2. A roadmap for delivering well-being in rural regions
This chapter outlines long-term trends in socio-economic development in OECD regions, with a focus on the effects of the 2008 financial crisis and the implications of the current COVID-19 crisis. The first section introduces a new way of classifying regions based on their density and access to cities. The second section discusses key demographic trends across rural regions, focusing on distinctive challenges brought about by population losses and ageing. The third section describes the changing economic conditions facing rural regions after the financial crisis and their effects on regional inequality. The chapter closes with a discussion on the role of skills and human capital, Internet connectivity and innovation as enabling factors of regional development.
Around 30% of the OECD population lived in rural regions in 2019, of which three-quarters lived in regions with close connections to cities. Still, in seven OECD countries, one-fifth of the population or more live in remote, sparsely populated regions.
Annual population growth in metropolitan regions has more than doubled versus rural regions in the past two decades. Among rural regions, remote regions gained more population than regions with or near a small/medium city, where population growth slowed down after the 2008 financial crisis.
Demographic pressures are stronger in regions far from cities. Half of OECD countries with remote regions and about one-third of countries with regions with or near a small/medium city dealt with population decline in the last two decades.
In 2019, the elderly as a percentage of the working population was above 30% across rural regions and was highest in rural regions near to large cities at 33%. Seventy-three regions had elderly dependency ratios above 50%, and were above 60% in 11 regions. The ageing dependency ratio has increased fastest in remote regions in the past two decades.
The gap between top and bottom regions and between metropolitan regions and regions far from large cities widened in the aftermath of the global financial crisis in most OECD countries.
Regional inequality, measured as the difference between the top 20% and bottom 20% of regions in gross domestic product (GDP) per capita, increased in 24 out of 28 OECD countries with available data in the post-crisis period and as compared to the pre-crisis period. In countries where regional inequality increased, changes were driven by improvements in top regions, except in the case of Greece and Italy.
The income level, productivity level and employment rate of regions near a large city are around 18, 10 and 8 percentage points below OECD average levels respectively. The gaps for regions with or near a small/medium city are even bigger, at 28, 20 and 14 percentage points below the OECD average, while the gap for remote regions is 21, 14 and 3 percentage points below the OECD average.
Larger and denser places were more resilient to the shock of the 2008 financial crisis. After the financial crisis, regions far from large cities grew slower than all other region types. Meanwhile, regions near a city grew faster than other region types and employment creation was concentrated largely in metropolitan regions.
The importance of the service sector increased in metropolitan and rural regions. Some 71% of jobs were in the service sector in remote regions while the percentage in metropolitan regions was 75% in 2017. While top-performing rural regions in terms of GDP per capita levels are specialised in high-value-added services, bottom-performing rural regions have not diversified away from traditional primary sectors and low value-added services.
Rural regions struggled to create new jobs after the global financial crisis. In the post-crisis period, rural regions contributed less than 10% to employment growth, down from a contribution of over 20% in the pre-crisis period.
Innovation performance, based solely on patent activity, is lower in rural regions compared to metropolitan regions. At the same time, evidence for Europe suggests individuals living in rural regions strongly lag behind their peers in cities with regard to their level of digital skills, paramount for many modern workplaces. Addressing the rural-urban digital divide in connectivity, education and skills is crucial to boost innovation in rural regions.
In recent decades, OECD countries and regions have faced a number of structural transformations affecting their development trajectories. Globalisation, digitalisation and the shocks of the global financial crisis and current COVID-19 crisis are deeply shaping the economic landscape of rural economies. Today more than ever, the distance between “winners” and “losers” feels ever-widening. The 2008 global financial crisis exacerbated the divergence between regions endowed with the key ingredients for high-income generation and those lacking them (Iammarino, Rodriguez-Pose and Storper, 2018[1]). Persistent inequality has divided societies, leading large swaths of the population to feel they belong to “places that do not matter” (Rodríguez-Pose, 2018[2]). This discontent has recently fuelled populist and anti-establishment sentiments in some OECD countries, underscoring the failure of traditional “place-insensitive” solutions to ensure prosperity and convergence (McCann, 2019[3]). Inequalities within countries will likely widen with the current COVID-19 crisis as both virus incidence and socio-economic consequences are highly asymmetrical across places. The financial and economic consequences of the ongoing COVI-19 pandemic threaten to become a catalyst for further discontent.
It is in this context that many rural communities will face further ageing, outmigration, service provision challenges and a shift in population compositions following international migration in the next decades. Rural economies will also continue to face sweeping megatrends including global shifts in production, new technological breakthroughs and environmental pressures from climate change. These megatrends offer new opportunities to rural economies, including the transition to renewable energy, benefitting from tourism and ecosystem services, and adopting artificial intelligence technologies to improve well-being. The same trends may on the other hand generate uncertainty stemming from job losses from increased digitalisation, environmental disasters, higher fiscal pressure tied to declining tax revenues, and uncertainty about adequate public service provision.
Making the most of these changes requires a forward-looking view of a sustainable, inclusive and balanced development path. Policies need to shift from space-blind to place- and people-based, from the passive use of transfers and subsidies to active efforts to make the best use of resources. By focusing on rural places, this chapter sheds light on the distinctive shifts taking place outside urban areas, the needs of different types of rural regions and the importance of rural-urban linkages.
The next section discusses the definition of rural in the context of regions and introduces an alternative typology of regions based on their density and level of access to cities. The second section discusses demographic trends amid increasing urban concentration, focusing on distinctive challenges for rural regions. The third section analyses economic trends across OECD regions, with an emphasis on the effect of the economic crisis in 2008. The final section discusses the role of skills and human capital, Internet connectivity and innovation as enabling factors of regional development.
The term “rurality” is generally recognised as a multidimensional concept, embodying different meanings for different purposes. Debates on the definition of rural spaces focus on how best to define the concept –e.g. as a geographical/spatial concept, a land use concept, a socio-economic or socio-cultural descriptor, a functional concept related to, for instance, labour market flows, or simply as “not urban”.
In March 2020, the United Nations (UN) Statistical Commission endorsed a new global definition of cities, urban and rural areas called the Degree of Urbanisation (UN Statistical Commission, 2020[4]). The Degree of Urbanisation is the first global definition of rural areas to be endorsed by the UN and it goes beyond the traditional rural-urban dichotomy by proposing concrete measures of places in the rural-urban continuum. What is more, the Degree of Urbanisation also provides a refined definition of moderate and low-density areas that include towns, villages, dispersed area and mostly uninhabited areas (Box 2.1). This new definition of the space outside cities opens new possibilities for measuring diversity within rural and emphasise the role of rural-urban linkages in future work. As a definition built for international comparability, the Degree of Urbanisation is not designed to replace national definitions. National definitions can incorporate more indicators, can be tailored to reflect specific circumstances and better serve the needs of national policies.
Six international organisations – the European Commission, the UN Food and Agriculture Organization (FAO), the UN Human Settlements Programme (UN-Habitat), the International Labour Organization (ILO), the OECD and the World Bank – have worked closely together to develop a harmonised methodology to facilitate international statistical comparisons, called the Degree of Urbanisation.
The Degree of Urbanisation was designed to create a simple and neutral method that could be applied in every country in the world. It relies primarily on population size and density thresholds applied to a population grid with cells of 1 km by 1 km. The different types of grid cells are subsequently used to classify small spatial units, such as municipalities or census enumeration areas. The Degree of Urbanisation was endorsed by the UN Statistical Commission in March 2020.1 The Degree of Urbanisation level 1 classifies the entire territory into: i) cities; ii) towns and suburbs; and iii) rural areas. At level 2, towns and suburbs are split into: i) dense towns; ii) semi-dense towns; and iii) suburbs. Rural areas are split into: i) villages; ii) dispersed rural areas; and iii) mostly uninhabited areas.
Cities have a population of at least 50 000 inhabitants in contiguous grid cells with a density of at least 1 500 inhabitants per km2.
Dense towns have a population of between 5 000 and 50 000 inhabitants in contiguous grid cells with a density of at least 1 500 inhabitants per km2.
Semi-dense towns have a population of at least 5 000 inhabitants in contiguous cells with a density of at least 300 inhabitants per km2 and are at least 2 km away from the edge of a city or dense town.
Suburbs have most of their population in contiguous cells with a density of at least 300 inhabitants per km2 that are part of a cluster with at least 5 000 inhabitants but are not part of a town.
Villages have between 500 and 5 000 inhabitants in contiguous cells with a density of at least 300 inhabitants per km2.
Dispersed rural areas have most of their population in grid cells with a density between 50 and 300 inhabitants per km2.
Mostly uninhabited areas have most of their population in grid cells with a density of fewer than 50 inhabitants per km2.
Source: UN Statistical Commission (2020[4]), “A recommendation on the method to delineate cities, urban and rural areas for international statistical comparisons”, https://unstats.un.org/unsd/statcom/51st-session/documents/BG-Item3j-Recommendation-E.pdf.
Traditional definitions of “rural” have relied on a number of noticeable characteristics including: i) classifying the entire territory as either urban or rural; ii) focusing the definition primarily on urban characteristics by defining rural as the residual of urban; iii) not differentiating among different types of rural areas; and iv) not recognising mixed areas with strong urban and rural interactions (OECD, 2014[5]).
An evolution of these definitions identified density as the differentiating factor conceptualising rural regions. In 2006, the New Rural Paradigm (OECD, 2006[6]) separated the concept of rural from the concept of urban by using actual rural characteristics. The New Rural Paradigm recognised the diversity of rural regions in terms of access to markets, economic competitiveness and structure that sets them apart from each other and urban regions (OECD, 2016[7]), and introduced the narrative of rural as places of opportunities.
Since context and geography matter, it is no surprise that OECD member countries have adopted a wide range of definitions delimiting and adapting urban and rural borders to their geographic characteristics (see Box 2.2). The wide diversity of rural definitions (see Table 3.A.3.1 in (OECD, 2016[7])) also reflects different criteria that exist to elaborate definitions including density, economic activity, size or distance to services, among others. Beyond the definition of rural, it should be emphasised that a strong rural development policy requires actions not only by local but also by regional and national levels of government.
Rural definitions adapted to local realities
OECD countries have moved away from traditional definitions of rural as simply the remaining “leftover” space that is not urban, to delimitations of rural that can help identify common challenges and opportunities to design better policy responses.
The design of new and territorial definitions is possible today more than ever thanks to the advancement of Geographic Information System (GIS) tools and better availability of data. The revisions of definitions have incorporated new criteria such as distance, commuting and accessibility to services, and are now recognising areas with strong urban and rural interactions. For instance:
France is advancing on a definition considering accessibility. The National Institute of Statistics and Economic Studies has developed an indicator that examines the accessibility of services and amenities that are important to daily life for communities of varying population densities (Barbier, Toutin and Levy, 2016[8]).
New Zealand has adopted a definition that distinguishes between rural areas with high, moderate or low urban influence and those deemed rural-remote, by drawing on both population density, place of employment and commuting data (Statistics New Zealand, 2004[9]).
Italy has developed a definition based on service accessibility and policy objectives. Rural areas in Italy are split into three categories: intensively cultivated and plain areas; intermediate rural areas; and areas with lagging development. This classification is based on population density indicators and share of agricultural land. Italy has also adopted a classification of rural areas based on policy objectives. It characterises “inner areas” as groups of municipalities facing “inadequate access to essential services”, including healthcare, education and transportation.
The Australian Bureau of Statistics (ABS) developed a typology of regions and cities that aims to capture the particularities of Australian geography, such as a small density of settlements and large portions of empty territory. It defines geographic areas including metropolitan, inner regional, outer regional, remote and very remote areas, and urban areas and localities of various sizes. The ABS geography also includes an accessibility and remoteness index.
Source: Barbier, M., G. Toutin and D. Levy (2016[8]), “L’accès aux services, une question de densité des territoires”, https://www.insee.fr/fr/statistiques/1908098; Statistics New Zealand (2004[9]), New Zealand: An Urban/Rural Profile, http://archive.stats.govt.nz/browse_for_stats/Maps_and_geography/Geographic-areas/urban-rural-profile.aspx#gsc.tab=0; European Network for Urban Development (n.d.[10]), Strategy for Inner Areas - Italy, https://enrd.ec.europa.eu/sites/enrd/files/tg_smart-villages_case-study_it.pdf; Australian Bureau of Statistics (n.d.[11]), The Australian Statistical Geography Standard (Asgs) Remoteness Structure, https://www.abs.gov.au/websitedbs/D3310114.nsf/home/remoteness+structure; Australian Bureau of Statistics (2018[12]), 1270.0.55.005 - Australian Statistical Geography Standard (ASGS): Volume 5 - Remoteness Structure, July 2016, https://www.abs.gov.au/ausstats/[email protected]/mf/1270.0.55.005.
Key dimensions defining rural
The geography of a place is effectively defined by a combination of physical (“first-nature”) and human (“second-nature”) geographies. The more people inhabit a place, the more its character will be defined by second-nature geography – by human beings and their activities. In contrast, less human presence implies a larger role for natural factors in shaping economic opportunities.
Economic remoteness or “peripherality” has three distinct features:
The first is the physical distance to major markets. Distance increases travel times and shipping costs, which must be borne by the buyer (in the form of higher prices) or seller (in the form of lower margins).
The second is the degree of economic connectedness. Lack of economic integration not only reduces current trade opportunities but also the ability of agents in a place to identify new opportunities. Thus, there are both static and dynamic associated costs.
The third is the degree of sector specialisation. Production is concentrated in relatively few sectors since it is impossible to achieve “critical mass” in more than a few activities. A narrower economic base implies greater vulnerability to sector-specific shocks, whether positive or negative.
In this context, rural places have “low-density economies”, specialised in niche markets or those linked to natural resources (e.g. agriculture, tourism, etc.). Geographical features and settlement patterns set rural areas apart from urban areas, as they differ in terms of local workforce size, sensitivity to transport costs, level of competition with similar regions, and reliance on innovations developed elsewhere. Because of their size and reliance on external markets, rural economies may be more vulnerable to external changes affecting economic and natural conditions. At the same time, many rural places have rich social capital resulting from community cohesion and strong informal and formal social networks capable of promoting social trust.
The Geography of Opportunities paradigm extends this diversity and acknowledges the existence of a rural-urban continuum so that it is not the presence of characteristics but rather the degree of a factor –rurality for example – which differentiates places. In addition to density as a central concept to rural economies, accessibility has taken a central role as a defining characteristic of places. While common perception suggests that all rural places typically face larger physical connectivity barriers to markets and services than cities, the level of access depends on the location of rural places relative to urban nodes. This will also determine the degree of interdependencies between rural and urban areas through different types of linkages that often cross traditional administrative boundaries.
Urban and rural places are highly interconnected across economic, social and environmental dimensions (Figure 2.1), and these linkages tend to be stronger in rural places that are closer to cities. Linkages can occur in many dimensions including amongst other commercial ties, environmental goods and population flows.
Rural places that are in close proximity to cities have much stronger linkages in transportation networks, commuting flows, spatial planning and the provision of goods and services. Furthermore, these rural places can also benefit from good access to markets, services and agglomeration of talent present in urban areas. These benefits are often referred to as “borrowed” agglomeration effects. In turn, rural places close to cities also enjoy environmental amenities and lower housing costs than cities making them attractive and liveable places.
Linkages are not limited to city-centred local labour market flows and include bi-directional relationships. Each type of interaction encompasses a different geography or “functional region”. Flexibility is required in the space considered for governing these complex relationships. Remote areas in contrast face the largest connectivity barriers due to their geographical location far away from transportation nodes. This distinction matters because lack of connectivity entails higher transportation, infrastructure and service provision costs that affect the well-being of rural residents.
Source: OECD (2013[13]), Rural-Urban Partnerships: An Integrated Approach to Economic Development, https://dx.doi.org/10.1787/9789264204812-en.
This complexity can be represented by an urban to rural continuum. While there are no sudden breaks in these spatial relationships, there is great diversity in the size and types of interconnections. Figure 2.2 depicts a continuity of urban and rural places based on location, proximity and density characteristics – moving from more to less concentrated settlements, with multiple connections and interactions among them. Such distinctions are important for public policy, with implications for jobs, services and infrastructure development, among other considerations. It also implies that the barriers between urban and rural are not dichotomous and clear-cut because territories display different degrees of interaction between urban and rural.
Rural in this continuum plays an important complementary role to urban. This means that the development path of most rural places is not to become themselves cities but instead to provide goods and services that are best produced in a rural setting and then delivered to national and international markets.
From the perspective of regions, the main difference is where the driving source of economic dynamism is located. In highly urbanised regions, it is clearly in the city, whereas in remote regions, it is in rural areas. The vast majority of rural territory falls into an “intermediate” situation where the urban and rural components of a region are more balanced in capacity and there are potentially substantial gains from co-ordination.
Source: OECD (2016[7]), OECD Regional Outlook 2016: Productive Regions for Inclusive Societies, https://dx.doi.org/10.1787/9789264260245-en.
Within this complex configuration, there are three types of rural places, each broadly defined with different characteristics and policy needs:
Rural within a functional urban area (FUA) – These types of rural places are part of the catchment area of the urban core and their development is fully integrated into the metropolitan strategy. The main challenges of these types of rural places are accessibility of services within the FUA, matching of skills to the wide range of supply and managing land use policy brought by increasing pressures of the urban core.
Rural close to cities – The main challenges in these types of places are: improving two-way connectivity and accessibility between the cities and rural territory; building short supply chains that link urban and rural firms; balancing population growth while preserving quality of life and green spaces; and enhancing the provision of secondary goods and services.
Remote rural – Remote places depend largely on the primary activities of the area. Growth relies on absolute and comparative advantage, improving connectivity to export markets, matching skills to areas of comparative advantage and ensuring the provision of essential services (e.g. tourism). In more densely settled but remote regions where farms are distributed across the open countryside, some small cities and towns serve the farm population as market points.
Classifying regions according to their level of access to cities of different sizes
To bring these conceptual ideas into measurement, the OECD typology of regions based on their level of access to cities builds on previous territorial definitions (see Box 2.3) to introduce the idea of spatial continuity between urban and rural. TL3 regions cover the entire territory within countries, while FUAs only capture a sub-sample of the territory.
The OECD typology based on the level of access to cities aims at taking into account the relative location of rural places with respect to FUAs. This typology is meant to be relevant for rural policies while ensuring international comparability. As such, it differentiates amongst different types of rural regions – those close to cities and those that are remote. Rural places close to cities require a much stronger integration of policies with cities in areas such as transportation, land use labour market or housing amongst others. Furthermore, the definition differentiates rural with access to large cities vis-à-vis small/medium places allowing for a better understanding and capturing differences in linkages.
The typology used in this document identifies five types of (TL3) regions based on the share of population living in metropolitan areas and an accessibility criterion. The 5 types of regions include 2 types of metropolitan regions – large metropolitan (with an FUA of more than 1 million people) and metropolitan regions (with an FUA of less than 250 000 people) –, and 3 types of rural regions – regions near a large city (i.e. regions with access to an FUA of more than 250 000 people within a 60-minute drive), regions with a small/medium city or near one (i.e. regions with an FUA of less than 250 000 people or with access to one within a 60-minute drive), and remote regions.
Throughout this report, reference will be made to “rural regions” when referring to the group of non-metropolitan regions. The term “rural regions” is not a synonym for “predominantly rural regions” as defined in the OECD regional typology developed in 2011 (see Box 2.3). The terms “city” and FUA will be used interchangeably. The document uses the term “large city” to signify a city (FUA) with more than 250 000 inhabitants and “very large” city when referring to a city with more than 1 million inhabitants. The term “areas outside FUAs” is meant to be comprehensive of territories with settlements with intermediate or low-density levels, such as towns and suburbs as defined by the Degree of Urbanisation. On the other hand, terms such as “rural economy”, “rural places” and “rural communities” are used conceptually for policy purposes and are not meant to reflect any particular territorial definition.
Three decades of territorial definitions for international comparisons
Country rural definitions are adapted to their specific needs and are mainly used for policy implementation. The OECD has developed regional typologies to allow for international comparisons and to be able to compare regions with similar characteristics.
The first OECD regional typology was elaborated in 1991 in collaboration with the European Commission (EC). It classified Territorial Level 3 (TL3) regions as predominantly urban, intermediate or predominantly rural using simple and commonly accepted criteria based on a three-step procedure:
1. Identify rural communities according to population density. A community is defined as rural if its population density is below 150 inhabitants per km² (500 inhabitants per km² for Japan).
2. Classify regions according to the percentage of the population living in rural communities. A TL3 region is classified as predominantly rural if more than 50% of its population lives in rural communities.
3. Adjust the classification of “predominantly rural” and “intermediate” regions based on the size of the urban centres.2
This typology was further developed in 2011 introducing an accessibility criterion sub-classifying rural regions into two sub-groups: rural close to cities and rural-remote regions based on a driving distance to urban centre criterion.3 The result is a fourfold classification of TL3 regions: predominantly urban (PU), intermediate regions (IN), predominantly rural regions close to a city (PRC) and predominantly rural-remote regions (PRR) (Brezzi, Dijkstra and Ruiz, 2011[14]). This extension already included two of the types of rural areas outlined above.
In 2014, the European Union (EU) modified the rural-urban typology, using one-km2 population grids as building blocks to identify rural or urban communities with the aim of improving international comparability by changing the previously arbitrarily defined local units (Dijkstra and Poelman, 2014[15]). The OECD updated this classification only for EU-OECD countries.
In parallel to the regional typology, in 2011, the OECD elaborated an FUA definition of cities and their broader area of influence based on commuting patterns. An FUA is constructed by concatenating grid cells with high population density (above 1 500 inhabitants per km2) into an urban core (Fadic et al., 2019[16]). These cells are then connected with surrounding lower-density cells when the flows of commuting between the 2 types of cells exceed a given threshold (i.e. at least 15% of the labour force commutes to the urban core). The set of the urban core and the hinterland compose the FUA.4
In 2018, the OECD developed an alternative definition of regions based on their level of access to cities of different sizes (see Box 2.4).
Source: Fadic, M. et al. (2019[16]), “Classifying small (TL3) regions based on metropolitan population, low density and remoteness”, https://dx.doi.org/10.1787/b902cc00-en; Brezzi, M., L. Dijkstra and V. Ruiz (2011[14]), “OECD Extended Regional Typology: The Economic Performance of Remote Rural Regions”, https://dx.doi.org/10.1787/5kg6z83tw7f4-en; Dijkstra, L. and H. Poelman (2014[15]), “A harmonised definition of cities and rural areas: The new degree of urbanisation”, European Commission Regional Working Papers, European Commission.
The alternative regional typology helps uncover the many existing shades of rural: while large metropolitan regions are clearly more “urban” and remote regions clearly more “rural”, other region types differ in their degree of rurality (i.e. the share of the regional population outside FUAs) (Figure 2.3). It also highlights the role of access in setting apart regions with a high degree of rurality with and without access to cities.
Note: City refers to an FUA.
Source: OECD (2020[17])OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en, EC (n.d.[18]), Global Human Settlement Layer 2015, https://ghsl.jrc.ec.europa.eu/
The regional classification based on access allows measuring socio-economic differences between regions, across and within countries. It takes into consideration the presence of and access to FUAs. Access is defined in terms of the time needed to reach the most proximate urban area, a measure that takes into account not only geographical features but also the status of physical road infrastructure.
The first tier adopts as a threshold of 50% of the population of the TL3 (small) region living in an FUA of at least 250 000 people; the second tier uses a 60-minute driving-time threshold, a measure of the access to an FUA.
The new methodology classifies TL3 regions into metropolitan and non-metropolitan according to the following criteria:
Metropolitan TL3 region, if more than 50% of its population live in an FUA of at least 250 000 inhabitants. Metropolitan regions are further classified into:
Large metropolitan TL3 regions, if more than 50% of its population lives in an FUA of at least 1.5 million inhabitants.
Metropolitan TL3 regions, if the TL3 region is not a large metropolitan region and 50% of its population lives in an FUA of at least 250 000 inhabitants.
Non-metropolitan TL3 region, if less than 50% of its population live in an FUA. These regions are further classified according to their level of access to FUAs of different sizes into regions:
With access to (near) a metropolitan TL3 region, if more than 50% of its population lives within a 60-minute drive from a metropolitan area (an FUA with more than 250 000 people); or if the TL3 region contains more than 80% of the area of an FUA of at least 250 000 inhabitants.
With access to (near) a small/medium city TL3 region, if the TL3 region does not have access to a metropolitan area and 50% of its population has access to a small or medium city (an FUA of more than 50 000 and less than 250 000 inhabitants) within a 60-minute drive; or if the TL3 region contains more than 80% of the area of a small or medium city.
Remote TL3 region, if the TL3 region is not classified as NMR-M or NMR-S, i.e. if 50% of its population does not have access to any FUA within a 60-minute drive.
Source: Fadic, M. et al. (2019[16]), (2019), “Classifying small (TL3) regions based on metropolitan population, low density and remoteness”, https://dx.doi.org/10.1787/b902cc00-en.
Rural places have common features: low density, peripherality and remoteness. In other words, they all lack economies of agglomeration that attract firms and workers to a given location. Firms tend to locate close to other firms and densely populated areas due to lower transportation costs, proximity to markets and wider availability of labour supply. People are also attracted to densely populated areas for the wider availability of job opportunities, goods and services. These mutually reinforcing forces yield economic premia for both consumers and firms through economies of scale, better matching and functioning of labour markets, spill-over effects and more technological intensity (Duranton et al., 2004[19]). To no surprise, productivity and wages tend to be on average higher in densely populated areas. The benefits, however, must be weighed against the costs of agglomeration – often referred to as diseconomies of scale – including congestion, higher land and housing prices, rising inequality and environmental pressures.
Yet, even without economies of agglomeration from high-density, rural economies can also benefit from agglomeration effects indirectly or at lower scales. Pockets of density outside large cities including villages, market towns and smaller cities can represent important development hubs for the broader rural economy. Rural places located near urban areas can also borrow agglomeration benefits and, at the same time, enjoy lower diseconomies of scale.
In other rural places, however, demographic decline might constitute an unavoidable long-term trend driven by structural factors. In these cases, rural policies should not fight against demographic patterns but rather respond with strategic, sustainable forward-looking policies to manage population decline.
Population distribution across regions
According to the OECD regional typology, 25% of OECD population lived in predominantly rural regions in 2017, 20% of which lived in rural regions close to cities and 5% in rural-remote regions. This means that 80% of the OECD rural population live in close proximity to cities and only 20% in remote regions. This definition, however, does not rely on functionality and classifies many rural places as intermediate regions. According to the alternative regional typology, in 2019, 42% of the OECD population lived in regions with a large city. Amongst the reminding 58%, approximately three-quarters lived in regions near cities, while one-third lived in remote regions (accounting for 8% of the total population). This evidence confirms that the bulk of residents of regions have a strong interaction with cities, or differently said, only a small share of the total population lives in remote areas with no interaction to nearby cities.
The distribution of the population of regions according to the alternative typology (Figure 2.4) captures some similarities of countries according to their geographic characteristics:
Although only 8% of the OECD population live in remote regions, in 7 OECD countries one-fifth or more of the national population live in remote regions. These include Norway (31%), Finland (28%) and Sweden (24%) from Scandinavia with sparsely populated regions, Greece (31%) with an island and mountainous geography, and 2 of the largest OECD countries in terms of area, Canada (23%) and Australia (20%).
In 15 OECD countries, more than one-fifth of the national population live in regions with or near a small/medium city. Countries with the highest shares of population in these types of regions include Iceland (84%), and former East European and Baltic countries including the Slovak Republic (63%), Latvia (57%), the Czech Republic (43%), Hungary (36%), Estonia (34%) and Lithuania (33%).
Regions near a large city are home to one-fifth of the national population or more in 10 OECD countries. These include small- and medium-sized European countries, namely Austria (21%), Belgium (50%), Denmark (30%), Germany (23%), Italy (22%), the Netherlands (25%), Portugal (20%), Slovenia (40%), Switzerland (40%) and the United Kingdom (22%).
Note: City refers to an FUA. 2018 values for Australia, Ireland, Japan and the United States.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Demographic dynamics in rural regions
A common characteristic of cities is their ability to attract people and firms to their location in a sustained form. This occurs since firms like to locate where other firms and/or suppliers are located given the lower transportation costs. They also like to locate where consumers and densities are higher, especially service-oriented firms. Workers in turn also like to locate close to firms, given the higher job opportunities available. Studies of this phenomenon include Perroux’s notion of “growth poles” (1995[20]). Myrdal’s analysis of “circular and cumulative causation” (Myrdal and Sitohang, 1957[21]) and Hirshman’s concept of “forward and backward linkages” (1958[22]).
Demographic patterns across OECD countries over the past two decades confirm these circular and cumulative causation dynamics. The share of population living in metropolitan regions against the share of rural regions increased in all but three OECD countries (Greece, Korea and the Netherlands).
Note: Metropolitan regions include regions with a city of at least 250 000 inhabitants. Based on available data for 2 147 TL3 regions.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Greece was the only OECD country that experienced absolute population losses in the aftermath of the financial crisis, with most of outmigration flows originating in metropolitan regions. Most countries concentrated even more population in metropolitan regions in the aftermath of the crisis, especially small countries such as Estonia and Lithuania and those with large sparsely populated areas such as Canada, Finland and Norway. As most of the largest increases happened in relatively small countries, the increase in the share of metropolitan regions is close to half a percentage point across 31 OECD countries.
Between 2001 and 2019, the population in metropolitan regions grew annually twice as fast (0.70%) as in rural regions (0.33%), driven by growth in large metropolitan regions (0.79%). Outside metropolitan regions, remote regions experienced the fastest growth rate (0.45%) and the second-largest absolute increase (7 million people) after regions near a large city (8 million) (Table 2.2).
Population growth slowed down after the crisis across all rural region types, except in remote regions where population growth slightly accelerated. After the crisis, population growth slowed down by 0.13 percentage points (p.p.) in regions near a large city. In regions with a small/medium city or near one, the slow-down was even sharper at 0.14 p.p.
Although the population in rural regions has grown at a slower pace than in metropolitan regions, around two-thirds of rural regions in each of the three types is gaining population. Still, population decline hit some remote regions the hardest in 2001-19: 36% of all OECD remote regions experienced population decline, with the population falling at a rate of 1% or more in 26 regions in Canada, Chile, Estonia, Germany, Latvia, Lithuania and Portugal.
Over 2001-19, metropolitan regions displayed the highest population growth rates and remote regions the slowest rates in the majority of OECD countries. Among 24 countries with at least 1 large metropolitan region, large metropolitan regions grew faster than other region types in 19 countries – in the remaining 5 countries, metropolitan regions grew faster. In contrast, in 19 out of 28 OECD countries with remote regions, population growth was lowest in that type of region. Half of OECD countries with remote regions (14 out of 28) and 9 out of 31 countries with regions with or near a small/medium city dealt with population decline in those types of regions in 2001-19 (Figure 2.6). Meanwhile, only 5 OECD countries (Japan, Hungary, Germany, Poland and Portugal) dealt with population decline in regions near a large city.
Available population projections for Europe show that, as a whole, regions with or near a small/medium city will have absolute population loses within a decade as early as 2040 and will continue to do so afterwards. The same will happen in metropolitan regions and regions near a large city by 2060. By 2060, regions with or near a small/medium city in Europe will have lost nearly 700 000 people compared to 2015, while metropolitan regions and regions near a large city will have gained nearly 22 million.
Note: City refers to an FUA. Based on available data for 2 147 TL3 regions.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Note: Population series downscaled from country level to NUTS3 level, consistent with the 2018 EC Ageing Report, https://ec.europa.eu/info/sites/info/files/economy-finance/ip065_en.pdf.
Source: Batista e Silva, F. et al. (2016[23]), Regionalisation of Demographic and Economic Projections, https://op.europa.eu/fr/publication-detail/-/publication/f6155238-3f4e-11e6-af30-01aa75ed71a1.
Population growth is mainly driven by three factors: migration, fertility and mortality. Metropolitan areas appear to be drivers of migration. In both metropolitan regions and regions near a large city, net migration was positive in 2015, whereas they are negative in regions with or near a small/medium city and in remote regions. This suggests that larger cities and their surrounding areas are important hubs attracting migrants, whereas smaller cities do not have the same level of attractiveness.
The comparison of net migration rates of total population versus young people reveals that: i) large metropolitan regions attract young people; ii) migration into regions near a large city corresponds to an older profile, as net migration flows for the 15-29 age bracket in this type of region are actually negative; and iii) compared to other age groups, young people disproportionally leave remote regions and regions with or near a small/medium city.
Regarding fertility, the relationship between the proportion of children to women and migration is not simple. Not all age groups and genders migrate at the same rate as different places bring different demands and changes in lifestyles that might affect fertility decisions. Still, child-woman ratios are higher in metropolitan regions across 15 out of 22 countries with available data and lower in rural regions in 8 countries (Figure 2.8).
These findings on higher fertility rates in remote places than in larger cities are consistent with previous studies in the literature (Kulu, 2013[24]). The studies identify compositional effects and contextual ones as the main drivers of the variation:
The compositional effects are due to the higher proportion of highly educated people in cities than in remote areas, and higher fertility tends to be lowest for university education and highest for individuals with only compulsory education (Andersson et al., 2009[25]; Hoem, 2005[26]). The variation may also result from the larger share of students in metropolitan regions and their surrounding areas than in remote regions (Hank, 2001[27]).
The over-representation of married people in small towns and rural areas may explain the higher fertility rates there, in particular the higher likelihood of family formation (Hank, 2001[27]). Couples who intend to have a child (or another child) may move from cities to small towns and villages because the latter are perceived as better suited to raising children and as offering more affordable and spacious child-friendly housing (Kulu, 2013[24]).
Note: Based on available data for 1 407 TL3 regions.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
In turn, mortality rates are expected to be lower in better-performing regions that attract population because of the effect of higher incomes and health infrastructures. The analysis across region types does not reveal a one-to-one correspondence between death rates and density (Figure 2.9) and regions near a large city have the highest maximum average age over the last decade (82 years in 2015). In fact, remote regions have similar death rates than metropolitan regions and both regions have displayed similar trends over time.
Note: The death rate is the number of deaths per 1 000 inhabitants. Based on available information for 1 729 TL3 regions in 29 countries.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Nevertheless, people in remote regions experience the lowest life expectancy on average by living two years less while in regions with or near a small/medium city live one year less. From a national perspective, only Swiss rural regions have a lower death rate than the metropolitan regions. Countries with the largest regional differences include Canada, Denmark, Estonia, Japan, Korea, Portugal and Sweden.
Rural regions face structural challenges of an ageing population
OECD countries are facing structural challenges of an ageing population. Current elderly dependency ratios – the share of the population aged 65 and over as a percentage of the population aged 20-64 – stands at 28.6%. This share is expected to increase to 35% by 2025 and to 53% by 2050 on average in OECD countries (Figure 2.10). Greece, Italy, Japan, Korea, Portugal and Spain are all expected to have elderly dependency ratios of over 70% by 2050.
These national figures, however, mask important regional variations within countries. The rates of change and impacts vary greatly from place to place, resulting in significant changes to both labour markets and the settlement pattern across types of regions. Ageing is a stronger structural phenomenon in rural regions vis-à-vis metropolitan regions. In only one OECD country (Poland), ageing dependency ratios are significantly lower in rural regions compared to metropolitan regions. In the large majority of countries (27 out of 31 countries with available data), the elderly dependency ratio is higher in rural regions by at least 1 percentage point. The countries with the largest gap in elderly dependency ratios in 2019 include Japan, Finland, Australia, United Kingdom, Sweden, Canada and Korea, all with a gap above 9 percentage points.
Source: OECD (2017[28]), Pensions at a Glance 2017: OECD and G20 Indicators, https://doi.org/10.1787/pension_glance-2017-en.
Note: Elderly dependency ratio defined as the number of individuals aged 65 and over per 100 people of working age defined as those aged between 20 and 64.
In 2019, regions near a large city had the highest average elderly dependency ratios (33%), followed by remote regions (31%) and regions with or near a small/medium city (31%). Remote regions experienced, on average, the largest increases between 2003 and 2019 (a 0.9 percentage point increase). In 2019, 73 regions had elderly dependency ratios above 50% and, in 11 regions (including Evrytania from Greece and Akita, Kochi, Shimane and Yamaguchi from Japan), they were above 60%.
Note: Elderly dependency ratio defined as the average share of +65 population with respect to the working working-age population (15-64 years old). Based on available data for 2 147 TL3 regions in 33 countries.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Rural regions will need to prepare to face the growing pressures of ageing. While elderly dependency ratios are highest in regions near a large city, many countries face growing pressures of ageing in regions far from large cities. In about two-thirds of OECD countries with remote regions (23 out of 32), elderly dependency ratios were the highest in remote regions and in 20% (6 out of 30), they were the highest in regions with or near a small/medium city. The gap of age dependency ratios between remote regions and other rural regions is particularly substantial in Denmark (15 percentage points) and Portugal (12 p.p.).
Note: Based on available data for 2 147 TL3 regions.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Ageing also has gender variations across types of regions. Remote regions comprise a lower share of females amongst the elderly (0.83 elderly males per elderly female) than other rural regions (0.76). Overall, females are over-represented amongst the elderly age group given their longer longevity but less so in remote places.
In conclusion, most of the population in rural regions have a strong interaction with urban economies. The share of metropolitan vis-à-vis rural regions has been increasing in almost all OECD countries. Yet in half of OECD countries, remote regions are losing population and one-third of regions near a city are losing population. Fertility rates appear to important drivers of the population for remote regions and migration flows for metropolitan regions and their surrounding areas. Rural regions face stronger ageing pressures than metropolitan regions. The highest pressures are in remote regions.
OECD countries and regions have faced a number of structural transformations over the past decades creating opportunities and challenges. The intensification of globalisation has delocalised many production tasks to emerging economies where labour costs are cheaper against capital-intensive ones contributing to the emergence of complex global value chains (GVCs). This delocalisation has contributed to the tertiarisation of economic activities across OECD countries, in which the relative share of services increased. Services nowadays represent around 80% of value-added across OECD countries increasing by 15 percentage points relative to the share of services 15 years ago.
These two interconnected forces have not been neutral in space. Manufacturing ceased to be the economic base of large cities against service-oriented activities because they require a pool of specialised labour, access to capital and knowledge networks that are found in cities, especially large ones. This transformation benefitted cities while low-density regions faced increased competition in tradeable goods over the past decades.
Beyond this structural transformation, territories are facing the effects of a number of economic shocks including the 2008 global financial crisis and the COVID-19 pandemic. Low-density regions that produce a limited range of goods and services have a greater vulnerability to economic shocks, whether positive or negative. All things equal, in a very large, dense economy, the greater range of activities typically offers a greater degree of resilience to external shocks.
This section examines the economic performance of TL3 regions since the early 2000s. It focuses particularly on the effect of the crisis on incomes, employment and productivity and examines the effects of these structural transformations on spatial inequality and the economic structure of regions.
The global financial crisis accentuated regional disparities in most OECD countries
The effects of growing and sustained inequalities have come to the forefront of the policy debate. In the past, spatial inequalities were regarded as a natural process of development, given that denser areas benefit more from economies of agglomeration yielding higher levels of productivity, wages and living standards than lower-density areas. Policy responses have focused on mitigating inequalities within cities (OECD, 2016[29]) but recently, the attention switched to the effects of growing and sustained territorial inequalities that bring about a “geography of discontent” especially during the aftermath of the global financial crisis (Dijkstra, Poelman and Rodríguez-Pose, 2018[30]; Hendrickson, Muro and Galston, 2018[31]; McCann, 2019[3]). Analysis in this section is limited to data availability at the regional level, up to 2017, thus capturing only the effects of the aftermath of the global financial crisis.
In 2017, regional disparities, measured as the difference between the top 20% and bottom 20% of regions in GDP per capita level, are substantial across many OECD countries. The absolute gap between incomes in top versus bottom region was highest in France, Germany, Norway and the United Kingdom and lowest in Hungary, New Zealand and Portugal (Figure 2.13). In 16 out of 26 countries with available data, per capita incomes in top regions were more than double that of bottom regions.
Regional inequality increased in 24 out of 28 OECD countries with available data in the post-global financial crisis period (2008-17) compared to the pre-crisis period (2000-07). The relative decline in regional performance in Greece, Italy and Portugal occurred in a context of severe austerity measures in the years following the crisis. The distributional impacts of public spending cuts may have affected bottom-performing regions the most because many regions with high unemployment tend to have relatively high concentrations of public sector jobs.
In absolute terms, the change in regional inequalities was largest in Slovakia, Poland, Lithuania and Czech Republic and the United Kingdom, where the gap in per capita incomes between top and bottom regions increased by at least USD 6 000 between 2000-07 and 2008-17. High levels of regional inequality coincide with a substantial number of high-growing regions of all types. This is due to substantial differences in regional growth rates, as well as the varied composition of top and bottom regions across countries. For instance, in Norway, 5 out of 6 bottom regions are remote, while in Germany only 9 out of 106 bottom regions are remote (and 52 are metropolitan). Across all countries, 30% of metropolitan regions, 24% of regions near a large city, 18% of regions with or near a small/medium city, and 28% of remote regions are bottom regions. (Figure 2.14). Importantly, however, an increase or a decrease in spatial inequality by itself is not necessarily a negative outcome. If spatial inequalities increase because the top regions become better off and the rest of regions remain as they were, the increase is not necessarily a negative outcome. If, on the other hand, inequalities increase because bottom or top regions fall further behind, the rise in inequality signals a problem.
In the case of the 24 countries where regional inequality increased, changes were driven by improvements in top regions in most cases. The exception was Greece, where bottom regions were worst off in terms of income per capita in 2017 compared to 2000. Amongst the four countries reducing inequality, it was in one case (Portugal) due to top regions falling behind. In Austria, Belgium and Finland, larger inequalities went along with an improvement of the bottom regions and, in Switzerland, with a worsening of both top and bottom regions’ per capita incomes.
Note: Top (bottom) refers to top (bottom) 20% regions with the highest (lowest) GDP per capita levels with populations adding up to at least 20% of the national population. 2016 values for France and Japan. Based on available data for 1 512 TL3 regions.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
The relative decline in regional performance in Greece, Italy and Portugal occurred in a context of severe austerity measures in the years following the crisis. The distributional impacts of public spending cuts may have affected bottom-performing regions the most because many regions with high unemployment tend to have relatively high concentrations of public sector jobs.
In absolute terms, the change in regional inequalities was largest in Slovakia, Poland, Lithuania and Czech Republic and the United Kingdom, where the gap in per capita incomes between top and bottom regions increased by at least USD 6 000 between 2000-07 and 2008-17. High levels of regional inequality coincide with a substantial number of high-growing regions of all types. This is due to substantial differences in regional growth rates, as well as the varied composition of top and bottom regions across countries. For instance, in Norway, 5 out of 6 bottom regions are remote, while in Germany only 9 out of 106 bottom regions are remote (and 52 are metropolitan). Across all countries, 30% of metropolitan regions, 24% of regions near a large city, 18% of regions with or near a small/medium city, and 28% of remote regions are bottom regions.
Note: 2017 extrapolated values for France and Japan based on 2001-16 regional growth rates. Based on available data for 1 629 TL3 regions. Unlabelled cases are better off (top or bottom). Top (bottom) refers to top (bottom) 20% regions with the highest (lowest) GDP per capita levels (PPP) with populations adding up to at least 20% of the national population.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
The distance penalty widened after the global financial crisis
The global financial crisis occurred more than a decade ago. Although the crisis affected all regions, the recovery has been much slower for rural economies. Low population growth, slow employment creation and sluggish productivity appear to be working against the recovery in hard-hit rural regions. This trend has been especially stark in regions far from large cities, which are diverging from other regions in terms of productivity and incomes, and in regions with a small/medium city or near one, where employment rates have fallen behind.
A well-established fact is that per capita income and productivity levels are higher in higher density areas across OECD countries due to the benefits associated with economies of agglomeration (OECD, 2015[32]; OECD, 2016[29]) The alternative TL3 typology provides further evidence on this well-known fact. It shows how incomes per person, productivity and employment rates decrease as distance to high-density areas increases (Table 2.5). The gaps between regions near a large city and the group of regions far from large cities are substantial:
Regions near a large city have a gap in GDP per capita with respect to metropolitan (large) regions of nearly USD 4 600 (USD 18 000). Their productivity levels and employment rate are around 10 and 8 percentage points below OECD average levels respectively.
The gaps for regions with or near a small/medium city are even larger. With respect to GDP per capita, they are 28 percentage points below the OECD average. In terms of productivity and employment rates, the gap is also still significant, at 20 and 14 percentage points.
For remote regions, the gap is 21 percentage points below the OECD average in GDP per capita, 14 percentage points in labour productivity and 3 percentage points in employment rates.
Convergence was brought to a halt by the crisis
The current gap between metropolitan and rural regions in GDP per capita is the result of long-standing differences that accentuated after the financial crisis of 2008, especially for regions far from large cities (Figure 2.15). Regions near a large city, in contrast, maintained and even marginally reduced their gap in GDP per capita gap with respect to the OECD average.
The global financial crisis had an asymmetric impact across region types and brought regional convergence to a halt. Before the crisis, regions far from large cities were growing faster than other region types. The crisis clearly slowed down growth rates across all region types (as seen by comparing the slope of the lines in Figure 2.16). The decline, however, was much higher in regions with or near a small/medium city and remote regions (as seen by comparing the slope of the lines connecting the dotted and full bubbles). Meanwhile, metropolitan regions and their surrounding regions weathered the effects of the crisis better than the rest of the regions.
One of the factors contributing to the resilience of metropolitan regions is the presence of skilled labour (Crescenzi, Luca and Milio, 2016[33]). On the other hand, the disproportionate effect of the crisis in regions far from large cities is related to their thinner and less diversified economic base (OECD, 2016[7]). To the effect of decreased and fragmented internal demand, low-density economies have faced competitive pressures from low-wage emerging economies over the past two decades. Without increased exports, the sources of productivity gains have remained limited for remote regions. Moreover, because low-density regions produce a limited range of goods and services, they are more vulnerable to industry-specific shocks that are neutralised by a broader and more diversified economic base in larger and denser regions.
Note: 2017 extrapolated values for France and Japan based on 2001-16 regional growth rates. Based on available data for 1 536 TL3 regions in 28 countries. GDP is in USD PPP with the base year 2015.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Note: 2017 extrapolated values for France and Japan based on 2001-16 regional growth rates. Based on available data for 1 530 TL3 regions in 28 countries. GDP is in USD PPP with the base year 2015.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
The global financial crisis brought about a starker division between winners and losers in terms of GDP per capita growth across rural regions. Before the crisis, most regions experienced growth in income per capita and there was convergence within each type as evidenced by higher growth rates in regions with initially lower income per capita levels (Figure 2.17). After the crisis, the variability in growth performance increased across all rural region types. A considerable number of regions far from large cities achieved relatively high growth in a broader context of sluggish economic growth.
In aggregate terms, as of 2017, 85% of large metropolitan regions, 87% of the metropolitan regions and 83% of regions near a large city had already recovered to pre-crisis levels in GDP per capita (Table 2.6). In contrast, only 69% of regions with or near a small or medium city and 74% of remote regions had recovered.
While the success recovery stories accrued all types of regions, they were highly concentrated in Germany and Poland. About 71% of regions in which GDP per capita in 2017 was at least 25% larger than in the pre-crisis period were from these 2 countries (Table 2.7). Germany concentrated about three in four high-growth regions near a large city and one out of two high-growth remote regions.
Note: 2017 extrapolated values for France and Japan based on 2001-16 regional growth rates. Based on available data for 28 countries. GDP is in USD PPP with the base year 2015.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
In contrast, regions suffering the highest economic decline were mostly rural regions, and most of them were in Greece or Italy. Three-quarters of the population of Greece (75%) and 38% of the population of Italy lived in regions where income per capita in 2017 was still 10% lower than during the pre-crisis period (Table 2.7). Overall, the lack of recovery was more frequent in regions far from large cities.
Productivity in remote regions drifted away after the crisis
After the shock of the financial crisis, labour productivity started to converge slowly in regions near cities but drifted away in remote regions (Figure 2.18). One explanation for this divergence is that productivity in regions with a small- and medium-sized city or near one may have benefitted from agglomeration benefits, though not on the same scale as metropolitan regions. In contrast, further concentration of productive industries in cities translated into productivity losses in remote regions that were highly dependent in a few industries with lower than average productivity performance.
Higher productivity levels in metropolitan regions compared to rural regions is the norm across OECD countries. Aside from Korea, all OECD countries with available data show higher productivity in metropolitan regions compared to rural regions (Figure 2.19). The difference is especially stark in small East European and Baltic countries with relatively low productivity levels (Latvia, the Slovak Republic, Lithuania and Estonia), where productivity in metropolitan regions is at least 50% higher than in rural regions. In contrast, the productivity gap is narrow in countries with diverse productivity levels, including Spain, Hungary, Denmark, Japan, Slovenia and Austria.
Note: 2017 productivity measured as GVA per worker. GVA extrapolated values for France, Japan and New Zealand based on 2001-16 regional growth rates. 2017 employment extrapolated values for France, Japan and Switzerland based on 2001-16 regional growth rates. Based on available data for 1 346 TL3 regions in 22 countries. GVA is in USD PPP with the base year 2015. Employment at place of work.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Note: 2017 GVA extrapolated values for France, Japan and New Zealand based on 2001-16 regional growth rates. 2017 employment extrapolated values for France, Japan and Switzerland based on 2001-16 regional growth rates. Based on available data for 1 491 TL3 regions. GVA is in USD PPP with the base year 2015. Employment at place of work.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Rural economies struggle to create employment in the new service economy
The contribution of rural regions to employment growth declined significantly after the crisis. In 2001-07, rural regions contributed 22% to an employment growth rate of 7.5%, similar to their contribution to GDP and GVA, and above their contribution to population growth (18%) (Table 2.8). After the crisis, the contribution of rural regions to employment growth fell to 7%, meaning that more than 90% of employment growth was contributed by metropolitan regions in the post-crisis period. The drop in contribution was particularly big for regions with or near a small/medium city, which moved from a contribution of 8% to a negative contribution of 0.9% after the crisis.
The closing of the productivity gap in regions with or near a small/medium city after the crisis is at odds with a diverging trend in employment rates. Even before the crisis in 2008, employment rate levels in regions with or near a small/medium city drifted away from other types of rural regions (Figure 2.20). In 2013, the employment rate gap reached a minimum of 16% below OECD levels. In contrast, in the same year, employment rates in large metropolitan regions were 7% above OECD levels.
The stark difference between remote regions and regions with or near a small/medium city may be due to the mobility and size of the working-age population. Small and medium cities have relatively smaller pools of workers and a different demographic composition, which means more competition for existing job posts. With slow employment creation, workers in small and medium cities may decide to wait for employment opportunities instead of migrating, as cities allow them to access health, education and other services. Policy responses can focus on addressing some structural challenges in smaller cities to tackle the lack of new employment opportunities.
Note: 2017 employment extrapolated values for France, Japan and Switzerland based on 2001-16 regional growth rates. Based on available data for 1 395 TL3 regions in 23 countries. Employment at place of work. Employment rates based on employment at place of work over working- age (15-64 year-old) population.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
The territorial disparities in employment performance have occurred in the context of a general and steady increase in the importance of services against manufacturing and agriculture. The share of total employment in services grew across all region types after 2008 but remote regions experienced faster tertiarisation (Figure 2.21). In fact, in 2017 the share of employment in services in remote regions was only 4 percentage points below the corresponding share for metropolitan regions with a city of 250 000 people or more (71% versus 75%).
Note: 2017 employment extrapolated values for France based on 2001-16 regional growth rates. Based on available data for 1 403 TL3 regions in 24 countries.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Bottom-performing regions remain over-specialised in primary sectors agriculture
Despite a strong tertiarisation trend, rural regions continue to be specialised in primary sectors, including agriculture, forestry and fishing. Although the share of primary sector employment is over-represented across all rural region types, a larger proportion of regions far from large cities show relatively high levels of employment specialisation (i.e. a specialisation index larger than 2) (Figure 2.22). On the other hand, rural regions have similar patterns of specialisation in manufacturing, which are in line with the median levels of specialisation in metropolitan regions.
Regions with very large cities are in the best position to reap the benefits of specialisation in high-value-added services. The productivity of services tends to increase in large cities with access to a pool of specialised labour and knowledge networks. Furthermore, many service-oriented businesses are less vulnerable to offshoring and therefore protected from international competition. To no surprise, large metropolitan regions are more specialised in high-value-added services that lower-density areas (Figure 2.22).
The slow-down in trade brought about by the financial crisis made regions far from cities more dependent on internal markets, as they cater a more limited range of the goods and services. These features make rural regions less prone to specialise in high-value-added services. In fact, bottom rural regions are overly specialised in primary sectors, while top rural regions are specialised in high-value-added services (Figure 2.23).
Note: Bars indicate median values. Each dot represents a TL3 region. High-level value-added services include information and communication, financial and insurance activities, and professional, scientific, technical, administrative, support service activities. Data for 19 OECD countries with available data (16 for high value-added services).
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Note: Bars indicate median values. Each dot represents a TL3 region. High-level value-added services include information and communication, financial and insurance activities, and professional, scientific, technical, administrative, support service activities. Top and bottom regions refer to the 20% highest and lowest-ranked regions in terms of GDP per capita in each country, accounting for at least 20% of the population. Data for 22 OECD countries with available data (19 for high value-added services).
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Women increasingly participate in rural labour markets
The evolution of employment after the crisis has favoured occupations that disproportionally employ women. While female employment rates had recovered their pre-crisis levels across all rural region types by 2014, male employment rates continue to be below 2007 levels across all region types (Figure 2.24). Employment rates of males were particularly slow to pick up in regions far from metropolitan regions. In regions with access to a small/medium city, female employment rates were 4 p.p. above 2007 levels, while male employment rates were 5 p.p. below.
These diverging trends relate to broad structural changes that have had localised impacts on rural labour markets. In general, rural labour markets tend to be divided by gender, with women more represented in lower-wage services sector jobs (e.g. health and social care services) and men more represented in higher wage primary sectors and associated manufacturing (e.g. agriculture, forestry and mining). Ongoing structural change in primary sectors and rural manufacturing have contributed to increasing differences between employment rates for men and women in regions far from large cities.
Note: Based on data available for 776 TL3 regions in 14 countries. Employment rate calculated as employment at place of residency over the working-age (15-64 year-old) population of each gender.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
The productivity paradox intensified in rural regions after the crisis
Productivity gains can be powerful engines of social transformation but can also be a vehicle for wider gaps across regions if they occur in a context of job-less growth (OECD, 2018[34]). The reasons why productivity gains do not translate into employment gains are manifold. One reason is a more difficult adjustment in the labour market following re-adjustments across and within industries. This is the case if the economic crisis brought about a reorientation towards industries intensive in highly specific skills (e.g. programming and data science) that are difficult to acquire for certain workers. Another reason is the structural unemployment arising from the exit of unproductive firms and unproductive workers that are not absorbed by more productive local firms that may source labour abroad or replace labour with capital.
Across OECD countries, 60% of employment concentrated in regions that experienced productivity and employment gains simultaneously. This “gain-gain” situation was far more common in metropolitan regions than in rural regions (Figure 2.22). Meanwhile, the mismatch between employment and productivity gains became more prevalent in all types of regions after the crisis but more pervasive in regions far from large cities. In the post-crisis period, about half of employment in regions with or near a small/medium city (51%) and remote regions (57%) concentrated in regions with employment and productivity gains.
The “productivity paradox”, a scenario of productivity gains with low employment, intensified outside metropolitan regions after the crisis. Indeed, in 2008-17, 18% of employment was concentrated in regions that experienced employment losses in the presence of productivity gains. While this touches all region types, it was more prevalent in regions with or near a small/medium city (concentrating 25% of employment) and in remote regions (28%).
Note: Based on available data for 1 225 TL3 regions in 20 countries. Productivity measured as GVA per worker.
Source: OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.
Moreover, employment and productivity losses combined affected total employment more strongly in rural regions. Within rural regions, remote regions had the biggest drop in productivity of almost a full percentage point the average negative rates of regions near a large city. Regions with or near a small/medium city (accounting for 10% of regions of this type) had the biggest drop in employment.
The relationship between regional productivity and employment growth varies widely across OECD countries. Table 2.9 shows the split of regions in each country between different scenarios in terms of employment and productivity growth in the post-crisis period. Several conclusions emerge:
The win-win situation of productivity growth paired with employment creation occurred in most regions of Austria, Belgium, Germany, the Netherlands, New Zealand, Sweden and East European countries, including the Czech Republic, Poland and the Slovak Republic. This is consistent with the evidence of concentration of rapid recovery from the economic crisis of 2008 in Germany, Eastern Europe and in northern European regions (OECD, 2018[35]).
In Hungary, Latvia, Lithuania, Portugal and Spain, more than half of employment occurred in regions where productivity gains occurred without employment gains. In the most extreme case, all regions in Portugal experienced productivity gains paired with employment losses.
Greece and Italy stand out as the countries concentrating the bulk of regions in decline. In Italy, about one-quarter of employment (27%) occurred in regions that experienced employment and productivity losses. In Greece, 46% of regions had both employment and productivity losses.
In conclusion, regions far from large cities were growing faster than the national average before the crisis, but the crisis brought convergence to a halt. In contrast, regions near a large city have shown more resilience and have performed as well as metropolitan regions after the crisis. The increases in productivity in regions far from large cities were accompanied by labour shedding in many cases. Regions with access to smaller cities experienced the largest drops in employment, with effects likely coming from the effect of international competition on tradeables. As ongoing trade tensions between countries can disproportionately affect these types of regions, there is an urgent need to restructure their economies toward sectors that can create local employment while adding value. On the other hand, large cities and their surrounding regions have weathered the effects of the crisis better than the rest of the regions.
The economic consequences of the ongoing COVID-19 pandemic threaten the incipient recovery of lagging regions in countries badly hit by the financial crisis. The negative shock of the ongoing health crisis will impact rural industries including tourism and agriculture and disproportionally affect the most vulnerable, including temporary and self-employed rural workers. Appropriate and timely place-based policy responses should go in the direction of bridging gaps and containing the increase in inequality across people and places, in order to ensure social cohesion and stability.
Innovation is today a major driving force for economic growth across OECD countries. The speed of innovation generation is constantly increasing, making innovation a basic requirement for national and regional competitiveness. Skilled human capital along with sound information and communication technology (ICT) and civil infrastructure are cornerstones to developing an ecosystem that sparks innovation at the local level.
Skills and human capital
Human capital and skills are critical drivers of regional growth and this is particularly challenging for rural regions that may suffer from “brain drain” (OECD, 2012[36]). Cities attract high-skilled workers from over the globe due to their amenities, presence of economies of agglomeration and higher paid jobs especially in services. In contrast, the market for low and technical skills is much more locally driven. This suggests that the productivity of rural economies depends on the successful upgrading of low-skill workers and an increase in workers with technical skills. Research finds strong benefits of reducing the share of low-skilled workers in the regional labour force supports economic growth (OECD, 2012[36]).
The quality and accessibility of rural education have a double role to play in addressing gaps in skills: starting from children’s early years, high-quality education and care can help raise outcomes in education and the labour market. At the same time, access to public services, such as childcare and schools, is a locational factor shaping the attractiveness of rural places, including for highly skilled workers. This also means that a lack of access to high-quality education and training provision in rural places can aggravate the rural-urban divide with regard to skill levels.
Low levels of high-skilled workers can be a bottleneck for growth in low-density economies. For instance, across European countries, individuals living in rural regions strongly lag behind their peers in cities with regard to their level of digital skills, paramount for many modern workplaces (Figure 2.26). Educational attainment provides another indicator of the average skill level in the labour force. The share of workers with tertiary education, i.e. a university degree, is lower in regions characterised by low-density economies, while the share of workers that do not have education beyond primary education (a proxy for low-skilled workers) tends to be higher in these regions (OECD, 2016[7]). Across all countries considered, the share of workers with tertiary education in the most urbanised regions is higher than in low-density regions ranging from 57 percentage points higher in the Czech Republic to 2.8 in the United States (US). All OECD countries have a higher share of primary educated workers in low-density regions except for Germany and the US. In Germany, this partly reflects the historic east-west divide and the significantly lower shares of workers with only primary education in the (less-densely populated) east of Germany and, for the US, the difference can be driven by states that are mostly urbanised and have a large percentage of foreign-born residents.
Note: Not all OECD countries covered by the data source. For further information on the Eurostat classification of areas by degree of urbanisation, see https://ec.europa.eu/eurostat/web/degree-of-urbanisation/background.
Source: Eurostat (2020[37]), European Social Survey, https://ec.europa.eu/eurostat/cros/content/european-social-survey_en.
In terms of the level of skills of students, results from the Programme for International Student Assessment (PISA) show that students in rural schools, defined as villages, hamlets or rural areas with fewer than 3 000 people, tend to underperform in secondary education outcomes in comparison to cities that have more than 100 000 inhabitants (Echazarra and Radinger, 2019[38]). On average, students in city schools across OECD countries scored 48 points higher in reading than their peers in rural schools, according to the PISA 2018 data – more than the equivalent of a year of schooling (new analysis of PISA 2018 data adapted from Figure 2.27). Yet, when the comparison accounts for the socio-economic status of students and schools, the performance gap between rural and city schools was no longer statistically significant. This means that differences in the socio-economic composition of the population tend to explain the rural-urban gap in academic performance.
Note: Rural schools are schools in villages, hamlets or rural areas with 3 000 or fewer inhabitants; city schools are those in places with a population of more than 100 000. Results based on linear regression models. Statistically significant coefficients are marked in a darker tone.
Source: OECD PISA 2018 database, adapted from Echazarra, A. and T. Radinger (2019[38]), “Learning in rural schools: Insights from PISA, TALIS and the literature”, https://doi.org/10.1787/8b1a5cb9-en.
The rural-urban education gap is even more visible when analysing rural students’ educational expectations. Based on a survey among 15-year-old students carried out by PISA 2018, on average across OECD countries, students in rural schools are half as likely to expect completing a university degree as those in city schools (new analysis of PISA 2018 data adapted from Echazarra and Radinger (2019[38])). This reflects students’ self-assessment of their opportunities and capacities regarding higher education (OECD, 2017[39]). In that sense, beyond financial facilities, other factors might discourage students in rural areas to advance further in their studies, including geographical barriers, lack of career role models and highly skilled jobs in their home areas.
Attracting highly skilled teachers to rural areas is key to improve student outcomes. While differences in the highest level of education are on average not statistically significant between rural and city schools OECD countries, there tends to be a greater share of new teachers and a higher turnover rate in rural schools (OECD, 2020[40]). As teachers in rural schools also tend to be more satisfied with their salaries and tend to report less stress than their peers in cities, policy makers need to take a broader approach to measures to attract and retain teachers to those locations that go beyond financial incentives. Those trends vary across countries but they highlight that a spatial lens is warranted when considering the support teachers need to deliver high-quality education in different locations, for instance, to enable collaborative professional learning when schools are small.
Policies related to skills development and education cannot be spatially blind across countries’ territories and must address rural regions’ specific challenges related to lower densities and longer distances in developing a strong skill base for the future local economic development.
Internet connectivity is key for the future of rural regions
Advances in technology and particular Internet infrastructure are quite relevant for low-density regions. Improvements in Internet connectivity can overcome some of the core challenges they face including isolation, high transportation costs, high costs to delivery services and distance to markets. Most Internet infrastructure investments were initially deployed in urban areas given their higher densities and commercially viable solutions. Over the past years, further improvements in ICT technology will have a proportionally higher impact in low-density regions since most urban areas are already well connected.
Furthermore, confinement measures during the Covid-19 crisis have fomented the use of teleworking, remote learning and e-services. These practices will accelerate the usage of these digital tools beyond the crisis period. With changing habits and more willingness to embrace these digital tools, government and private operators may increase investments to realise their potential benefits. In rural economies, the increased connectivity of services can further unlock opportunities for future work, synergies and regional integration between rural places and their surroundings.
In order to benefit from Internet infrastructure deployments, a multidimensional response is needed (as will be argued in Chapter 5); deployment by itself is a necessary but not sufficient condition to reap the potential benefits of Internet connectivity and the potential benefits for rural regions. These range from attracting new economic activity and skills, improving the productivity of firms, raising the quality and reducing costs of service delivery, connecting to a new market and overcoming isolation.
Economic remoteness, or peripherality, has three distinct features:
The first is simple physical distance to major markets. This increases travel times and shipping costs, which must be borne by the buyer (in the form of higher prices) or seller (in the form of lower margins).
The second dimension of peripherality is the degree of economic connectedness. Lack of economic integration not only reduces current trade opportunities but it also reduces the ability of agents in a place to identify new opportunities. Thus, there are costs in both static and dynamic perspectives.
Third, the economic structures of such places often have specific features. Production is concentrated in relatively few sectors since it is impossible to achieve “critical mass” in more than a few activities. Whatever the respective roles of the primary, secondary and tertiary sectors, a narrower economic base implies greater vulnerability to sector-specific shocks, whether positive or negative.
Broadband access is today a needed asset for economic progress and well-being. Quality broadband is instrumental to harness the benefits from new technologies, including the Internet of things, blockchain, artificial intelligence, big data and 5G networks (see Chapter 5).
Broadband access in rural areas has increased across OECD countries. Since 2010, the gap of broadband access between rural and urban areas, as defined for this measure, has decreased by half in almost all OECD countries (OECD, 2019[41]). In 2018, the average share of rural households with broadband connection in a sample of 31 OECD countries reached 82%, slightly below the 89% in urban areas (Figure 2.28).
Note: Brazil (2010, 2016), Chile (2017, 2012), Switzerland (2017, 2012), United States (2010, 2017). For EU countries, rural areas are those with a population density less than 100 per km². For Canada, rural areas are those with a population density less than 400 per km². For the United States, rural areas are those with a population density less than 1 000 per square mile or 386 people per km².
Source: OECD (2019[42]), Measuring the Digital Transformation: A Roadmap for the Future, https://dx.doi.org/10.1787/9789264311992-en; OECD (2018[43]), ICT Access and Usage by Households and Individuals (database), http://oe.cd/hhind.
In terms of speed capacity, there is still a gap between rural and urban regions. Based on data of 27 OECD countries for 2017, just 56% of rural households have access to fixed broadband with a minimum speed of 30 Mbps (speed required to support many consumer applications such as streaming high-definition video), in comparison to over 85% in urban areas (Figure 2.29). In countries like Finland for instance, while the share of rural households with an Internet connection is almost 90%, just 8.3% of households in rural areas had a connection to quality broadband. Slow or intermittent broadband connection reduces the opportunities for people to participate and benefit from economic gains and quality of life in the digital age.
Note: 2016 data for the United States.
Source: OECD (2019[42]), Measuring the Digital Transformation: A Roadmap for the Future, https://dx.doi.org/10.1787/9789264311992-en; OECD calculations based on CRTC (2017[44]), Communications Monitoring Report 2017, https://crtc.gc.ca/eng/publications/reports/policymonitoring/2017/cmr.htm; EC (2017[45]), Study on Broadband Coverage in Europe 2017, European Commission; FCC (2018[46]), 2018 Broadband Deployment Report, https://www.fcc.gov/reports-research/reports/broadband-progress-reports/2018-broadband-deployment-report; https://doi.org/10.1787/888933915050.
Rural economies lag in innovation as measured by patent activity
Innovation encompasses a wide range of activities from research and development (R&D) to organisational changes, training, testing, marketing and design. The Oslo Manual recognises four types of innovation: product innovation, process innovation, marketing innovation and organisational innovation (OECD, 2015[32]). Despite this broad definition, due to the availability of data patent application, a type of intellectual property rights (IPs) remains the most common indicator to measure innovation performance. Not only do they focus on a subset of innovation (science and technology) but there are measurement biases driven by the location of where the patent is recorded against where it was conceived.
Innovation performance, based solely on patent activity, is lower in rural regions compared to metropolitan regions (Figure 2.30). In 2016, the average number of patents per 10 000 inhabitants in metropolitan regions (1.9) almost doubled the number in regions near a large city (1.0) on average across OECD countries. Patent activity is even lower in regions with or near a small/medium city (0.6) and remote regions (0.5). However, out of 30 OECD countries with metropolitan areas, 6 countries (Chile, Hungary, Italy, Mexico, Slovenia, United Kingdom) exhibit more patent activity in at least one type of rural region compared to metropolitan regions. In the United Kingdom, for instance, regions near a large city, including those with university towns such as Cambridge and Oxford, display higher patent intensity than metropolitan regions.
Better data to measure innovation performance at the local level is needed to assess the different regional dynamics. Patents mainly measure the front-end – or invention – of the innovation process, giving less indication on the back-end or the commercialisation. Thus, patent data tends to overlook the firms that only apply existing technologies to their operations, without engaging in technological development that leads to a patentable invention (OECD/Eurostat, 2018[47]). Furthermore, not all technological development activities result in patentable inventions and firms do not seek patent protection for all of their inventions. Thus, measuring innovation through patents or IPs can penalise rural places since these metrics do not fully measure grassroots or user-developed innovation, which may be more important to rural firms (Whitacre, Meadowcroft and Gallardo, 2019[48]; Wojan and Parker, 2017[49]). Therefore, there is a need to come up with tailored indicators that are able to canvas how rural business innovate or use technologies in innovative ways.
Note: Variable constructed using total population and fractional count of patents.
Source: International Energy Agency (2018[50])CO2 Emissions from Fuel Combustion 2018; OECD (2020[17]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en; OECD (2019[41]), Going Digital: Shaping Policies, Improving Lives, https://doi.org/10.1787/9789264312012-en.
USDA undertook a broad study to better understand innovation in rural areas using a variety of data sets, including a large-scale survey that compares innovation in urban and rural areas based on some 11 000 business establishments with at least 5 paid employees in tradeable industries in rural and metropolitan areas. The survey divides business into either: i) substantive innovators, which amount to 30% of business launching new products and services, making data-driven decisions and creating intellectual property worth protecting; ii) nominal innovator, which comprise 33% of firms and engage in the more incremental improvement of their products and processes; or iii) non-innovators, amounting to 38% of firms showing little or no evidence of innovation. The study finds:
Establishments in metropolitan areas are slightly more innovative – around 20% of rural firms are substantive innovators, compared to 30% of firms in metropolitan areas.
Rural areas have a slightly higher overall rate of substantive innovation for large firms (those with 100 employees or more), while urban areas win out in their rate of substantive innovation by small- and medium-sized firms.
Rural areas have a slight advantage over their metropolitan counterparts in the rate of substantive innovation by the most innovative firms (those that are patent-intensive). That is because innovation in rural areas tends to be a product of patent-intensive manufacturing in industries like chemicals, electronics and automotive or medical equipment, while urban areas have higher rates of innovation in services.
Source: Wojan, T. and T. Parker (2017[49]), Innovation in the Rural Nonfarm Economy: Its Effect on Job and Earnings Growth, 2010-2014, https://doi.org/10.22004/ag.econ.264596 (accessed on March 2020).
Digitalisation and automation are the main global trends that will affect rural economies. The effects of these trends can radically transform life and work for rural inhabitants (see Chapter 5). Detailed data and indicators to measure the impacts of digital transformation at the local level will be instrumental for policies to adapt and make the most of the technological change.
This chapter has outlined population and economic trends shaping rural development and the status of skills, human capital, digital connectivity and innovation as enabling factors for rural development.
A policy roadmap for delivery of well-being in rural regions has to take into account the variety of development profiles of rural regions in OECD countries. The alternative typology of regions has uncovered the differentiating role of access to density in the economic performance of rural regions, particularly after the 2008 global financial crisis. In more remote regions, policies will have to place emphasis not only on bridging the “distance penalty” through the provision of quality and affordable digital access for people and entrepreneurs but on designing overarching policies targeting rural attractiveness that nurture existing and new economic activities. These plans can take advantage of new economic opportunities generated by the transition to a low-carbon economy, new business opportunities in the care sector and social innovation initiatives. Furthermore, although the effects of the COVID-19 crisis will likely deepen territorial inequalities, they will also potentially accelerate some megatrends, in particular digitalisation.
Confinement measures have brought changing habits and more willingness to embrace digital tools. Government and private operators will likely increase investments to realise their potential benefits. In rural areas, the increased connectivity of services can further unlock opportunities for future work, synergies and territorial integration.
In regions close to cities, policy strategies can leverage the natural attractiveness of proximity to dense labour and consumer markets by focusing on high-quality affordable housing and services, the attraction of high-value-added service industries and co-ordination solutions to maximise rural-urban linkages. Across all rural regions, ambitious and urgent strategies to increase digital skills and connectivity are required to bridge development gaps with metropolitan regions.
References
[25] Andersson, G. et al. (2009), “Cohort fertility patterns in the Nordic countries”, Demographic research, Vol. 20, pp. 313-352,https://doi.org/10.4054/DemRes.2009.20.14.
[12] Australian Bureau of Statistics (2018),1270.0.55.005 - Australian Statistical Geography Standard (ASGS): Volume 5 - Remoteness Structure, July 2016, https://www.abs.gov.au/ausstats/[email protected]/mf/1270.0.55.005.
[11] Australian Bureau of Statistics (n.d.), The Australian Statistical Geography Standard (Asgs) Remoteness Structure, https://www.abs.gov.au/websitedbs/D3310114.nsf/home/remoteness+structure.
[8] Barbier, M., G. Toutin and D. Levy (2016), “L’accès aux services, une question de densité des territoires”, Insee Première, No. 1579, Insee, https://www.insee.fr/fr/statistiques/1908098.
[23] Batista e Silva, F. et al. (2016), Regionalisation of Demographic and Economic Projections, Joint Research Centre, European Commission, https://op.europa.eu/fr/publication-detail/-/publication/f6155238-3f4e-11e6-af30-01aa75ed71a1.
[14] Brezzi, M., L. Dijkstra and V. Ruiz (2011), “OECD Extended Regional Typology: The Economic Performance of Remote Rural Regions”, OECD Regional Development Working Papers, No. 2011/6, OECD Publishing, Paris, https://dx.doi.org/10.1787/5kg6z83tw7f4-en.
[33] Crescenzi, R., D. Luca and S. Milio (2016), “The geography of the economic crisis in Europe: National macroeconomic conditions, regional structural factors and short-term economic performance”, Cambridge Journal of Regions, Economy and Society, Vol. 9/1, pp. 13-32, https://doi.org/10.1093/cjres/rsv031.
[44] CRTC (2017), Communications Monitoring Report, Canadian Radio-television and Telecommunications Commission, https://crtc.gc.ca/eng/publications/reports/policymonitoring/2017/cmr.htm.
[15] Dijkstra, L. and H. Poelman (2014), “A harmonised definition of cities and rural areas: The new degree of urbanisation”, European Commission Regional Working Papers, European Commission.
[30] Dijkstra, L., H. Poelman and A. Rodríguez-Pose (2018), “The geography of EU discontent”, Regional Studies, Accesed on 15 July 2019, https://doi.org/10.1080/00343404.2019.1654603.
[19] Duranton, G. et al. (2004), “Micro-foundations of urban agglomeration economies”, pp. 2063-2117.
[45] EC (2017), Study on Broadband Coverage in Europe, European Commission.
[18] EC (n.d.), Global Human Settlement Layer 2015, European Commission, https://ghsl.jrc.ec.europa.eu/.
[38] Echazarra, A. and T. Radinger (2019), “Learning in rural schools: Insights from PISA, TALIS and the literature”, OECD Education Working Papers, No. 196, OECD Publishing, Paris, https://doi.org/10.1787/8b1a5cb9-en.
[10] European Network for Urban Development (n.d.), Strategy for Inner Areas - Italy, https://enrd.ec.europa.eu/sites/enrd/files/tg_smart-villages_case-study_it.pdf.
[37] Eurostat (2020), European Social Survey, European Commission, https://ec.europa.eu/eurostat/cros/content/european-social-survey_en.
[16] Fadic, M. et al. (2019), “Classifying small (TL3) regions based on metropolitan population, low density and remoteness”, OECD Regional Development Working Papers, No. 2019/06, OECD Publishing, Paris, https://dx.doi.org/10.1787/b902cc00-en.
[46] FCC (2018), 2018 Broadband Deployment Report, Federal Communications Commission, https://www.fcc.gov/reports-research/reports/broadband-progress-reports/2018-broadband-deployment-report.
[27] Hank, K. (2001), “Regional fertility differences in western Germany: An overview of the literature and recent descriptive findings”, International Journal of Population Geography, Vol. 7/4, pp. 243-257, https://doi.org/10.1002/ijpg.228.
[31] Hendrickson, C., M. Muro and W. Galston (2018), Countering the geography of discontent: Strategies for left-behind places, https://www.brookings.edu/research/countering-the-geography-of-discontent-strategies-for-left-behind-places/ (accessed on 26 July 2019).
[22] Hirshman, A. (1958), “The strategy of economic development”, The Annals of the American Academy of Political and Social Science.
[26] Hoem, J. (2005), “Why does Sweden have such high fertility?”, Demographic Research, Vol. 13, pp. 559-572, https://doi.org/10.4054/DemRes.2005.13.22.
[1] Iammarino, S., A. Rodriguez-Pose and M. Storper (2018), “Regional inequality in Europe: Evidence, theory and policy implications”, Journal of Economic Geography, Vol. 19/2, pp. 273-298, https://doi.org/10.1093/jeg/lby021.
[50] International Energy Agency (2018), CO2 Emissions from Fuel Combustion 2018.
[24] Kulu, H. (2013), “Why do fertility levels vary between urban and rural areas?”, Regional Studies, Vol. 47/6, pp. 895-912, https://doi.org/10.1080/00343404.2011.581276.
[3] McCann, P. (2019), “Perceptions of regional inequality and the geography of discontent: Insights from the UK”, Regional Studies, pp. 1-12, https://doi.org/10.1080/00343404.2019.1619928.
[21] Myrdal, G. and P. Sitohang (1957), Economic theory and under-developed regions, Duckworth, London.
[17] OECD (2020), OECD Regional Statistics (database), OECD, Paris, https://doi.org/10.1787/region-data-en.
[40] OECD (2020), TALIS 2018 Results (Volume II): Teachers and School Leaders as Valued Professionals, TALIS, OECD Publishing, Paris, https://dx.doi.org/10.1787/19cf08df-en.
[41] OECD (2019), Going Digital: Shaping Policies, Improving Lives, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264312012-en.
[42] OECD (2019), Measuring the Digital Transformation: A Roadmap for the Future, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264311992-en.
[43] OECD (2018), ICT Access and Usage by Households and Individuals (database), OECD, Paris, http://oe.cd/hhind.
[35] 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.
[34] OECD (2018), The Productivity-Inclusiveness Nexus, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264292932-en.
[28] OECD (2017), Pensions at a Glance 2017: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/pension_glance-2017-en.
[39] OECD (2017), PISA 2015 Results (Volume III): Students’ Well-Being, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264273856-en.
[29] OECD (2016), Making Cities Work for All: Data and Actions for Inclusive Growth, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264263260-en.
[7] OECD (2016), OECD Regional Outlook 2016: Productive Regions for Inclusive Societies, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264260245-en.
[32] OECD (2015), The Innovation Imperative: Contributing to Productivity, Growth and Well-Being, OECD Publishing, Paris, https://doi.org/10.1787/9789264239814-en.
[51] OECD (2015), The Metropolitan Century: Understanding Urbanisation and its Consequences, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264228733-en.
[5] OECD (2014), OECD Rural Policy Reviews: Chile 2014, OECD Rural Policy Reviews, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264222892-en.
[13] OECD (2013), Rural-Urban Partnerships: An Integrated Approach to Economic Development, OECD Rural Policy Reviews, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264204812-en.
[36] OECD (2012), Promoting Growth in All Regions, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264174634-en.
[6] OECD (2006), The New Rural Paradigm: Policies and Governance, OECD Rural Policy Reviews, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264023918-en.
[47] OECD/Eurostat (2018), Oslo Manual 2018: Guidelines for Collecting, Reporting and Using Data on Innovation, 4th Edition, The Measurement of Scientific, Technological and Innovation Activities, OECD Publishing, Paris/Eurostat, Luxembourg, https://dx.doi.org/10.1787/9789264304604-en.
[20] Perroux, F. (1995), “Note sur la notion de pôle de croissance”, Économie appliquée.
[2] Rodríguez-Pose, A. (2018), “The revenge of the places that don’t matter (and what to do about it)”, Cambridge Journal of Regions, Economy and Society, Vol. 11/1, pp. 189-209, https://doi.org/10.1093/cjres/rsx024.
[9] Statistics New Zealand (2004), New Zealand: An Urban/Rural Profile, http://archive.stats.govt.nz/browse_for_stats/Maps_and_geography/Geographic-areas/urban-rural-profile.aspx#gsc.tab=0.
[4] UN Statistical Commission (2020), “A recommendation on the method to delineate cities, urban and rural areas for international statistical comparisons”, https://unstats.un.org/unsd/statcom/51st-session/documents/BG-Item3j-Recommendation-E.pdf.
[48] Whitacre, B., D. Meadowcroft and R. Gallardo (2019), “Firm and regional economic outcomes associated with a new, broad measure of business innovation”, Entrepreneurship & Regional Development, Vol. 31/9-10, pp. 930-952, https://doi.org/10.1080/08985626.2019.1630486.
[49] Wojan, T. and T. Parker (2017), Innovation in the Rural Nonfarm Economy: Its Effect on Job and Earnings Growth, 2010-2014, https://doi.org/10.22004/ag.econ.264596 (accessed on March 2020).
Notes
← 1. See https://unstats.un.org/unsd/statcom/51st-session/documents/BG-Item3j-Recommendation-E.pdf.
← 2. A region that would be classified as “predominantly rural” in the second step is classified as “intermediate” if it has an urban centre of more than 200 000 inhabitants (500 000 for Japan) representing no less than 25% of the regional population. Similarly, a region that would be classified as “intermediate” in the second step is classified as “predominantly urban” if it has an urban centre of more than 500 000 inhabitants (1 million for Japan) representing no less than 25% of the regional population.
← 3. The distance from urban centres is measured by the driving time necessary for a certain share of the regional population to reach an urban centre with at least 50 000 people.
← 4. The OECD Metropolitan Database contains a range of socio-economic indicators at the FUA level and can be accessed at https://measuringurban.oecd.org/.
