vol. 10/3, pp. 42-70, http://dx.doi.org/10.1257/mac.20160312.

[71] Mertins-Kirkwood, H. (2018), Making decarbonization work for workers: Policies for a just transition to a zero-carbon economy in Canada, Canadian Centre for Policy Alternatives, https://www.policyalternatives.ca/publications/reports/making-decarbonization-work-workers (accessed on 17 March 2020).

[40] Moretti, E. (2012), The new geography of jobs, Houghton Mifflin Harcourt, Boston and New York.

[93] Moretti, E. (2011), “Local Labor Markets”, in Handbook of Labor Economics, Elsevier, http://dx.doi.org/10.1016/s0169-7218(11)02412-9.

[121] Moretti, E. (2010), “Local multipliers”, American Economic Review 100(2), pp. 373-377, http://dx.doi.org/10.1257/aer.100.2.373.

[14] Muro, M. et al. (2017), Digitalisation and the American Workforce, The Brookings Institution, https://www.brookings.edu/research/digitalization-and-the-american-workforce/.

[22] Muro, M., R. Maxim and J. Whiton (2020), The robots are ready as the COVID-19 recession spreads, The Brookings Institution, https://www.brookings.edu/blog/the-avenue/2020/03/24/the-robots-are-ready-as-the-covid-19-recession-spreads/ (accessed on 22 April 2020).

[18] Nedelkoska, L. and G. Quintini (2018), “Automation, skills use and training”, OECD Social, Employment and Migration Working Papers, No. 202, OECD Publishing, Paris, https://dx.doi.org/10.1787/2e2f4eea-en.

[99] OECD (2020), Adult education level (indicator), https://dx.doi.org/10.1787/36bce3fe-en (accessed on 14 September 2020).

[57] OECD (2020), “COVID-19 and global value chains: Policy options to build more resilient production networks”, OECD Policy Response to Coronavirus, http://www.oecd.org/coronavirus/policy-responses/covid-19-and-global-value-chains-policy-options-to-build-more-resilient-production-networks-04934ef4/ (accessed on 14 September 2020).

[103] OECD (2020), Employment by education level (indicator), https://dx.doi.org/10.1787/26f676c7-en (accessed on 29 April 2020).

[111] OECD (2020), “Managing international migration under COVID-19”, OECD Policy Responses to Coronavirus (COVID-19), OECD, Paris, http://www.oecd.org/coronavirus/policy-responses/managing-international-migration-under-covid-19-6e914d57/ (accessed on 18 June 2020).

[61] OECD (2020), OECD Economic Outlook, Interim Report September 2020, OECD Publishing, Paris, https://dx.doi.org/10.1787/34ffc900-en.

[56] OECD (2020), OECD Economic Outlook, Volume 2020 Issue 1, OECD Publishing, Paris, https://dx.doi.org/10.1787/0d1d1e2e-en.

[51] OECD (2020), OECD Employment Outlook 2020: Worker Security and the COVID-19 Crisis, OECD Publishing, Paris, https://dx.doi.org/10.1787/1686c758-en.

[101] OECD (2020), OECD Regions and Cities at a Glance 2020, OECD Publishing, Paris, https://dx.doi.org/10.1787/959d5ba0-en.

[28] OECD (2020), Preparing for the Future of Work in Canada, OECD Reviews on Local Job Creation, OECD Publishing, Paris, https://dx.doi.org/10.1787/05c1b185-en.

[109] OECD (2019), International Migration Outlook 2019, OECD Publishing, Paris, https://dx.doi.org/10.1787/c3e35eec-en.

[58] OECD (2019), OECD Economic Outlook, Volume 2019 Issue 2, OECD Publishing, Paris, https://dx.doi.org/10.1787/9b89401b-en.

[92] OECD (2019), OECD Economic Surveys: Japan 2019, OECD Publishing, Paris, https://dx.doi.org/10.1787/fd63f374-en.

[4] OECD (2019), OECD Employment Outlook 2019: The Future of Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9ee00155-en.

[87] OECD (2019), OECD Regional Development Ministerial Issue Notes, https://www.oecd.org/regional/ministerial/documents/Issue-notes.pdf (accessed on 28 April 2020).

[17] OECD (2019), OECD Skills Outlook 2019 : Thriving in a Digital World, OECD Publishing, Paris, https://dx.doi.org/10.1787/df80bc12-en.

[31] OECD (2019), Regions in Industrial Transition: Policies for People and Places, OECD Publishing, Paris, https://dx.doi.org/10.1787/c76ec2a1-en.

[16] OECD (2019), Skills Matter: Additional Results from the Survey of Adult Skills, OECD Skills Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/1f029d8f-en.

[96] OECD (2019), Working Better with Age, Ageing and Employment Policies, OECD Publishing, Paris, https://dx.doi.org/10.1787/c4d4f66a-en.

[98] OECD (2018), “Back to work: Lessons from nine country case studies of policies to assist displaced workers”, in OECD Employment Outlook 2018, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2018-8-en.

[5] OECD (2018), Good Jobs for All in a Changing World of Work: The OECD Jobs Strategy, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264308817-en.

[27] OECD (2018), Job Creation and Local Economic Development 2018: Preparing for the Future of Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264305342-en.

[10] OECD (2018), OECD Employment Outlook 2018, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2018-en.

[100] OECD (2018), OECD Regions and Cities at a Glance 2018, OECD Publishing, Paris, https://dx.doi.org/10.1787/reg_cit_glance-2018-en.

[55] OECD (2018), Productivity and Jobs in a Globalised World: (How) Can All Regions Benefit?, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264293137-en.

[112] OECD (2018), Working Together for Local Integration of Migrants and Refugees, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264085350-en.

[106] OECD (2017), Getting Skills Right: Skills for Jobs Indicators, Getting Skills Right, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264277878-en.

[68] OECD (2017), Investing in Climate, Investing in Growth, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264273528-en.

[42] OECD (2017), OECD Employment Outlook 2017, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2017-en.

[39] OECD (2017), The Next Production Revolution: Implications for Governments and Business, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264271036-en.

[9] OECD (2016), Back to Work: United States: Improving the Re-employment Prospects of Displaced Workers, Back to Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264266513-en.

[54] OECD (2016), OECD Regional Outlook 2016: Productive Regions for Inclusive Societies, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264260245-en.

[89] OECD (2016), OECD Regions at a Glance 2016, OECD Publishing, Paris, https://dx.doi.org/10.1787/reg_glance-2016-en.

[104] OECD (2016), Skills Matter: Further Results from the Survey of Adult Skills, OECD Skills Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264258051-en.

[8] OECD (2009), OECD Employment Outlook 2009: Tackling the Jobs Crisis, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2009-en.

[29] OECD (forthcoming), Preparing for the Future of Work in the Basque Country, Spain, OECD Publishing, Paris.

[108] OECD/AFD (2019), “The new immigrants”, Migration Data Brief, https://www.oecd.org/migration/mig/Migration-data-brief-4-EN.pdf (accessed on 6 March 2020).

[110] OECD/Bertelsmann Stiftung (2019), “How do OECD countries compare in their attractiveness for talented migrants?”, Migration Policy Debates, No. 19, https://www.oecd.org/els/mig/migration-policy-debates-19.pdf (accessed on 6 March 2020).

[70] Popp, D. et al. (2020), “The Employment Impact of Green Fiscal Push: Evidence from the American Recovery Act”, No. 27321, National Bureau of Economic Research Working Paper, Cambridge, MA, http://dx.doi.org/10.3386/w27321.

[11] Quintini, G. and D. Venn (2013), Back to Work: Re-employment, Earnings and Skill Use after Job Displacement, https://www.oecd.org/els/emp/Backtowork-report.pdf (accessed on 26 March 2020).

[75] Reserve Bank of Australia (2020), “Statement on Monetary Policy – February 2020” February, https://www.rba.gov.au/publications/smp/2020/feb/overview.html (accessed on 16 September 2020).

[49] Rodrik, D. (2016), “Premature deindustrialization”, Journal of Economic Growth, Vol. 21/1, http://dx.doi.org/10.1007/s10887-015-9122-3.

[41] Sandler, R. (2020), “Facebook Is The Latest Company To Extend Remote Work Until July 2021—Here’s When Other Companies Plan To Go Back”, Forbes, https://www.forbes.com/sites/rachelsandler/2020/08/06/facebook-is-the-latest-company-to-extend-remote-work-until-june-2021-heres-when-other-companies-plan-to-go-back/#6cb403fb18a3 (accessed on 12 August 2020).

[21] Sillitoe, B. (2020), Coronavirus impact: where will automation in retail be accelerated?, IMRG, https://www.imrg.org/blog/coronavirus-impact-where-will-automation-in-retail-be-accelerated/ (accessed on 16 September 2020).

[12] Silva, F. et al. (2019), “Structural Adjustment, Mass Lay-offs and Employment Reallocation”, OECD Science, Technology and Industry Policy Papers, No. 72, OECD Publishing, Paris, http://doi.org/10.1787/90b572f3-en (accessed on 27 March 2020).

[13] Simonen, J., J. Herala and R. Svento (2020), “Creative destruction and creative resilience: Restructuring of the Nokia dominated high‐tech sector in the Oulu region”, Regional Science Policy & Practice, http://dx.doi.org/10.1111/rsp3.12267.

[6] Smit, S. et al. (2020), The future of work in Europe, McKinsey Global Institute, https://www.mckinsey.com/featured-insights/future-of-work/the-future-of-work-in-europe (accessed on 18 June 2020).

[119] Terzidis, N., R. Maarseveen and R. Argiles (2017), “Employment Polarization in local labor markets: the Dutch case”, CPB Discussion Paper, https://ideas.repec.org/p/cpb/discus/358.html (accessed on 22 July 2019).

[90] Thomas, E., I. Serwicka and P. Swinney (2015), Urban demographics: Where do people live and work in England and Wales?, Centre for Cities, https://www.centreforcities.org/publication/urban-demographics.

[46] Tüzemen, D. and J. Willis (2013), The Vanishing Middle: Job Polarization and Workers’ Response to the Decline in Middle-Skill Jobs, The Kansas City Federal Reserve, https://www.kansascityfed.org/publicat/econrev/pdf/13q1tuzemen-willis.pdf (accessed on 30 March 2020).

[86] United Nations, Department of Economic and Social Affairs, P. (2019), World Population Prospects 2019, https://population.un.org/wpp/ (accessed on 15 March 2020).

[88] US Census Bureau (2019), Metropolitan and Micropolitan Statistical Areas Population Totals and Components of Change: 2010-2019, https://www.census.gov/data/tables/time-series/demo/popest/2010s-total-metro-and-micro-statistical-areas.html (accessed on 6 November 2020).

[83] US Census Bureau (2017), “United States Mover Rate at a New Record Low”, https://www.census.gov/newsroom/blogs/random-samplings/2017/01/mover-rate.html (accessed on 27 March 2020).

[30] van Agtmael, A. and F. Bakker (2016), The smartest places on earth - why rustbelts are the emerging hotspots of global innovation, PublicAffairs, New York.

[73] Vona, F., G. Marin and D. Consoli (2018), “Measures, drivers and effects of green employment: evidence from US local labor markets, 2006–2014”, Journal of Economic Geography, Vol. 19/5, pp. 1021-1048, http://dx.doi.org/10.1093/jeg/lby038.

[115] What Works Centre for Local Economic Growth (n.d.), Toolkit: Multiplier Effects, https://whatworksgrowth.org/public/files/Toolkits/Multipliers_Toolkit.pdf (accessed on 19 March 2020).

Notes

← 1. Population ageing will also impact the demand side of the labour market as consumer demands shift over the lifecycle, for example an increase in demand for health care services.

← 2. OECD estimates are based on the analysis of PIAAC data for 32 OECD countries from 2012 and 2015 (Nedelkoska and Quintini, 2018[18]).

← 3. There are ongoing debates about the scale of the local job multipliers, as well as the differing effects of and impacts on jobs in tradeable and non-tradeable sectors and the local economic and institutional conditions that impact these multipliers (What Works Centre for Local Economic Growth, n.d.[115]; Bartik and Sotherland, 2019[116]). For example, studies in the US have found that for each manufacturing job created in a given city, 1.6 jobs in the non-tradeable sector are created. This figure rises to 2.5 for skilled tradeable jobs (Moretti, 2010[121]), while high-tech jobs have been found to have a multiplier of 5 (Moretti, 2012[40]). Other research has found a local job multiplier of between 3.9 and 4.4 for high-tech jobs in Europe, although with significant regional differences. (Goos, Konings and Vandeweyer, 2018[117]), while others have found significantly lower multipliers for high-tech jobs, such as .7 in the UK (Lee and Clarke, 2019[118]).

← 4. These findings are robust with and without country fixed effects, although stronger without country fixed effects.

← 5. See, for example Davis, Mengus and Michalski (2020[50]) for France, Terzidis, Maarseveen and Argiles, 2017 (2017[119]) for the Netherlands, and Autor (2019[120]) for the United States.

← 6. See, for example, https://www.bloomberg.com/news/articles/2020-04-08/japan-to-fund-firms-to-shift-production-out-of-china

← 7. See OECD (2020), “COVID 19 and the low carbon transition: impacts and possible policy responses” http://www.oecd.org/coronavirus/policy-responses/covid-19-and-the-low-carbon-transition-impacts-and-possible-policy-responses-749738fc/ for a further discussion of these issues.

← 8. These figures are based on the percentage of households that changed residence within the last 5 years, and thus includes both local residential moves and longer-distance moves. Job-related reasons are likely more important for the latter, but comparative international data is not available.

← 9. Challenges include differences in data collection (censuses, registers, surveys), time frames used, and spatial frameworks. See Bell et al. (2015[82]) for a further discussion.

← 10. For example, Hal Vernon, the Chief Economist of Google estimates that the net effect of demographic changes on wages will be 53% greater than that of automation. See https://voxeu.org/article/automation-versus-procreation-aka-bots-versus-tots for more information.

← 11. Estonia, Latvia, and Lithuania also saw their labour force shrink over this time period, although there is only one TL2 region in each of these countries.

← 12. See, for example, Program that pays workers $10,000 to move to Vermont and work remotely is now accepting applications (14 January 2019) and Tulsa wants to pay you $10,000 to move there and work remotely (29 October 2019).

← 13. For the purposes of PISA, rural schools are considered those located in rural areas or villages with fewer than 3 000 inhabitants, while urban schools are located in cities with 100 000 inhabitants or more.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2020

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.