2. Trends, opportunities and challenges in rural Korea

Abstract

Chapter 2 provides a diagnosis of the performance of rural regions as compared to OECD trends. It first examines demographic patterns in rural Korea, focusing on population levels, growth rates and elderly dependency ratios, and how these influence overall settlement structures. The chapter then benchmarks the performance of Korean rural regions, examining trends in gross domestic product (GDP) per capita and productivity. This section also examines the main sectors of specialisation in rural Korea. The chapter finally examines several dimensions of well-being against OECD trends. In order to draw international comparisons, this chapter makes use of the OECD regional typology, the OECD regional definition based on access to cities and the OECD functional urban areas definition (FUAs), as these apply a consistent definition across OECD countries.

Defining rural in Korea

Korea’s official definition

The definition of what constitutes rural areas in Korea is not unique or straightforward. Different agencies define rural areas differently. According to the Ministry of Land Infrastructure and Transport (MOLIT), there are no rural areas defined as such, rather they emerge as leftovers of urban areas. Urban areas are areas that require systematic development, maintenance, management and conservation, where population and industry are or are expected to be dense (Article 6 of the National Land Planning Act). Therefore, the areas that are not systematically developed and dense are non-urban areas or, in other words, rural areas.

Rural areas are also defined by policy goals. For instance, the Happy Living Zones policy defines three types of living zones: rural, rural-urban and metropolitan.

The rural living zone is the smallest with a total population of around 100 000 inhabitants.
A rural-urban affiliated living zone is centred around a small- or medium-sized city with a population of between 100 000 and 500 000.
A metropolitan living zone is centred around a large city with a population of more than 500 000.

As of 2015, according to the Happy Living Zones policy, a total of 63 zones have been established so far: 21 rural, 14 rural-urban, 20 metropolitan and 8 pilot zones in the capital area.

From the agricultural policy perspective, the Framework Act on Agriculture, Rural Community and Food Industry defines “rural community” in line with the act’s objectives, including eup and myeon, or other areas designated at the discretion of the minister (MAFRA, 2015[1]).

The OECD regional typology

The OECD regional typology simplifies regional data comparability across OECD countries. It classifies two levels of geographic units within each member country: i) large regions (TL2), which generally represent the first administrative tier of subnational government; and ii) small regions (TL3), which aggregate local administrative units (e.g. communes in France or municipalities in Mexico). TL3 regions are divided into predominantly urban (PU), intermediate (IN) and predominantly rural (PR) based on population density and size. Rural areas are further categorised into different types according to their proximity to urban centres for the purpose of defining specific challenges and opportunities related to their geographic location.

Using the OECD typology defining 3 types of TL3 regions (urban, intermediate and rural), there are 17 such regions in Korea, 9 of which are defined as PU, 3 as IN and 5 as PR. Of the five PR regions, all are further categorised as PR close to a city. There are no rural regions in Korea that are considered remote rural regions under the OECD typology.

The OECD uses the concept of functional urban areas (FUAs) as a complementary territorial definition (see Box 2.1). FUAs define urban areas encompassing daily flows of people for work, leisure and social activities as functional socio-economic units, rather than relying on official administrative boundary definitions. Applying the FUA territorial definition, Korea has 22 FUAs covering 26% of the national territory. Out of the 22 FUAs, 5 are classified as large metropolitan areas (with a population of 1 500 000 or more) and 6 as metropolitan areas (with a population of 500 000 to 1 500 000), 8 as medium-sized urban areas (population between 250 000 to 500 000 people) and 3 as small urban areas (population between 50 000 and 250 000 people).

Box 2.1. The EU-OECD definition of an FUA

The EU-OECD definition of FUAs consists of highly densely populated urban cores and adjacent municipalities (“commuting zones”) with high levels of commuting (travel-to-work flows) towards the cores. This definition overcomes previous limitations for international comparability linked to administrative boundaries. This methodology is a clear example of how geographic/morphological information from geographic sources and census data can be used together to get a better understanding of how urbanisation develops.

As the first step, the distribution of the population at a fine level of spatial disaggregation of 1 square kilometre (1 km²) is used to identify the urban cores, which are constituted by aggregations of contiguous municipalities that have more than 50% of their population living in high-density clusters. These clusters are made of contiguous 1 km² grid cells with a population density of at least 1 500 inhabitants per km² and a total population of at least 50 000 people.

As the second step, 2 urban cores are considered part of the same (polycentric) FUA if more than 15% of the population of any of the cores commute to work to the other core.

The third step defines commuting zones using the information on travel-to-work commuting flows from surrounding municipalities to the urban core. Municipalities sending 15% of their resident employed population or more to the urban core are included in the commuting zones, which thus can be defined as the “worker catchment area” of the urban labour market, outside the densely inhabited urban core.

The methodology makes it possible to compare FUAs of similar size across countries, proposing four types of FUAs according to population size:

Small urban areas, with a population of between 50 000 and 250 000 inhabitants.
Medium-sized urban areas, with a population of between 250 000 and 500 000.
Metropolitan areas, with a population of between 500 000 and 1.5 million.
Large metropolitan areas, with a population of 1.5 million or more.

The definition is currently applied to 34 OECD countries (of the 37 OECD members, data are not available for Israel, New Zealand and Turkey) and identifies 1 199 FUAs of different sizes. Among them, 351 FUAs have a population larger than 500 000 and 668 FUAs have a population larger than 250 000.

Source: OECD (2018[2]), OECD Regions and Cities at a Glance, https://dx.doi.org/10.1787/reg_cit_glance-2018-en.

A regional typology based on access to cities

The two territorial definitions – the OECD TL3 Regional Typology and the complementary FUA territorial definition – lead to different analytic frameworks. The TL3 regions cover the entire territory within countries, while FUAs only capture a subsample of the territory. Furthermore, the OECD typology may lead to a certain dichotomy between urban and rural areas.

Against this backdrop, the OECD has recently developed an alternative definition introducing some spatial continuity between metropolitan and non-metropolitan areas. This definition of FUAs classifies 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. Then, these cells are connected with surrounding lower density cells when the flows of commuting between the two types of cells exceed a given threshold (i.e. at least 15% of the labour force commutes to the urban core).

The alternative TL3 classification is based on the presence of FUAs within TL3 borders and the proximity of regions to FUAs of different sizes. 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 more than 250 000 people). It also includes 3 types of non-metropolitan 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-sized city or near one (i.e. regions with an FUA of more than 250 000 people or with access to one within a 60-minute drive),and remote regions (see Box 2.2 for details).

Throughout the document, reference will be made to “rural regions” when referring to the group of non-metropolitan regions, to a “large city” when referring to a city with more than 250 000 inhabitants and a “very large” city when referring to a city with more than 1 million inhabitants. Also, the terms “city” and FUA will be used interchangeably. This alternative regional classification, based on access to cities 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.

Box 2.2. A typology of small regions based on access to cities

The first tier adopts as a threshold 50% population of the TL3 (small) region living in an FUA of at least 250 000 people; the second tier uses as threshold 60 minutes’ driving time, 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 (MRs) are further classified into:
- Large TL3 MRs, if more than 50% of its population lives in an FUA of at least 1.5 million inhabitants.
- TL3 MRs, if the TL3 region is not a large MR and 50% of its population live in an FUA of at least 250 000 inhabitants.
Non-metropolitan TL3 region, if less than 50% of its population live in an FUA. Non-metropolian regions (NMRs) are further classified according to their level of access to FUAs of different sizes:
- NMR-M: With access to a TL3 MR, 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.
- NMR-S: With access to a small/medium-sized 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-sized 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-sized city.
- NMR-R: 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[3]), “Classifying small (TL3) regions based on metropolitan population, low density and remoteness”, https://doi.org/10.1787/b902cc00-en.

These definitions, initially elaborated for international comparability, also represent important tools for policymaking purposes. The access to cities regional definition is relevant for rural policies, since it differentiates amongst different types of rural regions – those close to cities and those that are remote. Rural areas 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 areas with access to large cities vis-à-vis small/medium-sized ones allowing to better understand and capture differences in the linkages. In contrast, rural remote regions may require much-differentiated policy responses that address their particularities. Thus, spatial scales are critical tools for the design of regional policies.

For the case of Korea according to the access to cities definition, seven regions are classified large MRs, six are classified as MRs, three (Chungcheongnam-do, Gangwon-do and Sejong Special Self-Governing City), as regions close to large cities and just one (Jeollanam-do) as a region close the small- and medium-sized cities.

Table 2.1. Korean regions using the regional typology based on access to cities
TL3 region	Access to cities definition
KR011: Seoul	Large metropolitan region
KR012: Incheon	Large metropolitan region
KR013: Gyeonggi-do	Large metropolitan region
KR021: Busan	Large metropolitan region
KR022: Ulsan	Metropolitan region
KR023: Gyeongsangnam-do	Metropolitan region
KR031: Daegu	Large metropolitan region
KR032: Gyeongsangbuk-do	Metropolitan region
KR041: Gwangju	Large metropolitan region
KR042: Jeollabuk-do	Metropolitan region
KR043: Jeollanam-do	Near small- and medium-sized cities
KR051: Daejeon	Large metropolitan region
KR052: Chungcheongbuk-do	Metropolitan region
KR053: Chungcheongnam-do	Near a large city
KR054: Sejong	Near a large city
KR061: Gangwon-do	Near a large city
KR071: Jeju-do	Metropolitan region
Source: Fadic, M. et al. (2019[3]), “Classifying small (TL3) regions based on metropolitan population, low density and remoteness”, https://doi.org/10.1787/b902cc00-en.

Demographic patterns of rural Korea and land distribution

Korea has experienced rapid population changes since the period of industrialisation, which started in the 1960s. The share of the rural population has decreased gradually based on the national definition, from representing 60% of the total population (or 18.2 out of 30.8 million) in the 1970s to only 19% of total population (9.7 out of 51.6 million) in 2018. These trends have been driven by migration patterns from rural to urban due to the rapid industrialisation experience by the country.

Korea has a low share of its national population in rural regions

The OECD has developed regional typologies to allow for international comparisons and to be able to compare regions with similar characteristics as explained in the previous section. Based on the OECD TL3 classification, in 2018, 17.1% of the national population live in rural regions (Table 2.2), namely rural regions close to cities. The population of the PU regions accounted for 69.7% of the total population, while 13.2% lived in IN areas. According to this definition, Korea has the fourth-largest urban population amongst OECD countries after the United Kingdom, the Netherlands and Australia. According to the OECD access to cities definition, rural regions are home to 11.3% of the national population and, according to the FUA definition, the share of the national population living outside of FUAs amount to 17%. In other words, rural regions in Korea are less than 20% but more than 10% according to these indicators.

Table 2.2. Share of rural population according to different definitions, percentage, 2018
	OECD TL3 typology	Non-FUA	Access to cities OECD TL3 typology	World Bank
Share of rural population	17.1	17	11.3	18.5
Note: Table refers to year 2015 for FUA.
Source: (OECD, n.d.[4]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en; World Bank (2019[5]), Rural Population, (% of Population) - Korea (dataset), https://data.worldbank.org/indicator/SP.RUR.TOTL.ZS?locations=KP-KR.

As described in the previous section, different definitions capture different elements of rurality. The OECD regional typology uses density criteria and proximity to urban areas to measure rurality. The FUA captures cities and their broader areas of influence but this definition does not provide any measure or criterion outside FUAs and does not cover the entire country. The access to cities OECD definition overcomes the limitations of both these definitions by providing a territorial definition covering the entire country and using an FUA criterion.

Figure 2.1. Distribution of population by metro and non-metro TL3 regions, 2018

According to the access to cities OECD definition, 11.3% of the national population lived in rural regions in 2018, amongst which 7.8% lived in regions near a large city (Figure 2.2) and 3.4% in regions near a small/medium-sized city (no region of Korea is defined as a remote rural region using this definition). The share of Korea’s national population living in rural regions (11.3%) is significantly lower than the OECD average (29%). Furthermore, those Koreans who do live in rural regions have a strong interaction with cities. Understanding and making the most of urban-rural linkages is consequently of paramount importance in the design of rural policies for Korea.

Figure 2.2. Population share by TL3 regions, access to cities OECD definition, 2018

In term of the OECD FUA definition, 83% of the national population lives in cities of more than 50 000 inhabitants and 76% were living in cities with more than 500 000 inhabitants, significantly higher than the OECD average share of 55%. When compared to OECD countries (Figure 2.3), only Luxembourg has a higher national share living in FUAs, making Korea one of the countries with the highest share of the national population living in FUAs.

Figure 2.3. Population share living in functional urban areas over total population, 2015

These figures show a high concentration in settlement patterns (Figure 2.4) in Korean regions when compared to OECD standards, both among TL3 regions and in FUAs.

Figure 2.4. Index of geographic concentration of population, TL3, 2001 and 2019

Settlement growth and implications for concentration

The high levels of concertation in Korea have driven national policy responses over the past decades with measures put in place to promote more balanced national development and in particular to reduce the pressures that concentration has placed on the capital region of Seoul. This section measures the growth dynamics in Korea’s settlement patterns over the past years and their implications for the country’s high level of concentration.

In terms of population growth, according to the OECD TL3 regional typology, PU regions in Korea experienced the largest increase in their population share over the period 2001-17, increasing by 2.3 percentage points over this period (Figure 2.5). In contrast, the population share declined in IN and PR regions by 0.6 and 1.6 percentage points respectively. The population share in PR regions decreased in all except five OECD countries (the exceptions are: Belgium, Chile, Mexico, the Slovak Republic and the United States).

Figure 2.5. Change in the share of population by type of region (TL3), 2000-17

In terms of the OECD access to cities typology, only Greece increased its national share of the rural population over the period 2001-19. In the rest of the countries, the population share of MRs to total population increased. Korea also experienced an increase in the total population living in MRs (0.76 percentage points) but this increase was the sixth-lowest when compared to OECD countries (Figure 2.6).

Despite the overall decline in the population share in rural regions, the decline was mainly driven by Jeollanam-do, the only Korean TL3 region classified with access to small- and medium-sized cities, which shrank annually by an average 0.7%. In those rural regions classified as close to large cities, the population increased annually by close to a half of a percent (0.48%) annually. This increase was the 12^th largest when compared to the trends in OECD countries. These population dynamics across different types of regions in Korea are in line with OECD trends (Figure 2.7). Rural regions with access to small- and medium-sized cites are facing stronger demographic pressures than regions close the large cities. During 2008-18, 29% of OECD countries with regions with access to a small/medium-sized city (9 out of 31) experienced population decline. This percentage is smaller (20% or 5 out of 25) in regions with access to a large city.

Amongst the non-metropolitan (rural) regions classified in the access to cities definition, two of them, Chungcheongnam-do and Sejong, increased their population share from 2000-18 while Jeollanam-do saw its share decline (Figure 2.8). In absolute terms, Sejong increased its population by 219 161 inhabitants from 2012 to 2018, more than doubling its initial population. Chungcheongnam-do increased its population by 284 160 during 2001-18 representing a 13% increase. Jeollanam-do, in contrast, lost 12% of its 2001 population amounting to 216 500 inhabitants and Gangwon-do gained 7 750 inhabitants in absolute terms over the period, though saw its share of the national population decline slightly.

Figure 2.6. The share of the population in metropolitan regions increased in the last two decades, 2001-19

Figure 2.7. Population growth in TL3 non-metropolitan regions, 2000-18

Figure 2.8. Growth in population share by four non-metropolitan (rural) TL3 regions in Korea, 2001-18

Amongst FUAs, those with more than 500 000 inhabitants but less than 1.5 million experienced the fastest population growth with an approximately 18% overall increase over 2000-18. Large metropolitan and medium-sized urban areas grew by 9% and 7% overall. In turn, small urban area populations decreased by 11% in total.

Within FUAs, the commuting zones of large metropolitan areas grew by 48%, the highest growth over the period of 2000-15. Similarly, the commuting zones of medium-sized urban areas grew by 23%. In contrast, the core of the large metropolitan and medium-sized urban areas did not grow by as much, by 8% and 7% respectively, while the growth of core population in metropolitan areas was higher at 17%.

At the regional scale, the two TL3 regions from the capital TL2 region which concentrate the highest population shares, Gyeonggi-do and Seoul, experienced different population dynamics over 2001-18. Seoul in 2001 was home to 21.3% of the national population followed by Gyeonggi-do 20%. Over this time, the population share in Seoul decreased down to 18.8% in 2018, in contrast to Gyeonggi-do which experienced an increase to 25% (Figure 2.9).

Figure 2.9. Population shares in Seoul and Gyeonggi-do, 2001-18

The population share amongst region types according to the access to cities definition also show different trends. The share of large MRs increased by 1 percentage point (from 65.7% to 66.8%), while the share of MRs decreased by 0.9 percentage points (from 22.7% to 21.9%). Amongst rural regions, those with access to large cities increased by 0.6 percentage points and regions close to small- and medium-sized regions decreased by 0.8 percentage points (Figure 2.10).

Figure 2.10. Population shares Korean TL3 regions based on OECD access to cities definition

According to the geographic concentration index, Korea’s high level of demographic concentration (Figure 2.11) first increased from 2001 to 2011 and then this trend reverted and started to decline from 2011 to 2018.

Figure 2.11. Geographic concentration index among Korean TL3 regions, 2001-18

The young population migrates from rural to urban regions

Data suggests that PU regions (OECD TL3 regional typology) are net recipients of migrants from other types of regions within Korea. Based on the access to cities typology, not all MRs are net recipients of migrants. In fact, the only type of region which saw a positive average of migration flows was that with access to a large city (Table 2.3). In 2017, regions with access to a large city recorded a positive average net migration rate, while other region types saw a negative average rate. The comparison of net migration rates of total population versus young people reveals that: i) large MRs attract young people; ii) migration into regions with access to large cities 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) young people disproportionally leave regions with access to a small/medium-sized city compared to other age groups.

Table 2.3. Net migration rates, young and total population, by type of TL3 region (average), 2017
		Net migration rate (%)	Net migration rate (%), population 15-29 years of age
OECD TL3 regional typology	Predominantly urban	0.31	0.56
OECD TL3 regional typology	Predominantly rural	-0.07	-1.24
OECD access to cities typology	Large metropolitan regions	-0.14	1.72
	Metropolitan regions	-0.04	-4.92
	Regions with access to a large city	1.42	0.95
	Regions with access to a small/medium-sized city	-0.17	-8.15
Note: Net (young) migration rate is defined as the median value of inflows minus outflows of (young) people over the total population. Inflows are defined as the group of new residents in the region coming from another region of the same country; outflows are defined as the group of persons who left the region to reside in another region of the same country.
Source: (OECD, n.d.[4]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.

Korea’s rural population is fast-ageing

The composition of the population differs between cities and rural regions. While the share of the younger population in urban and rural regions are at the same level, the share of the elderly population is larger in rural regions (low density) than in MRs (Figure 2.12). The share of the elderly population in rural regions stood at 18.3% which was 4.5 percentage points larger than the share in MRs. On the contrary, the proportion of the working-age population in rural regions is 68.6%, 4.7 percentage points smaller than the proportion of the working-age population of the total population in MRs.

Korea’s population age structure follows the recent trends in OECD countries. Ageing is a stronger structural phenomenon in rural regions vis-à-vis MRs and therefore, rural and lower density regions face stronger ageing pressures than MRs (OECD, 2020[7]). Elderly dependency rates in rural regions stood at 27% in 2018. The ratio was slightly lower than the OECD average in TL3 rural regions, on average the difference was 2.7 percentage points (Figure 2.13). Korea is amongst the countries with the largest gap in elderly dependency ratios in 2018 between rural and metropolitan regions together with Australia, Canada, Finland, Japan, Sweden and the United Kingdom, all with a gap above 8 percentage points.

In 2018, regions with access to a small/medium-sized city had the highest average elderly dependency ratios (33%), followed by regions with access to an MR (24%). This differs from the trend typically seen in the OECD context, since in 2018, in 13 out of 18 OECD countries, the elderly dependency ratio was higher in regions with access to an MR than in regions with access to a small/medium-sized city (Figure 2.14).

Figure 2.12. Population pyramid by age, gender and place of residency, Korea, 2017

Figure 2.13. Elderly dependency ratio gap between rural and metropolitan regions, 2018

Over the last 18 years, the population living in rural regions has become increasingly older. The growth rate of the elderly population, 65 years of age and older, has been the largest in rural regions when compared to MRs, recording an annual average growth rate of 4.2% during the period of 2000-18. Within rural regions, the elderly population grew faster in the region with access to a small/medium-sized city compared to regions with access to an MR, where the per annual growth rate was actually negative.

Figure 2.14. Elderly dependency ratio by country and type of non-metropolitan region, 2018

In terms of access to cities, rural Korea has an advantage compared to other OECD TL3 rural regions. Compared to those, rural regions in Korea have the fourth shortest travelling time to the closest city (Figure 2.15). On average, the median travel time to the closest city is about 25 minutes. Within rural regions in Korea, the median travel time varies from 16 (Jeju-do) to 37 minutes (Gangwon-do).

Figure 2.15. Median travel time to the closest city

In sum settlement patterns in Korea tend to concentrate in large cities when compared to OECD countries. The share of the national population living in rural regions in Korea ranges from 11.3% to 18% depending on the definition employed (e.g. OECD regional typology or OECD access to cities typology). Despite the low share, rural regions have strong linkages with cities, particularly with those above 250 000 inhabitants. This implies that rural policy responses will need to take advantage of strong interlinkages. Settlement patterns are concentrated in Korea when compared to OECD standards but the rate of concentration has declined in recent years and started to revert towards a more balanced development pattern. Population in Korea is growing faster in MRs, particularly in MRs against large MRs. In rural regions, population growth is higher in regions close to large cities and negative in rural close to small- and medium-sized cities. Elderly dependency ratios in rural Korea are slightly lower when compared to OECD rural regions but, when compared to urban, they are significantly higher. In fact, Korea shows the seventh-highest gap in dependency ratios between rural and urban amongst OECD countries. Elderly dependency ratios are increasing at a fast pace in rural regions in comparison to MRs, particularly in regions close to small- and medium-sized cities.

Rural areas of Korea are performing well economically

Korea’s economic expansion is considered a success story of catching-up economies across OECD member countries. In 2003, Korea’s GDP per capita was 28 percentage points below the OECD average. In just a decade, Korea has been able to reduce the gap by 8 full percentage points. The industrialisation of Korea’s economy over the last 60 years which shifted its specialisation from agricultural to industry and now services has been largely responsible for the country convergence. The annual GDP per capita growth rate has been 2.6 times higher in Korea than in OECD member countries on average, growing annually at a rate of 3.07 in GDP per capita during 2003-16. This catching-up process, however, has not occurred in all Korean regions. In fact, regional disparities are very present in the geographical economic landscape of Korea today.

Economic development tends to progress unevenly among regions within OECD countries due to the benefits associated with economies of agglomeration. “Economies of agglomeration” is the term used to describe how firms like to locate close to other firms and to densely populated areas due to lower transportation costs, proximity to markets and a wider availability of labour supply. People also tend to be attracted to densely populated areas for the wider availability of job opportunities, goods and services. These mutually reinforcing forces yield important economic premium for both consumers and firms: economies of scale, better matching and functioning of labour markets, spill-over effects and more technological intensity. It is to no surprise that productivity, and therefore wages, tend to be higher in densely populated areas. These benefits, however, must be weighed against the costs of densely populated areas such as congestion, negative social effects of a possible oversupply of labour, higher land prices, rising inequality and environmental pressures. The net impact varies from one urban area to another.

Similar to settlement patterns, economic activities in Korea are also fairly concentrated. PU regions attract the largest share of economic activities in Korea with approximately 45% of the national GDP being produced in just 2 large MRs (the capital city Seoul and Gyeonggi, the region that surrounds the capital) in 2017. By looking at the international benchmark, Korea is among the top ten OECD countries with the highest index of geographic concentration of GDP among TL3 regions (Figure 2.16) though notably, Korea is the only OECD country where GDP is less concentrated than the population.

Figure 2.16. Geographic concentration index of GDP (TL3 regions), 2017

A closer look at the distribution of GDP by type of region reveals that the GDP share of PR regions stood at 19.5% of Korea’s GDP in 2018 and this corresponded to about KRW 353 trillion. The contribution of the rural economy to the national GDP has been constant during the past 2 decades, including during the global financial crisis, standing at around 20% of the national GDP. The rural economy achieved its highest share of 20.9% in 2011. The GDP of Korea’s rural regions grew overall by 95.3% during 17 years. Rural regions have been able to sustain their share of national GDP against PU and IN regions. Based on the distribution of GDP by type of regions, PU regions were hit hardest by the 2008 financial crisis.

Figure 2.17. GDP contribution by type of TL3 Korean region, 2018

According to the access to cities typology, rural (non-metropolitan) regions account for 13.3% of Korea’s GDP in 2018, or KRW 241 trillion. The share peaked in 2011, amounting to 13.6%. Since this peak, the share of rural regions has stabilised to 13.2% of Korea’s GDP. During the period 2001 to 2018, the GDP of rural regions grew overall by 108% which was 1.3 times more than the growth of GDP in Korea’s metropolitan areas.

Figure 2.18. GDP contribution by type of TL3 Korean region using OECD access to cities typology, 2018

According to the OECD typology, the average level of GDP per capita in rural regions was USD 45 762 in 2018, which was USD 5 220 higher than the national average of USD 40 542 and USD 17 600 higher than the OECD PR region average of USD 28 162 (2017). Similarly, according to the access to cities OECD definition, GDP per capita in Korea’s non-metropolitan (rural) regions in 2018 was USD 48 279. The high level of GDP per capita in rural regions is largely driven by the higher levels of GDP per capita in regions close the large cities against the lower levels in regions closer to small/medium-sized cities.

Korea’s rural economies are performing well compared to OECD countries in terms of per capita GDP growth. Before the 2008 global financial crisis, the annual growth rate of GDP per capita in rural Korea was 1.5 times higher than the average of the PR OECD TL3 region. Once the crisis erupted in 2008, Korea’s GDP per capita remained above the average of OECD rural economies. As a result, the PR economies of Korea have been able to achieve high levels of per annum growth in GDP per capita growth with high initial levels of GDP per capita (on average).

Figure 2.19. Performance of TL3 rural regions in Korea and other OECD TL3 rural regions, 2000-17

When comparing the performance by type of region, Korea is the only OECD country where GDP per capita in rural regions is higher than the GDP per capita in urban regions (Figure 2.20). The GDP per capita ratio to the national average in Korea’s PR regions was 113% in 2017, which was 19 percentage points higher than the GDP per capita ratio to the national average in PU regions in Korea. This can be largely due to the high concentration of settlement patterns in MRs and large cities in Korea.

Figure 2.20. GDP per capita gap between PU and rural regions, 2017

Within the national context, Korea’s rural regions are also amongst the top performers. The majority of Korea’s PR regions record higher GDP per capita growth than the national average during 2000-17 (Figure 2.21). The highest growth rate occurred in Chungcheongnam-do, where the GDP per capita grew by 5.2%. In addition to this, more than half of the rural areas had higher initial levels of GDP per capita in 2000 than the national average.

In terms of labour productivity, Korea’s rural regions have lower levels of productivity compared to OECD rural regions but productivity in rural Korea has been converging. Korea’s rural economies have recorded higher per annum growth rates than on average OECD rural regions. During the period 2008-17, labour productivity grew by 2% per annum in rural Korea surpassing the average OECD TL3 rate of 0.7% per annum. Korea’s rural (non-metropolitan) regions also have on average higher productivity growth rates than other types of regions. The gains in productivity were greatest in regions with access to large cities (2.4%) while regions with access to a small/medium-sized city grew at a rate of 1.3%. Across OECD countries, rural regions in Austria, Belgium, Germany, the Netherlands and Switzerland show similar trends.

As is the case with GDP per capita, Korea is the only country with no gaps in labour productivity among rural and urban regions (Figure 2.22). Rural regions display higher levels of labour productivity against urban regions. In the rest of the countries, labour productivity tends to be higher in urban regions.

The lower levels of labour productivity also occur at the national scale. Previous OECD studies have attributed the relative lower productivity levels amongst other items resulting from the working culture. Long working hours have resulted in lower levels of well-being and female labour participation (OECD, 2019[9])).

Figure 2.21. GDP per capita as a share of the national average by type of region, Korea, 2000 vs. 2017

Figure 2.22. Labour productivity gab between urban and rural regions, 2016

The strong performance of Korea’s rural regions may result from the effects of geography and their close proximity to cities. Since Korea is an extremely urbanised country, rural regions have strong interactions with cities. OECD research suggests surrounding regions can also “borrow” agglomeration from neighbouring cities. For a doubling of the population living – at a given distance – in urban areas within a 300 km radius, the productivity of the city in the centre increases by between 1% and 1.5% (OECD, 2014[10]). Given that residents and businesses of low-density regions have a strong interaction with cities, they can also borrow agglomeration benefits from the surrounding areas.

Rural regions specialised in tradeable sectors

The economy in rural regions, in general, is less diversified than in urban regions. In Korea, rural regions are highly specialised in tradeable sectors. The share of tradeable goods and services to the region’s GVA represents 58% in 2017, 8 percentage points higher than the national average of 50%. Manufacturing alone contributed to over two-fifths of the rural regions’ GVA and was also higher (42%) than the national average (32%).

Rural regions in Korea show differences in their economic structure. These can be divided into two categories: rural regions specialised in tradeable sectors and rural regions more specialised in non-tradeable sectors.

The economy of three out of five rural regions (Chungcheongnam-do, Gyeonsangbuk-do and Jeollanam-do) is specialised in tradeable goods and services sectors since these sectors explain more than half of the GVA of the regions.
The remaining two rural regions, Gangwon-do and Jeju-do, are more specialised in non-tradeable activities.

The tradeable sector is a key driver of competitiveness, given that it competes in global markets. The GVA of the tradeable sector in Korea represents 50% of the GVA in 2017 (Table 2.4) and the share was higher (58%) in rural regions, especially manufacturing (42%). Urban regions tend to specialise more in non-tradeable activities. In Korea, they contributed to 55% of total GVA.

Table 2.4. Distribution of GVA by economic sectors, 2017
Sector	Rural regions (%)	Urban regions (%)	Korea (%)
Manufacturing (T)	42	27	32
Public admin., compulsory social security, education, human health (NT)	18	15	16
Distributive trade, repairs, transport, accommodation, food service activities (NT)	9	18	15
Construction (NT)	6	5	5
Agriculture, forestry and fishing (T)	6	0	2
Real estate activities (NT)	4	9	7
Financial and insurance activities (T)	3	8	7
Non-manufacturing industry, including energy (T)	3	1	2
GVA in other services (T)	3	3	3
Professional, scientific, technical activities, admininistration, support service activities (NT)	3	9	7
Information and communication (T)	1	6	4
Tradeable	58	45	50
Non-tradeable	42	55	50
Source: OECD (n.d.[4]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.

In terms of growth rates, the tradeable sector in Korea grew by 22% overall during 2000-17 in rural regions in contrast to the non-tradeable sector, which shrank by 14%. Manufacturing was the largest driver of GVA growth for rural regions.

Despite the significance of the tradeable sector for rural Korea, the non-tradeable sectors are also important drivers for the economy and jobs. In Korea, they account for 65% of all jobs and, in rural areas, 59% of all rural jobs in 2019. In Korea there are a number of interesting trends:

About one-fourth of the total employment (24%) was in the distribution, trade, repairs, transport, accommodation and food service activities sector in PR regions in 2019.
Since the crisis period (2007-09), the agriculture, forestry and fishing sectors is losing its relative weight. The share of total employment in agriculture, forestry and fishing sectors fell from 27% to 16% in 2008-19.
The public administration sectors grew from 14% to 21% in 2008-19.
Other sectors have not gone through as dramatic changes as the agriculture and public administration sectors in terms of employment.

Figure 2.23. Change in sector share of GVA, 2000-17

These trends show the structural transformation in rural Korea. Although the tradeable sector, particularly manufacturing, remains the largest contributor to the rural GVA (42%), it only employs 15% of the rural workforce. The relative weight of traditional rural sectors of agriculture, forestry and fishing are also losing their relative weight against non-tradeable sectors.

Table 2.5. Distribution of employment by economic sectors, 2019
Sector	Rural (%)	Intermediate (%)	Urban (%)	Korea (%)
Manufacturing (T)	15	20	16	16
Agriculture, forestry and fishing (T)	16	12	1	5
Construction (NT)	7	6	8	7
Distributive trade, repairs, transport, accommodation, food service activities (NT)	24	24	29	27
Financial and insurance activities (T)	2	2	3	3
Non-manufacturing industry, including energy (T)	1	1	1	1
Information and communication (T)	1	1	4	3
Other services (T)	6	6	7	7
Professional, scientific, technical activities, admininistration, support service activities (NT)	6	6	10	9
Public admininistration, compulsory social security, education, human health (NT)	21	20	19	19
Real estate activities (NT)	1	2	2	2
Tradeable	41	42	32	35
Non-tradeable	59	58	68	65
Note: NT=Non-tradeable sector; T=Tradeable sector.
Source: (OECD, n.d.[4]), OECD Regional Statistics (database), https://doi.org/10.1787/region-data-en.

Well-being of rural dwellers

The well-being agenda has come to the forefront of OECD countries in recent years

For many years, GDP was the metric by which countries measured economic performance. However, GDP fails to measure important quality of life elements, including leisure time, health, social connections or environmental quality. At the same time, GDP does not account for inequality or how growth is affecting the resources available for future well-being. In light of growing inequalities and negative externalities stemming from increased production in certain sectors, policymakers can no longer look to GDP to provide an accurate assessment of progress. Since the financial crisis, policy leaders have acknowledged a need for a framework that recognises broader measures of social progress alongside more traditional “production-oriented” measurements (Stiglitz, Sen and Fitoussi, 2009[11]). Today, governments are paying greater attention to dimensions of well-being, such as housing, education, access to water and civic engagement (Cornia et al., 2017[12]). The concept of well-being recognises that economic progress works within these dimensions, encompassing a broader view of social progress beyond production and market value.

Studies reveal individuals who have made significant income gains often report their economic situation to be worse than much poorer rural individuals who have not achieved any income gains (Graham, 2018[13]). To avoid this paradox, any effort to understand welfare gains must consider non-income dimensions. The OECD began publishing comparative measures of well-being in 2011 in response to the growing recognition that income is not the only factor affecting how individuals experience economic growth. Individual countries have also established their own frameworks and indicators to reflect on well-being, such as those listed in the table below. New Zealand has taken this a step further by seeking ways to improve quality of life for citizens through its Wellbeing Budget (New Zealand Treasury, 2019[14]). The budget prioritises mental health, child well-being and Indigenous aspirations alongside more traditional economic growth goals.

The OECD framework provides a lens through which to consider current and future well-being through measures of quality of life, material conditions and sustainability. The first two measures provide a comparison of current well-being between regions. Quality of life considers the role of health, education, environmental quality and air pollution among other factors. The framework uses primarily objective indicators, such as voter turnout to measure civic engagement, while also including an indicator of subjective well-being through life satisfaction surveys (OECD, 2011[15]). Material conditions include measures of income and wealth, jobs and earnings and housing. These measures rely on indicators such as disposable income, net wealth and long-term unemployment rate. Finally, future well-being represents the stock of natural, economic, human and social capital available to provide lasting well-being to future generations.

Table 2.6. Selected national well-being measurement initiatives and indicator sets
Country	Measurement initiative	Leading agency	Description
Austria	How’s Austria	Statistics Austria	Since 2002, Statistics Austria reports on 30 indicators focused on 3 dimensions: material wealth, quality of life and environmental sustainability.
Israel	Well-being, Sustainability and National Resilience Indicators	Central Bureau of Statistics	Since 2015, the government publishes a set of indicators focused on the following domains: quality of employment, personal security, health, housing and infrastructure, education, higher education and skills, personal and social well-being, environment, civic engagement and governance, and material standard of living.
Slovenia	Indicators of Well-Being in Slovenia	Institute of Macroeconomic Analysis and Development	Since 2015, a consortium of four institutions updates indicators on a yearly basis. These indicators are presented in three categories: material, social and environmental well-being.
Wales	Well-being of Wales	Welsh Governments Chief Statistician	Since 2015, the Well-being of Future Generations (Wales) Act is aimed at incorporating social, economic, environmental and cultural well-being into policymaking. The act recognises 7 well-being goals and 46 indicators.
Source: Exton, C. and M. Shinwell (2018[16]), “Policy use of well-being metrics: Describing countries’ experiences”, https://doi.org/10.1787/d98eb8ed-en (accessed on 22 July 2019).

Figure 2.24. OECD framework for measuring well-being and progress

Measuring well-being at the regional level

In order to improve the use of well-being measures for policymaking at the regional and local levels, the OECD launched the publication How’s Life in Your Region (OECD, 2014[18]) that developed a common framework for measuring people’s well-being at the regional level. The framework has been designed to improve policy coherence and effectiveness by looking at nine dimensions that shape people’s material conditions (income, jobs and housing) and their quality of life (health, education, environment, safety, access to services and civic engagement). These nine dimensions derive from both characteristics of individuals and those of each specific territory. They are best gauged through indicators of real outcomes rather than inputs or outputs.

The regional well-being framework captures a number of factors that are important to the competitiveness of places and helps to reinforce the importance of complementarities between different sectoral policies. The OECD framework for measuring regional and local well-being has the following seven distinctive features:

It measures well-being where people experience it. It focuses both on individuals and on place-based characteristics, as the interaction between the two shapes people’s overall well-being.
It concentrates on well-being outcomes that provide direct information on people’s lives rather than on inputs or outputs.
It is multi-dimensional and includes both material and non-material dimensions.
It assesses well-being outcomes not only through averages but also by how they are distributed across regions and groups of people.
It is influenced by citizenship, governance and institutions.
It takes account of complementarities and trade-offs among the different well-being dimensions.
It looks at the dynamics of well-being over time, at its sustainability and the resilience of different regions.

To make the OECD Regional Well-Being Framework operational, indicators of well-being that are comparable across countries were developed for the OECD’s 402 large regions and to a much lesser extent for the 275 metropolitan areas (FUAs) across the 9 dimensions of well-being.

These indicators, comparable across OECD countries, come from official sources in most cases and are available over different years. They are publicly available in the OECD Regional Well-Being Database. At present, regional measures are available for OECD countries in 11 well-being topics: income, jobs, housing, education, health, environment, safety, civic engagement and governance, access to services, community, and life satisfaction, as specified in Table 2.7.

Table 2.7. Well-being topics selected for visualisation
	Topics	Indicators
Material conditions	Income	Household disposable income per capita (in real USD PPP)
	Jobs	Employment rate (%) Unemployment rate (%)
	Housing	Number of rooms per person (ratio)
Quality of life	Health	Life expectancy at birth (years) Age adjusted mortality rate (per 1 000 people)
	Education	Share of labour force with at least secondary education (%)
	Environment	Estimated average exposure to air pollution in PM2.5 (μg/m³), based on satellite imagery data
	Safety	Homicide rate (per 100 000 people)
	Civic engagement	Voter turnout (%)
	Accessibility of services	Share of households with broadband access (%)
Subjective well-being	Community	Percentage of people who have friends or relatives to rely on in case of need
Subjective well-being	Life satisfaction	Average self-evaluation of life satisfaction on a scale from 0 to 10
Source: OECD Regional Well-being, https://www.oecdregionalwellbeing.org (accessed on 3 December 2020).

Regional measures, comparable across countries, are not currently available on work-life balance. For each topic, one or two indicators have been selected. Improvements in the way the OECD measures the well-being topics in regions are still underway. For example, additional measures of access to services or indicators that measure other environmental performance are being developed particularly at lower-level geographies and for rural regions.

How’s life in your regions in Korea?

Comparable subnational data are only available across large TL2 regions. For the case of Korea, there are 7 TL2 regions. Figure 2.25 compares Korean TL2 regions across the distribution of all 440 OECD TL2 regions amongst the bottom 20%, middle 60% and top 20%.

Figure 2.25. Distribution of Korean TL2 regions compared to all OECD TL2 regions, 2020

The main trends amongst Korean TL2 regions are:

The well-being domain with the largest disparities between Korean regions is related to job outcomes (employment and unemployment rate), with Gyeongnam region in the bottom 40% of OECD regions and Jeju close to the top 20%.
All Korean regions rank in the top 20% of the OECD regions in access to broadband.
In contrast, all Korean regions except Jeju rank in the bottom 20% in perceived social support network (community).
The top performing Korean regions is above the average of the top OECD regions in 3 out of 13 well-being indicators, namely access to services, adjusted mortality rates and life expectancy.

Figure 2.26. Korean TL2 regions compared to national and median OECD TL2 regions, 2020

Since data for regional well-being are not available for TL3 regions, which would provide a tool to benchmark well-being in rural Korean regions against national and international standards, the analysis focuses on regional well-being of those TL2 regions with a high degree of rurality. The degree of rurality captures the percentage of the population living in rural communities. Rural communities are defined as municipalities with less than 150 inhabitants per km². The rural TL2 regions sample consist of 64 TL2 regions. The degree of rurality ranges from 67.9% to 100%. Korea has only 1 TL2 region (Gangwon-do) with a relatively high degree of rurality (87.5%).

Relative to other TL2 regions with a high degree of rurality, Gangwon-do’s performance across the different regional well-being dimensions is mixed:

The region’s comparative advantages are in the jobs, health, education, civic engagement, safety and access to services dimensions.
- Gangwon-do is a top performer in safety and access to services compared to other selected OECD TL2 regions (Figure 2.27).
- Approximately 97.2% of the region’s households had a broadband connection.
- Moreover, the region had about 4.5 percentage points lower unemployment rate than the average of selected OECD TL2 regions.
In contrast, the region recorded lower levels of well-being in the environment, housing, income and life satisfaction than with comparable TL2 regions.

Figure 2.27. The OECD Regional Well-being in selected TL2 regions with a high degree of rurality, 2014

The good performance in the jobs dimension can be linked to higher rates of education. The importance of human capital and skills as drivers of regional growth is challenged by the increasing outmigration of the young skilled population to urban areas. In 2014, the educational attainment rate of the region was 73%, approximately 10 percentage points smaller than the national average. However, the educational attainment rate of the region is about 7 percentage points higher than the average rural TL2 region’s rate.

Income was one of the dimensions where the region had lower levels than the average of the selected TL2 regions with a high degree of rurality (around USD 14 150 compared to an average of USD 19 300). Another area of comparative weakness is environmental quality: the level of air pollution was approximately 21.7 in µg/m³ in 2014, positioning Gangwon-do in the bottom rankings in terms of environmental quality performance across OECD TL2 regions.

References

[12] Cornia, G. et al. (2017), “Economic inequality and social progress”, in Socio-Economic Transformations, Cambridge University Press, https://doi.org/10.1017/9781108399623.004 (accessed on 29 May 2019).

[16] Exton, C. and M. Shinwell (2018), “Policy use of well-being metrics: Describing countries’ experiences”, OECD Statistics Working Papers, No. 2018/07, OECD Publishing, Paris, https://doi.org/10.1787/d98eb8ed-en (accessed on 22 July 2019).

[3] 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://doi.org/10.1787/b902cc00-en.

[13] Graham, C. (2018), “Subjective well-being in economics”, in The Oxford Handbook of Well-Being and Public Policy, https://doi.org/10.1093/oxfordhb/9780199325818.013.14.

[1] MAFRA (2015), Framework Act on Agriculture, Rural Community and Food Industry, Article 2 - Basic Philosophy.

[14] New Zealand Treasury (2019), The Wellbeing Budget, https://treasury.govt.nz/sites/default/files/2019-06/b19-wellbeing-budget.pdf (accessed on 3 July 2019).

[19] OECD (2020), Regions and Cities at a Glance 2020 – Korea, https://www.oecd.org/cfe/Korea-Regions-and-Cities-2020.pdf.

[7] OECD (2020), Rural Well-being: Geography of Opportunities, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/d25cef80-en.

[21] OECD (2019), OECD Regions and Cities at a Glance 2018 – Korea, OECD, Paris, https://www.oecd.org/cfe/KOREA-Regions-and-Cities-2018.pdf.

[9] OECD (2019), Women at work in G20 countries, https://www.oecd.org/g20/summits/osaka/G20-Women-at-Work.pdf.

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

[17] OECD (2017), How’s Life? 2017: Measuring Well-being, OECD Publishing, Paris, https://doi.org/10.1787/how_life-2017-en.

[18] OECD (2014), How’s Life in Your Region?: Measuring Regional and Local Well-being for Policy Making, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264217416-en.

[10] OECD (2014), The Metropolitan Century: Understanding urbanisation and its consequences.

[6] OECD (2012), Redefining “Urban”: A New Way to Measure Metropolitan Areas, OECD Publishing, Paris, https://doi.org/10.1787/9789264174108-en.

[15] OECD (2011), Compendium of OECD Well-being Indicators, OECD Better Life Initiative, OECD, Paris, http://www.oecd.org/sdd/47917288.pdf (accessed on 29 May 2019).

[4] OECD (n.d.), OECD Regional Statistics (database), OECD, Paris, https://doi.org/10.1787/region-data-en.

[20] OECD (n.d.), OECD Regional Well-being, OECD, Paris, https://www.oecdregionalwellbeing.org (accessed on 3 December 2020).

[11] Stiglitz, J., A. Sen and J. Fitoussi (2009), Report by the Commission on the Measurement of Economic Performance and Social Progress, https://ec.europa.eu/eurostat/documents/118025/118123/Fitoussi+Commission+report (accessed on 29 May 2019).

[8] Weiss (2018), Mapping inequality in accessibility: a global assessment of travel time to cities in 2015, https://ora.ox.ac.uk/objects/uuid:5c2ebe20-a1c2-44cb-bd71-95042fa49fbe/download_file?file_format=pdf&safe_filename=weiss_et_al_2017_accepted.pdf&type_of_work=Journal+article.

[5] World Bank (2019), Rural Population, (% of Population) - Korea (dataset), https://data.worldbank.org/indicator/SP.RUR.TOTL.ZS?locations=KP-KR (accessed on 13 December 2018).

╳

Metadata, Legal and Rights

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

https://doi.org/10.1787/a3c685a7-en

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