To Have and Have Not – How to Bridge the Gap in Opportunities

Report

22 September 2025

3. Geographic inequalities in access to opportunities

Copy link to 3. Geographic inequalities in access to opportunities

Abstract

This chapter examines how the place where people are born and grow up shapes their opportunities in life. Opportunities to access quality education, jobs, and services vary significantly across and within countries. These differences matter, because most people stay close to their birthplace well into adulthood. Residents of metropolitan and higher-income regions tend to have better access to services, infrastructure, and employment, which leads to better educational and labour market outcomes. In contrast, people in poorer or remote regions face persistent disadvantages. The chapter documents these disparities drawing on regional and urban typologies based on OECD territorial definitions. It shows how place shapes opportunity and contributes to long-term inequalities in life outcomes.

3.1. Introduction

Copy link to 3.1. Introduction

Addressing inequalities in opportunities, understood as disparities in people’s life chances that arise from circumstances beyond their control rather than individual effort (see Chapter 1), remains a significant challenge across OECD countries. As shown in Chapter 2, on average across OECD countries, over a quarter at least of observed inequalities in household market incomes can be attributed to circumstances beyond people’s control, such as their sex and country of birth or the country of birth and socioeconomic status of their parents. The reason is that these circumstances can reduce access to opportunities and drive discriminatory behaviour, impacting lifetime educational attainment and labour market trajectories (Adema, Fluchtmann and Patrini, 2023[1]; Akee, Jones and Porter, 2019[2]; O’Connell, 2019[3]) and thereby determining incomes and living standards.

This chapter provides evidence on another crucial factor that affects people’s opportunities in life: the place where they are born and live. Geographic location matters for inequality of opportunities because people have no control over where they are born and raised. As this chapter shows, opportunities in access to education, services and jobs can vary significantly within countries, and relocating later in life to areas where opportunities – such as well-paid jobs – are greater, requires overcoming significant barriers. These include the financial costs associated with switching homes, particularly when moving into higher-income areas (Ganong and Shoag, 2017[4]; Ferreira, Gyourko and Tracy, 2010[5]), job search costs (Gobillon and Selod, 2021[6]), a desire to remain close to established social networks (Spring, Gillespie and Mulder, 2023[7]), and responsibilities for caregiving (Artamonova and Syse, 2021[8]). As a result, many people stay close to their birthplace during their adult life. For instance, over half of Swedish people still live in their municipality of birth by age 30 (Thomassen, Lundholm and Malmberg, 2023[9]), and half of the UK population have never worked outside the local authority in which they were born (Bosquet and Overman, 2019[10]). Moreover, geographic mobility has declined in several large OECD countries over the last decades, including in Australia, Canada, Japan, Korea, Poland, and the United States (Alvarez, Bernard and Lieske, 2021[11]).

The characteristics of the places where people grow up significantly influence their later-life outcomes, notably educational attainment and lifetime earnings, as documented in recent studies for Australia (Deutscher, 2020[12]), the United Kingdom (McNeil, Luca and Lee, 2023[13]) and the United States (Chetty and Hendren, 2018[14]). Public perception aligns with these findings: two-thirds of respondents in the latest wave of the OECD’s Risks that Matter survey, which assesses people’s perceptions of the social and economic risks they face, report that the neighbourhood where people grow up impacts on their ability to get ahead in life (OECD, 2023[15]).

This chapter documents that, across the OECD, people in the same country face unequal access to education, employment, essential services and infrastructure, depending on where they live. This, in turn, contributes to persistent geographic disparities in economic opportunities and living standards. Metropolitan and higher-income regions provide greater physical accessibility to essential services, such as childcare facilities and schools (Almeida et al., 2024[16]), as well as better digital and transport infrastructure (OECD, 2023[17]). Although they tend to be more expensive, cities also tend to offer greater opportunities for employment (Ormerod, 2013[18]) and earnings progression (Roca and Puga, 2016[19]). A handful of large cities, often capital cities, concentrate most of the innovation (Paunov et al., 2019[20]; Cantwell and Zaman, 2024[21]). Substantial differences in opportunities also exist at finer geographic scales. Within cities, poorer neighbourhoods offer fewer and lower-quality services, such as schools (Owens and Candipan, 2019[22]) and public transport (Nie et al., 2024[23]). Such regional and local disparities in opportunities matter for people’s outcomes: the inhabitants of metropolitan and higher-income regions, and those living in higher-income neighbourhoods, benefit from higher upward mobility, both across generations (Chetty et al., 2014[24]) and throughout their adult lives (Aghion et al., 2023[25]; Roca and Puga, 2016[19]).

As significant differences in opportunities also exist within regions, the scale at which inequality of opportunity is measured matters. Smaller geographical units make spatial differences more visible, while larger units obscure these differences by averaging out smaller-scale variation. Inequalities in access may be felt most strongly in the places where people spend most of their daily lives (e.g., the neighbourhoods or functional urban areas they live in). Still, measuring inequalities at larger scales, such as at regional level, matters – e.g., because it can inform the redistribution of funds within countries and the design of place-based policies.

Because internationally comparable data on measures of opportunity at smaller scales are often unavailable, this chapter mainly provides evidence on disparities in people’s opportunities across regions and functional urban areas. It classifies "places" using OECD territorial definitions, including small regions (Territorial Level 3, TL3), distinguished by their access to cities (Fadic et al., 2019[26]) and the degree of urbanisation (OECD et al., 2021[27]), as well as large regions (Territorial Level 2, TL2). For further information on different geographical units and typologies used in this chapter, see Annex 3.A.

The remainder of this chapter is organised as follows. Section 3.2 discusses the uneven geography of income poverty and financial fragility risk. Section 3.3 explains how location matters for access to, and quality of, educational opportunities, and Section 3.4 provides evidence on geographic inequalities in labour market opportunities. Section 3.5 documents geographic disparities in access to healthcare and other essential services.

3.2. Place matters for income poverty and financial fragility

Copy link to 3.2. Place matters for income poverty and financial fragility

Where a person grows up or lives plays a critical role in shaping their life chances. Higher-income people are more likely to live in regions with better schools, healthcare and job opportunities. This spatial concentration of advantage reinforces existing socio-economic inequalities (van Ham, Manley and Tammaru, 2024[28]). In contrast, people living in poorer regions often face weaker public services and limited access to quality jobs. This reduces their prospects and can trap individuals in cycles of disadvantage (OECD, 2018[29]; Banzhaf and Walsh, 2008[30]).

3.2.1. People’s risk of facing income poverty varies greatly across regions

Poverty rates vary widely across regions within countries, pointing to persistent regional inequalities in income and opportunity in some countries. In 2022, 15% of people in OECD regions lived in relative income poverty, i.e., in a household with an income below 50% of the national median after adjusting for household size. However, this average conceals large territorial differences.

Some countries show deep regional divides in poverty, while others are more equal. For example, Mexico, the country where differences are widest, shows more than a tenfold difference in poverty rates across regions – a gap of around 35 percentage points – with a regional poverty rate in Baja California Norte at around 30% of the median region against 355% of the median region for Chiapas (Figure 3.1). Similarly, in Colombia and Italy, the gap between the regions with the highest and lowest poverty rates exceeds 30 percentage points, which corresponds to a factor of six to ten. In contrast, differences are smaller in countries like Hungary and Sweden, where the gap between the regions with the highest and lowest poverty rate is less than 25 percentage points.1

People’s chances to move up from the bottom of the income distribution are also influenced by where they live. Evidence from tax records in Belgium, Estonia and Spain suggests that people in higher-income regions experience greater upward mobility over five years than those in lower-income regions (Box 3.1)

Figure 3.1. Regional poverty rates vary significantly in some countries
Copy link to Figure 3.1. Regional poverty rates vary significantly in some countries

Relative income poverty rates in TL2 regions (median region = 100), in 2022 or latest available year

Box 3.1. Geographic inequalities in short-term income mobility: Evidence from Belgium, Estonia and Spain using tax-record data

Copy link to Box 3.1. Geographic inequalities in short-term income mobility: Evidence from Belgium, Estonia and Spain using tax-record data

People’s risk of living in poverty varies substantially depending on where they live (see Figure 3.1). Beyond comparing regional income levels, it is much harder to show, for lack of data, that there are also regional differences in the extent to which low incomes are persistent over time, i.e., in income mobility over the life course. This box presents evidence on short-term income mobility in Belgium, Estonia and Spain. The analysis draws on tax record data from an ongoing project that uses administrative microdata to study income dynamics (Königs and Terrero-Dávila, 2025[31]). Specifically, this box examines disparities in upward mobility across regions for working-age people in the lower part of the income distribution,¹ looking at (i) the average change in people’s position in the income distribution over five years; and (ii) the share of people who move up the income distribution over that period.

People in lower-income regions find it harder to move up the income distribution

In Belgium, Estonia and Spain, people face unequal prospects for climbing the income ladder depending on where they live. For those starting around the 15th percentile of the national income distribution, mobility over five years varies substantially by region. This holds both for people’s average advancement along the distribution (Figure 3.2, Panel A), as well as for the share of people who experience upward mobility (Figure 3.2, Panel B).²In Spain, for example, working-age people who start off around the 15th percentile of the income distribution move up, on average, by 4 percentiles in regions where upward mobility is weakest, compared to more than 10 percentiles in regions where mobility is stronger. Likewise, in both Belgium and Spain, rates of upward mobility differ widely: in some regions, less than half of working-age people starting around the 15th percentile move up over the period, while in others more than two-thirds do. In Estonia, where relative income mobility is higher and where there are fewer regions, regional disparities are somewhat narrower.

Figure 3.2. People in lower-income regions exhibit less favourable mobility outcomes
Copy link to Figure 3.2. People in lower-income regions exhibit less favourable mobility outcomes

Regional income levels help explain part of these disparities. Across the three countries, people living in higher-income regions generally face better prospects for upward mobility than those in lower-income regions. While some of these regional differences may reflect the sorting of individuals with different characteristics (see the discussion about urban-rural divides in test scores in Section 3.3), evidence from Belgium suggests that regional income levels may play a role even after accounting for such differences. Figure 3.3 shows the gap in mobility outcomes for people living in regions with different income levels, measured relative to residents of middle-income regions, both before and after accounting for individual characteristics. Among working-age people starting around the 15th percentile with the same age, gender, household type, employment status and education, those living in lower-income regions advance 3 percentiles less (Figure 3.3, Panel A), and are 6 percentage points less likely to experience upward mobility (Figure 3.3, Panel B). As discussed in the remainder of this chapter, higher-income regions provide better infrastructure, as well as more employment and education opportunities, which are key to fostering upward mobility.

Figure 3.3. Regional disparities in mobility outcomes persist even after accounting for individual characteristics
Copy link to Figure 3.3. Regional disparities in mobility outcomes persist even after accounting for individual characteristics

Gaps in five-year mobility outcomes for people in lower- and higher-income regions relative to those in middle-income regions around the 15th income percentile, before and after accounting for individual characteristics, in Belgium, TL3 level

Figure 3.4. Child poverty rates and the gender gap in poverty both vary widely within countries
Copy link to Figure 3.4. Child poverty rates and the gender gap in poverty both vary widely within countries

Relative child poverty rate (Panel A) and difference between adult female and male relative poverty rates (Panel B), in TL2 regions, in 2022 or latest available year

Children’s exposure to poverty varies also depending on where they grow up. In OECD regions with available data, the average child poverty rate is 17%, but differences within countries are wide. Mexico and Lithuania show the largest regional gaps, each exceeding 30 percentage points, with poverty rates for children in the poorest regions more than eleven times higher than in the least poor regions (Figure 3.4, Panel A). Child poverty rates tend to be lower in capital-city regions. In 15 out of 23 countries, the rate in the capital-city region is below the median region. Across all 23 countries, capital-city regions report an average child poverty rate of around 11%, which is about 3 percentage points lower than the national average.

Gender inequalities in poverty rates are also evident across regions, with women being, on average, more likely to live on very low incomes than men.2 Out of 323 regions with available data, 267 show higher poverty rates for women than for men. In countries such as Belgium, Chile, Czechia, Estonia, Finland and Portugal, every region reports a positive gender gap. In contrast, countries like Italy, Lithuania and Colombia display large within-country variation. For example in Italy, in the Basilicata region, the poverty rate for women exceeds that of men by 10 percentage points, while in Molise, it is 8 points lower (Figure 3.4, Panel B). Gender gaps in poverty rates tend to be lower in capital-city regions. Across the 23 countries with available data, capital-city regions report an average gender poverty gap of 1.3 percentage points, a little more than half the national average of 2 percentage points.

Gender disparities in poverty may translate into unequal access to economic opportunities, for instance if living in poverty creates obstacles to obtaining education and training, or if it is compounded by poorer access to essential services, such as childcare. For women in capital-city regions, gender gaps in poverty rates are lower, possibly due to selection effects, better job opportunities, greater access to services and stronger social support systems. However, higher living costs in capital-city regions could offset some of these potential benefits.

3.2.2. Poverty rates vary substantially across cities and rural areas

Aggregations by large regions may mask important differences in poverty between cities and rural areas. Although cities, especially large ones, typically provide better employment options, education and services, these advantages usually come at a price of higher living costs, especially housing, and more competition for access to services. Meanwhile, rural areas face distinct challenges, including longer travel times to services and higher per capita costs of provision (OECD, 2021[32]).

Poverty gaps between cities and rural areas vary significantly across countries. In Austria, Belgium and Denmark, the share of people living in relative poverty is higher in cities than in towns, semi-dense, or rural areas (Figure 3.5). For example, poverty rate in Austrian cities is 172% of the national median, while it is 79% of the median in rural areas. By contrast, in Hungary, Lithuania and Poland, poverty is more prevalent in rural areas. In Hungary, the difference is particularly large: the poverty rate in rural areas is 160% of the national median, while in cities it is 70%.

Figure 3.5. Poverty rates are higher in cities
Copy link to Figure 3.5. Poverty rates are higher in cities

Relative poverty rate by degree of urbanisation (median region = 100), in 2023 or latest available year

Poverty rates also differ substantially across city centres – the densely populated cores of Functional Urban Areas (FUAs). Across the five countries with available data, differences between city centres can be large (Figure 3.6). In Spain, for example, the poverty rate in the city centre of Torrevieja is 234% of the median city centre, the largest gap among countries with available data. In contrast, disparities are smaller in Norway, where the city centre of Asker, the most deprived, has a poverty rate just 139% of the median. These patterns point to considerable disparities in poverty within cities, underscoring the importance of place-based strategies in addressing urban deprivation.

Figure 3.6. Poverty rates in city centres of Functional Urban Areas differ substantially
Copy link to Figure 3.6. Poverty rates in city centres of Functional Urban Areas differ substantially

Relative poverty rate by city centre (median city centre = 100), in 2022 or latest available year

3.2.3. Households in capital-city regions are less likely to face financial fragility

Financial fragility, defined as living without sufficient financial assets equivalent to three months of income at the national poverty line, significantly undermines people’s opportunities by reducing their capacity to handle economic shocks, invest in education, and secure better housing or business opportunities. Regions with higher levels of financial fragility often show lower intergenerational mobility, reinforcing inequality over time (OECD, 2021[33]; Chetty et al., 2014[24]). Additionally, limited financial assets restrict access to credit, which can play an essential role for upward mobility and entrepreneurship (Balestra and Oehler, 2023[34]; Mian, Sufi and Verner, 2017[35]). This financial fragility perpetuates intergenerational inequality, as families without resources are less able to invest in their children's education and well-being, widening socio-economic gaps (OECD, 2023[36]; Fagereng et al., 2020[37]).

Like poverty, the share of individuals considered financially fragile also differs significantly across regions within countries (Figure 3.7). It is lower in capital-city regions than in the rest of the country, by 7 percentage points on average across countries with available data (36% vs 43%). Italy, the country with the widest differences, shows a nearly fivefold difference in the share of individuals considered financially fragile across regions, with Abruzzo at around 43% of the median region while Campania is at 260% of the median region. In Austria, Canada, Germany and Japan, certain regions also face substantially higher levels of financial fragility than the rest of the country.3

Figure 3.7. Cross-regional variation in financial fragility is substantial in some OECD countries
Copy link to Figure 3.7. Cross-regional variation in financial fragility is substantial in some OECD countries

Relative share of people in financial fragility by TL2 and TL3 regions (median region=100), 2022 or latest available year

3.3. Geographic inequalities in access to and quality of educational opportunities

Copy link to 3.3. Geographic inequalities in access to and quality of educational opportunities

Education and training are essential for helping people from lower socio-economic backgrounds improve their economic standing by increasing their earnings potential and building resilience to financial challenges. As highlighted in Chapter 2, parental socio-economic background is an important driver of inequality of opportunity. Children whose parents have lower levels of education often face disadvantages because their families have fewer resources to support learning. Highly educated parents tend to place greater resources to do so. As discussed in Chapter 4, early childhood education can help bridge these gaps and help equalise opportunities later in life (Heckman, 2006[38]).

However, individuals face unequal access to good-quality education not just in those crucial early stages, but at all stages in life, depending on where they are born and live. Unequal access to quality education can contribute to perpetuate disparities in skills and socio-economic outcomes, as families will usually be bound to education options in their surrounding areas (OECD, 2021[32]). This has greater implications in cities where socio-economic segregation is higher, and in rural areas where access may be difficult.

Beyond the family background, where children grow up also contribute to shape their attitudes towards education, their social networks and access to opportunities (van Ham, Manley and Tammaru, 2024[28]). In disadvantaged regions, schools often have fewer resources, less experienced staff and weaker support systems. These challenges affect all students but hit low-income families hardest.

Moreover, children in these schools may have limited exposure to diverse role models or career paths, which can lower aspirations and reduce the likelihood of completing upper secondary or tertiary education (OECD, 2024[39]). In contrast, children in wealthier regions tend to attend better-resourced schools and benefit from additional support outside the classroom. These early differences contribute to unequal life chances.

3.3.1. Lower test scores in rural areas largely reflect family background

School quality is crucial for ensuring equal opportunities for all students. Differences in resources, teacher qualifications and facilities can greatly impact students' learning and career prospects. Not all public schools get appropriate funding, modern facilities and experienced teachers (OECD, 2017[40]). Schools in poorer and remote areas often struggle with limited resources, outdated infrastructure, high teacher turnover and difficulty attracting good teachers. These disparities can affect academic achievement and perpetuate socio-economic inequalities by limiting students' opportunities for personal and professional growth. While it is not possible to measure differences in school quality within cities, this section reviews available evidence on quality differences across schools depending on the size of the settlement where they are located (see Annex 3.A for further details on the geographical units).

OECD Programme for International Student Assessment (PISA) results show wide geographic differences in student performance, partly linked to where families live. In mathematics, for instance, students in settlements with fewer than 3 000 residents (rural settlements) tend to score lower than those in settlements with more than 100 000 residents (urban areas) in 28 out of 31 OECD countries with available data (Figure 3.8, Panel A).4 However, when accounting for parental socio-economic background, a statistically significant disadvantage for rural students remains in only 10 countries; the rural-urban gap disappears in 10 countries, and it even reverses in 8 (Figure 3.8, Panel B).

While these results highlight the important role of the socio-economic composition of parents in explaining differences in learning outcomes of students across places, some regional differences in student performance persist even after accounting for parental background. This suggests that other place-based factors, such as school quality, learning environments, infrastructure, or teacher experience, also play a role. These factors are often linked to remoteness and may reflect deeper structural inequalities that are not fully captured by individual characteristics. Moreover, people do not randomly “sort” into places: families may choose where to live based on school quality or other unobserved advantages, making it difficult to fully isolate the effect of location.

Figure 3.8. Rural-urban gaps in test scores are partly explained by differences in socio-economic status
Copy link to Figure 3.8. Rural-urban gaps in test scores are partly explained by differences in socio-economic status

3.3.2. Promoting equal educational opportunities for rural students through better accessibility and resources

Students in rural areas often encounter distinct challenges, such as longer travel times to school. These distances could negatively impact attendance, academic performance and progression to higher education, especially where public transport is limited (OECD, 2021[32]). Rural-urban inequalities in accessibility are especially pronounced for primary school students, such that access to a motor vehicle or school bus services are often essential to transport kids from and to school (Almeida et al., 2024[16]). Even for families with access to a motor vehicle, long distances to school impose additional costs, which weigh on the budgets of low-income households. Transport barriers can also affect educational outcomes beyond test scores. Long commutes increase the risk of absenteeism and early school leaving, especially where public transport does not offer a viable alternative. The extra time and effort required for travel may lead to fatigue and reduced engagement with school (OECD, 2022[41]).

Still, in OECD countries with available data, there is no evidence that students in rural settlements do worse than those in cities. After accounting for parental socio-economic characteristics, students do not perform systematically worse in countries where rural schools are on average less accessible relative to urban schools (Figure 3.9). In some countries, including Estonia, New Zealand and Portugal, students in rural settlements outperform their urban peers in mathematics. This is true even though, in these countries, the share of people with access to a school within a 15-minute drive is much lower in rural settlements than in cities (by at least 10 percentage points). Conversely, in Australia, Czechia, Ireland, Latvia and Norway, rural students tend to score lower and face accessibility gaps compared to students in cities.

Figure 3.9. Geographic disparities in school accessibility can be large, but this is not always reflected in test scores
Copy link to Figure 3.9. Geographic disparities in school accessibility can be large, but this is not always reflected in test scores

The rural-urban gap in PISA assessment in mathematics after accounting for socio-economic conditions and share of population with access to a school within a 15-minute drive in rural settlements relative to urban settlements, in 2022 or latest available year

Redistributive policies play a key role in addressing the challenges faced by rural schools. Targeted funding helps improve infrastructure and attract qualified teachers. Complementary policies, such as incentives for rural teaching placements, investments in digital learning and stronger local transport networks, are also essential. These measures help ensure that students in rural areas have access to a similar quality of education as those in urban settings (OECD, 2022[42]).

Rural schools often face more infrastructure challenges than urban ones, though patterns vary across countries. In 21 of 33 OECD countries with available data, a higher share of rural school principals report that poor-quality infrastructure is a barrier to effective instruction (Figure 3.10). This is the case for countries including Colombia, Mexico and the United Kingdom. In others – such as Estonia, Latvia and Switzerland – urban schools are more likely to report infrastructure issues. These differences suggest that the quality of school infrastructure reflects not only the type of settlement, but also national policies and investment choices.

Figure 3.10. Rural schools are more likely to have inadequate or poor-quality physical infrastructure
Copy link to Figure 3.10. Rural schools are more likely to have inadequate or poor-quality physical infrastructure

Share of school principals expressing concerns about the quality of physical infrastructure by school location, in 2022 or latest available year

Box 3.2. Role of neighbourhoods for receiving social assistance benefits: Evidence from the Netherlands

Low-income people, including recipients of means-tested social assistance benefits, often cluster in certain neighbourhoods, for instance because of the availability of affordable (social) housing. This spatial concentration can make existing inequalities (e.g., labour market outcomes, education) stronger, so that living in a neighbourhood with a higher local concentration of social assistance benefit recipients may result in a higher likelihood of relying on social assistance benefits oneself. This may occur through multiple channels, including peer effects that lower the risk of stigmatisation; reduced informational costs of accessing benefits; weaker employment networks; labour market discrimination; and lower-quality local services.

Measuring the role of high benefit receipt in neighbourhoods on their residents is challenging precisely due to selection effects. Since people do not randomly choose where they live, those living in poorer neighbourhoods may also face disadvantages related to their education, employment or health. As a result, any association observed in the data between residents’ outcomes and the type of neighbourhood they live in may partly reflect who chooses to live where, rather than the causal impact of place on people’s life outcomes. In other words: there is likely a problem of endogeneity, as individual and household characteristics that influence people’s incomes also drive residential location choices. Addressing this challenge, also known as spatial sorting, requires methods that separate neighbourhood influences from individual traits.

Recent OECD analysis asks whether neighbourhood-level social assistance receipt is associated with an individual’s likelihood of relying on means-tested benefits (Moreno Monroy et al., 2025, forthcoming[43]). The analysis draws on rich longitudinal administrative records covering all working-age individuals in the Netherlands between 1999 and 2019 (over 7 million people), with precise geographic information. This data makes it possible to account for unobserved, time-invariant characteristics such as preferences, ability, or long-term disadvantage through individual fixed effects, as well as both time-invariant and available time-varying neighbourhood characteristics. The analysis examines whether changes in the neighbourhood share of benefit recipients are linked to changes in an individual’s likelihood of receiving benefits, while also controlling for other observable time-varying neighbourhood effects. It further tests whether these associations differ by labour market size or by type of area, such as cities compared to rural regions. However, the analysis cannot control for individual-level shocks (e.g., sudden job loss or divorce) that may both increase the likelihood of benefit receipt and trigger moves to poorer neighbourhoods. For this reason, the results should not be interpreted as causal.

The results indicate that living in a neighbourhood where social assistance benefit receipt is widespread is associated with a higher likelihood of an individual relying on social assistance. This effect is significant even after accounting for individual and household characteristics, and spatial sorting. On average, individuals living in a neighbourhood with twice the rate of social assistance receipt have a benefit receipt rate that is 2.6 percentage points larger (Figure 3.11, Bar 1). After accounting for individual characteristics and residential sorting, this difference falls to 1.2 percentage points, a 13% rise relative to the 9% average rate of social assistance receipt in the sample (Figure 3.11, Bar 2).

Densely populated cities, particularly those within the four largest labour-market areas, show the strongest association between neighbourhood- and individual-level benefit receipt, suggesting that population density plays a part. The neighbourhood’s role is most pronounced in cities and weaker in towns and rural areas (Figure 3.11, Bars 3-5). Similarly, the effect is stronger in large urban labour markets – particularly in cities such as Amsterdam, Rotterdam, The Hague and Utrecht – than in smaller ones (Bars 6-7).

These findings suggest policies should address both individual needs and place-based disadvantages. Policies that strengthen neighbourhood conditions while supporting vulnerable individuals are likely to be most effective in supporting individuals at the bottom of the income ladder.

Figure 3.11. Neighbourhood share of benefits recipients matter for the individual probability of receiving benefits in the Netherlands (1999-2019)
Copy link to Figure 3.11. Neighbourhood share of benefits recipients matter for the individual probability of receiving benefits in the Netherlands (1999-2019)

Marginal effects from a linear probability model

3.4. Geographic inequalities in labour market opportunities

Copy link to 3.4. Geographic inequalities in labour market opportunities

The employment opportunities available for young people when leaving education have a strong impact on their later careers. Empirical evidence suggests that early‑career joblessness can have long-term scarring effects, including a higher probability of later unemployment (Schmillen and Umkehrer, 2017[44]; Brandt and Hank, 2014[45]) and lower future earnings (De Fraja, Lemos and Rockey, 2021[46]). Throughout adulthood, factors such as the local availability and quality of jobs, training opportunities and access to employment support can all have an impact on participation and earnings, as well as on overall well‑being. However, these opportunities differ widely across places.

3.4.1. Young people’s career opportunities depend on where they grow up

Young people face unequal prospects for a successful school-to-work transition depending on where they live. Across countries where data are available, the share of young people aged 18 to 24 not in employment, education, or training (NEET) differs by an average of 12.7 percentage points between the best- and worst‑performing regions. The gap is substantially wider in some Southern European countries and in Mexico, exceeding 20 percentage points (Figure 3.12). In several regions in these countries, one in four young people or more are NEET. By contrast, Nordic countries have low NEET rates and limited regional disparities.

These differences in school-to-work transitions mirror geographic inequalities in educational outcomes: young people are more likely to be NEET in regions with a higher share of early school leavers, i.e., those aged 18 to 24 who have completed no more than lower secondary education (Figure 3.12). Although there may be some overlap, early school leavers and NEET young people are not always the same group: early school leavers may find work even without having obtained a formal qualification, while many NEETs do have upper-secondary education or more (Carcillo et al., 2015[47]). In the quartile of regions with the highest shares of early school leavers, NEET rates are, on average, 7 percentage points higher than in the quartile of regions with the lowest shares. The gap is most pronounced in countries where school-to-work transitions are particularly challenging, such as Italy and Türkiye, but can also be substantial in countries with low national NEET rates, such as Australia and Hungary. The results suggest that certain regions, often those with lower GDP per capita, struggle both to retain students in school and to provide adequate employment opportunities for young people.

Figure 3.12. Geographic inequalities in educational outcomes carry over into school-to-work transitions
Copy link to Figure 3.12. Geographic inequalities in educational outcomes carry over into school-to-work transitions

Early school leavers and young people not in employment, education or training (NEET), TL2 regions, 2023 or latest available year

In countries with lower GDP per capita, young women face greater regional inequalities in their school-to-work transition, and their labour market outcomes are more strongly linked with geographic inequalities in educational outcomes than those of young men. Specifically, in countries such as Costa Rica, Mexico and Türkiye, young women are significantly less likely to transition successfully from school to work in regions where a greater share of them did not complete upper secondary education (Figure 3.13, Panel B). For young men, this association is much less pronounced (Figure 3.13, Panel A). Although gender disparities also exist in some larger economies, such as the United States, their magnitude is substantially smaller. Cross-country studies suggest that, even in OECD countries, lower GDP per capita is associated with an earlier age at marriage for women, as well as with higher fertility rates (Jelnov, 2021[48]; Campisi et al., 2020[49]; Abeynayake, Bomhoff and Lee, 2012[50]). Both factors negatively affect the educational and labour market outcomes of women (Villalobos-Hernández et al., 2015[51]; İlkkaracan, 2012[52]).

Figure 3.13. Gender disparities in school-to-work transition can be large
Copy link to Figure 3.13. Gender disparities in school-to-work transition can be large

Early school leavers and young people not in employment, education or training (NEET) by gender, TL2 regions, 2023 or latest available year

3.4.2. People living in lower-GDP-per-capita regions have fewer employment opportunities

Geographic disparities in young people’s employment outcomes continue throughout adulthood, with people in regions with lower GDP per capita having systematically fewer employment opportunities. Regions that have had persistently lower GDP per capita over the last 20 years exhibit systematically lower employment rates (Figure 3.14, Panel A). One reason for these employment disparities is that firms tend to establish themselves in higher-income, higher-productivity regions more rapidly than workers can relocate to those same areas (Bilal, 2023[53]; Lindenlaub, Oh and Peters, 2022[54]). Higher-income regions also offer a deeper pool of skilled workers, reducing costs for firms in finding adequately skilled candidates, which can boost job creation (Di Cataldo and Rodríguez-Pose, 2017[55]). Differences in employment rates between lower- and higher-GDP-per-capita regions within countries are often larger than differences across countries for regions with a similar GDP per capita. The gap can exceed 10 percentage points in some countries, particularly those where national employment rates are relatively low, such as France, Italy, Spain and Türkiye. The gap tends to be smaller in countries with high employment rates, such as Australia, the Netherlands and Sweden. People in lower-GDP per capita regions also have less access to jobs in high-value-added sectors. In most OECD countries, such employment opportunities are disproportionately located in a few high-income regions, often those hosting capital cities (Figure 3.14, Panel B).

Figure 3.14. Regions with lower-GDP-per-capita also have lower employment rates and host fewer high‑value-added employment opportunities
Copy link to Figure 3.14. Regions with lower-GDP-per-capita also have lower employment rates and host fewer high‑value-added employment opportunities

3.4.3. People living in lower-GDP-per-capita regions also have less access to employment services and training

Public employment services (PES) play a key role in connecting workers with employment opportunities by providing job search assistance, individualised employment support and potentially training (Dromundo, Lüske and Tuccio, 2023[56]). The geographic inequalities in employment opportunities presented in Figure 3.14 coincide with, and may be reinforced by, disparities in access to such employment support and training. Results from a recent project, which collected and exploited geolocation data on PES centres, show that the accessibility of PES centres – measured in terms of travel times – is lower for people in regions with lower GDP per capita. This holds true even after accounting for factors such as population density, the degree of access to cities, and regional unemployment (Box 3.3).

Adults in regions with lower GDP per capita are also significantly less likely to participate in training, which limits their chances to acquire new skills, progress in their careers and build resilience against economic shocks. Historically, regions in the bottom 20% of GDP per capita over the last 20 years exhibit systematically lower rates of adult participation in formal and/or non-formal training and education (Figure 3.15). Across countries where data are available, adults in regions with persistently higher GDP per capita are on average almost 5 percentage points more likely to have participated in training and education in the past four weeks than those with persistently lower GDP per capita. This regional gap is independent of overall training opportunities in the country: the largest disparities are observed in Austria, Czechia, Denmark and France – countries with very different overall rates of adult participation in training and education. The regions with the highest adult training rates are typically those with the greater share of jobs in high value-added sectors – i.e., large metropolitan regions, and often those hosting capital cities. These tend to be also regions with a higher proportion of skilled workers, which highlights the challenge of providing training and education opportunities to lower-skilled workers, for whom the benefits of training may be largest.

Figure 3.15. Adults in regions with lower GDP per capita are less likely to participate in education or training
Copy link to Figure 3.15. Adults in regions with lower GDP per capita are less likely to participate in education or training

Share of adults (25-64) participating in training and/or education in the past four weeks, by GDP per capita over the last 20 years, TL2 regions, 2023

Box 3.3. Geographic accessibility of PES centres: cross-country evidence from geolocation data

Copy link to Box 3.3. Geographic accessibility of PES centres: cross-country evidence from geolocation data

Cross-country evidence on the physical accessibility of public employment services (PES) is limited, as there is no centralised, publicly available repository of the locations of PES centres across countries. A recent OECD project gathered data on the location of these services in over 30 countries and provided a first assessment of their accessibility across different regions (Almeida et al., 2024[57]).

This box summarises some main findings.

Methodology

To assess the regional characteristics associated with PES accessibility, the study uses the following OLS regression model:

$Y_{r c} = α + β_{1} {U n e m p l o y m e n t}_{r c} + {β_{2} G D P p e r c a p i t a}_{r c} + {β_{3} D e m o g r a p h i c s}_{r c} + β_{4} {M e t r o p o l i t a n}_{r c} + β_{5} {P o p u l a t i o n d e n s i t y}_{r c} + φ_{c} + ε_{r c}$

where $Y_{r c}$ captures the share of the population in region r, country c, who can access a PES centre within 15 minutes by motor vehicle.¹ ${U n e m p l o y m e n t}_{r c}$ captures the unemployment rate in each region. ${G D P p e r c a p i t a}_{r c}$ is a vector measuring both the level and growth in regional GDP per capita, while ${D e m o g r a p h i c s}_{r c}$ captures both the level and growth in the regional population. ${M e t r o p o l i t a n}_{r c}$ is a categorical variable with five outcomes capturing the degree of access to cities, as defined in Annex 3.A. ${P o p u l a t i o n d e n s i t y}_{r c}$ captures the regional population density. $φ_{c}$ are country fixed effects, which account, for instance, for cross-country differences in GDP per capita and institutional arrangements, including national employment policy. In the estimation, regions are given equal relative weight within countries, such that all countries carry equal weight in the regression.

Results

Three key findings can be derived from the analysis (Table 3.1):

PES are more accessible in regions with higher GDP per capita, even when accounting for other regional characteristics. The magnitude is relevant: a 10% higher GDP per capita is associated with a 2 percentage-point greater share of people who can reach a PES centre within 15 minutes by motor vehicle.
Metropolitan regions offer higher PES accessibility than non-metropolitan regions, even after accounting for demographic and economic characteristics. In large metropolitan regions, the share of people who can reach a PES centre within 15 minutes by motor vehicle is nearly 10 percentage points higher than in non-metropolitan remote regions.
PES accessibility is greater in regions with higher unemployment, after accounting for other regional characteristics. This could be interpreted as tentative evidence that governments may adjust service provision to meet regional demand. This may partly be offset by an effect working in the opposite direction: a PES centre in a region may contribute to a better matching of jobseekers to vacancies, which would lower regional unemployment.

	% of people within 15 minutes to nearest PES by motor vehicle
	(1)	(2)	(3)	(4)
Unemployment rate in 2019 (%)	0.233			0.807***
	(0.307)			(0.253)
Children aged 5 to 9 in 2022 (%)

GDP p.c. in 2019 (Ln)		23.389***		20.531***
		(2.920)		(4.029)
Annual GDP p.c. growth 2005-2019 (%)		-2.951***		-4.028***
		(0.866)		(0.987)
Total population in 2022 (Ln)			8.744***	0.480
			(0.755)	(1.331)
Annual population growth 2015-2022 (%)			2.325	-7.631***
			(1.467)	(2.021)
Population density in 2022 (Ln)				6.615***
				(1.157)
2. Metropolitan – Medium				-2.797
				(3.153)
3. Non-metropolitan – Medium				-4.510
				(3.838)
4. Non-metropolitan – Small				-3.867
				(4.017)
5. Non-metropolitan – Remote				-9.723**
				(4.557)
Country FE	YES	YES	YES	YES
Observations	692	922	1612	661
Number of countries	18	25	32	16
Adjusted R-squared	0.383	0.455	0.546	0.617

Note: Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1. The number of countries included in the regressions varies across models depending on data availability. Regional observations are weighted by the inverse of the number of regions within a country, such that all countries carry equal weight. The island regions Gotland (SWE), Eivissa y Formentera (ESP) and Mayotte (FRA) are not included in the analysis. GDP per capita and unemployment data are measured in 2019 to avoid potential distortions because of the COVID-19 crisis. The reference category for access to cities is large metropolitan regions (see Annex 3.A).

Source: Accessibility data derived from OECD calculations based on location data obtained from national authorities. Data on regional characteristics are retrieved from the OECD Regional Statistics database: https://www.oecd.org/regional/regional-statistics/.

1. 15 minutes is the time threshold that maximises regional variation in the share of people who can access PES centres by motor vehicle.

3.4.4. Geographic inequalities in labour market opportunities persist over time

Lower access to training and employment support contributes to the persistence of labour market disadvantage in economically lagging regions. In most OECD countries, regions with low employment rates two decades ago, i.e., in the mid-2000s, continue to have low employment rates today (Figure 3.16). NEET rates and the share of jobs in high value-added sectors show similar persistence over time. This suggests that people who stay in regions with weaker labour markets may benefit from fewer job opportunities and reduced potential for career progression throughout their lives. These results align with previous evidence highlighting limited regional employment convergence in labour market outcomes in OECD economies since the 1990s, as observed in Japan and the United States (Kondo, 2015[58]), EU countries (Iammarino, Rodriguez-Pose and Storper, 2018[59]) and Türkiye (Gil-Alana, Ozdemir and Tansel, 2018[60]).

Employment barriers, fewer jobs in high value-added sectors and limited training opportunities can hinder earnings progression for residents in disadvantaged regions, unless they move elsewhere. Empirical evidence suggests not only that disparities in earnings across regions are large (Overman and Xu, 2024[61]; Balauz et al., 2023[62]), but also that skilled workers in large cities accumulate more valuable work experience over time, which translates into faster earnings’ progression (Roca and Puga, 2016[19]).

Figure 3.16. Regional disparities in employment are highly persistent over time
Copy link to Figure 3.16. Regional disparities in employment are highly persistent over time

Correlation between employment rates in 2005 and 2023, selection of countries

3.5. Inequalities in access to essential infrastructure and services

Copy link to 3.5. Inequalities in access to essential infrastructure and services

Good health supports people’s ability to work, learn, and participate in society. It improves job performance, reduces absenteeism and lowers healthcare costs for individuals and for the wider economy. Health also contributes to social mobility: healthier individuals are more likely to complete education, enter stable employment, and progress in their careers.

People living in higher-income regions and those with higher levels of education tend to experience better health, which in turn supports further economic opportunity. These advantages reflect not only socio-economic conditions but also differences in access to health services. Well-equipped regions enable residents to maintain good health more easily, while people in regions with limited access to care may face persistent barriers that affect both their well-being and participation in the labour market (OECD/European Union, 2020[63]).

Access to hospitals is one of several factors driving these regional gaps. Most people in OECD countries live within 30 minutes of a general hospital. In large regions (TL2) that have consistently ranked in the top 20% of GDP per capita over the past 20 years, 96% of the population can reach a hospital within 30 minutes, compared to 92% in the bottom 20% (Figure 3.17, Panel A). While this overall difference is modest, country-level gaps can be much larger – reaching 37 percentage points in Greece, 32 in Finland, and 18 in Portugal. In remote or mountainous regions, distance and terrain further limit access.

Regions with higher GDP per capita also tend to have more physician per inhabitant, reinforcing healthcare disparities. These regions average 3.5 physicians per 1 000 inhabitants, compared to 2.9 in regions with the lowest GDP per capita (Figure 3.17, Panel B). In most OECD countries, the physician-to-population ratio increases with regional GDP per capita. This contributes to a cycle where regions with fewer economic resources also face weaker healthcare capacity. The gaps are particularly pronounced in Colombia, Mexico and Türkiye, where regions with lower GDP per capita have fewer than 2 physicians per 1 000 inhabitants, limiting both access to care and the quality of services.

Differences in service provision contribute to unequal health outcomes across regions. People in regions with higher GDP per capita tend to live longer and report better overall health. Across OECD countries, life expectancy in the top 20% of regions by GDP per capita is, on average, two years higher than in the bottom 20% within the same country (Figure 3.17, Panels C and D). Most countries show cross-regional gaps, but they are particularly wide in Colombia, Mexico and the United States, where the differences exceed three years. These patterns point to a persistent link between regional economic conditions and population health.

Figure 3.17. Health outcomes and healthcare infrastructure are better in regions with higher GDP per capita
Copy link to Figure 3.17. Health outcomes and healthcare infrastructure are better in regions with higher GDP per capita

3.5.1. People in regions with lower GDP per capita face poorer digital connectivity

Fast and reliable internet has become essential for full participation in economic and social life. It enables people to search for jobs, work remotely, pursue online education, and start or grow businesses. In this way, digital infrastructure directly shapes opportunities for employment, learning and entrepreneurship, making it a critical factor for economic mobility.

Yet, digital access remains uneven across regions. Gaps in internet quality and coverage – especially between urban and rural areas, and between higher- and lower-income regions – continue to limit the ability of some communities to benefit from the digital economy. These disparities compound other forms of disadvantage and can undermine efforts to promote digital inclusion (OECD, 2021[64]).

Despite progress in broadband coverage and adoption, significant regional differences persist. In large regions (TL2) that have consistently ranked in the top 20% of GDP per capita over the past 20 years, access to broadband tends to be higher than in other regions (Figure 3.18, Panel A). However, the extent of these regional gaps varies across countries, with the largest regional disparities observed in countries where broadband access remains low. In Colombia, Greece and Türkiye, the gap in broadband coverage between the best- and worst-performing regions exceeds 14.7 percentage points. Differences in broadband speed further widen this divide: top regions enjoy average download speeds 8% above the national average, while speeds in bottom regions lag by 10% (Figure 3.18, Panel B). This digital divide constrains access to education, remote work and essential services in less connected places.

Figure 3.18. Digital infrastructure is more developed in regions with higher GDP per capita
Copy link to Figure 3.18. Digital infrastructure is more developed in regions with higher GDP per capita

Bridging this gap requires more than investment in physical infrastructure. It also demands policy approaches that account for local needs, such as supporting digital skills, offering affordable connectivity, and ensuring that digital services are accessible in remote areas. Without such efforts, regional disparities in internet access risk becoming a lasting barrier that may hinder economic opportunities for certain populations and prevent a more level playing field across territories.

3.5.2. Accessible public transport is key for economic opportunities in urban agglomerations

Public transport is essential for ensuring access to jobs, education, and services in urban areas. A well-connected network allows people to commute efficiently, broadens the range of job opportunities they can reach and supports more inclusive labour markets (Glaeser, Kahn and Rappaport, 2008[65]). In contrast, limited public transport restricts mobility and can reinforce spatial inequalities, particularly for those without access to private vehicles.

Access to public transport remains uneven across and within countries. On average, about two in three residents in mid-size and large Functional Urban Areas (FUAs) live within a 10-minute walk of a public transport stop. In countries like Australia, Czechia, Germany and Switzerland, this share approaches 90%. In contrast, fewer than half of residents have such access in Mexico and the United States. In the United States, accessibility to public transport varies more across functional urban areas (FUAs) than in any other OECD country. For example, in 2023, Hamilton (Tennessee) scored just 33% of the national median, while Sonoma (California) reached 185% – a gap of more than 150 percentage points. (Figure 3.19). Other countries show much narrower internal differences, with most FUAs clustering closer to the national median. These differences underscore the uneven geography of transport access, even within the same country.

Transport gaps are particularly challenging for low-income households, who often rely on affordable public mobility. Yet the relationship between poverty and transport access is not straightforward. In some countries, lower-income areas receive less investment due to funding constraints or low population density. In others, targeted policies improve access in high-poverty neighbourhoods. For example, FUAs with higher poverty rates tend to have poorer access to public transport in Spain, while the opposite is true in France (Figure 3.19).

Figure 3.19. Accessibility to public transport varies across cities
Copy link to Figure 3.19. Accessibility to public transport varies across cities

Relative share of the population with access to public transport within a 10-minute walk (median value =100), FUAs, 2023

3.6. Conclusion

Copy link to 3.6. Conclusion

Opportunities in life are not only shaped by individual and family circumstances. Important differences are also observed at regional and local level in terms of access to some of the key drivers of economic opportunity including quality education, employment, services and infrastructure. Many people continue to live near the places where they were born, because of significant barriers to mobility such as existing social ties and local caregiving responsibilities, as well as the financial costs of moving. As a result, the quality of local services and infrastructure, including childcare, schools, transport, and digital connectivity, is a key determinant of people’s life outcomes.

This chapter showed that metropolitan and higher-income regions tend to provide better access to economic opportunities and support greater upward mobility. In contrast, people living in poorer or more remote regions, as well as in disadvantaged neighbourhoods within cities, face persistent challenges that limit their prospects and skew the level playing field. By doing so, the chapter highlighted the importance of measuring inequalities at multiple spatial scales. While smaller geographical units can reveal fine-grained differences, regional-level indicators are essential for informing national policies and the allocation of resources. Despite limitations in data availability, especially at the local level, the evidence presented here points to the significant role that place plays in shaping people’s opportunities. Reducing these place-based disparities remains essential for promoting more equal opportunities and equitable outcomes across OECD countries.

Future work could deepen the analysis of spatial inequalities by exploring additional factors such as exposure to environmental risks, social capital or the quality of local institutions. More granular and comparable data at the local level would support more comprehensive analysis of how neighbourhood conditions affect people's life chances. In particular, administrative data with geographic indicators can help uncover the long-term impacts of growing up in disadvantaged areas. These efforts will support the design of more targeted policies to address geographic dimensions of equal opportunity.

References

[50] Abeynayake, P., E. Bomhoff and G. Lee (2012), “Female Age at First Marriage and Fertility Levels: A Comparison of Developed and Developing Countries”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.2174737.

[1] Adema, W., J. Fluchtmann and V. Patrini (2023), “Mainstreaming gender equality”, in Joining Forces for Gender Equality: What is Holding us Back?, OECD Publishing, Paris, https://doi.org/10.1787/a3452d6b-en.

[25] Aghion, P. et al. (2023), “Anatomy of Inequality and Income Dynamics in France”, https://www.college-de-france.fr/sites/default/files/media/document/2023-05/Anatomy%20of%20Inequality%20and%20Income%20Dynamics%20in%20France%20Philippe%20Aghion.pdf.

[2] Akee, R., M. Jones and S. Porter (2019), “Race Matters: Income Shares, Income Inequality, and Income Mobility for All U.S. Races”, Demography, Vol. 56/3, pp. 999-1021, https://doi.org/10.1007/s13524-019-00773-7.

[57] Almeida, V. et al. (2024), “Geographic inequalities in accessibility of essential services”, OECD Social, Employment and Migration Working Papers, No. 307, OECD Publishing, Paris, https://doi.org/10.1787/12bab9fb-en.

[16] Almeida, V. et al. (2024), “Geographic Inequalities in Accessibility of Essential Services”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.4813454.

[11] Alvarez, M., A. Bernard and S. Lieske (2021), “Understanding internal migration trends in OECD countries”, Population, Space and Place, Vol. 27/7, https://doi.org/10.1002/psp.2451.

[8] Artamonova, A. and A. Syse (2021), “Do older parents’ assistance needs deter parent-child geographic divergence in Norway?”, Health & Place, Vol. 70, p. 102599, https://doi.org/10.1016/j.healthplace.2021.102599.

[62] Balauz et al. (2023), “Spatial wage inequality in North America and Western Europe: changes between and within local labour markets 1975-2019”, CEP Discussion Papers (CEPDP1941). London School of Economics and Political Science. Centre for Economic Performance, London, UK..

[34] Balestra, C. and F. Oehler (2023), “Measuring the joint distribution of household income, consumption and wealth at the micro level”, OECD Papers on Well-being and Inequalities, No. 11, OECD Publishing, Paris, https://doi.org/10.1787/f9d85db6-en.

[30] Banzhaf, H. and R. Walsh (2008), “Do People Vote with Their Feet? An Empirical Test of Tiebout’s Mechanism”, American Economic Review, Vol. 98/3, pp. 843-863, https://doi.org/10.1257/aer.98.3.843.

[53] Bilal, A. (2023), “The Geography of Unemployment”, The Quarterly Journal of Economics, Vol. 138/3, pp. 1507-1576, https://doi.org/10.1093/qje/qjad010.

[10] Bosquet, C. and H. Overman (2019), “Why does birthplace matter so much?”, Journal of Urban Economics, Vol. 110, pp. 26-34, https://doi.org/10.1016/j.jue.2019.01.003.

[45] Brandt, M. and K. Hank (2014), “Scars that will not Disappear: Long-term Associations between Early and Later life Unemployment under Different Welfare Regimes”, Journal of Social Policy, Vol. 43/4, pp. 727-743, https://doi.org/10.1017/s0047279414000397.

[49] Campisi, N. et al. (2020), “Spatial variation in fertility across Europe: Patterns and determinants”, Population, Space and Place, Vol. 26/4, https://doi.org/10.1002/psp.2308.

[21] Cantwell, J. and S. Zaman (2024), “International knowledge connectivity and the increasing concentration of innovation in major global cities”, Journal of Economic Geography, Vol. 24/3, pp. 415-440, https://doi.org/10.1093/jeg/lbae013.

[47] Carcillo, S. et al. (2015), “NEET Youth in the Aftermath of the Crisis: Challenges and Policies”, OECD Social, Employment and Migration Working Papers, No. 164, OECD Publishing, Paris, https://doi.org/10.1787/5js6363503f6-en.

[14] Chetty, R. and N. Hendren (2018), “The Impacts of Neighborhoods on Intergenerational Mobility I: Childhood Exposure Effects*”, The Quarterly Journal of Economics, Vol. 133/3, pp. 1107-1162, https://doi.org/10.1093/qje/qjy007.

[24] Chetty, R. et al. (2014), “Where is the land of Opportunity? The Geography of Intergenerational Mobility in the United States *”, The Quarterly Journal of Economics, Vol. 129/4, pp. 1553-1623, https://doi.org/10.1093/qje/qju022.

[46] De Fraja, G., S. Lemos and J. Rockey (2021), “The Wounds That Do Not Heal: The Lifetime Scar of Youth Unemployment”, Economica, Vol. 88/352, pp. 896-941, https://doi.org/10.1111/ecca.12384.

[12] Deutscher, N. (2020), “Place, Peers, and the Teenage Years: Long-Run Neighborhood Effects in Australia”, American Economic Journal: Applied Economics, Vol. 12/2, pp. 220-249, https://doi.org/10.1257/app.20180329.

[55] Di Cataldo, M. and A. Rodríguez-Pose (2017), “What drives employment growth and social inclusion in the regions of the European Union?”, Regional Studies, Vol. 51/12, pp. 1840-1859, https://doi.org/10.1080/00343404.2016.1255320.

[67] Dijkstra, L. et al. (2021), “Applying the Degree of Urbanisation to the globe: A new harmonised definition reveals a different picture of global urbanisation”, Journal of Urban Economics, Vol. 125.

[68] Dijkstra, L., H. Poelman and P. Veneri (2019), “The EU-OECD definition of a functional urban area”, OECD Regional Development Working Papers, No. 2019/11, OECD Publishing, Paris, https://doi.org/10.1787/d58cb34d-en.

[56] Dromundo, S., M. Lüske and M. Tuccio (2023), “Innovative approaches to tackle long-term unemployment”, OECD Social, Employment and Migration Working Papers, No. 300, OECD Publishing, Paris, https://doi.org/10.1787/e1f7e16e-en.

[73] Espasa Reig, J. et al. (2025), “Household financial fragility and asset poverty in OECD regions: New indicators and an experimental imputation method”, OECD Regional Development Papers, No. 160, OECD Publishing, Paris, https://doi.org/10.1787/ee9f2f16-en.

[26] 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.

[37] Fagereng, A. et al. (2020), “Heterogeneity and Persistence in Returns to Wealth”, Econometrica, Vol. 88/1, pp. 115-170, https://doi.org/10.3982/ecta14835.

[5] Ferreira, F., J. Gyourko and J. Tracy (2010), “Housing busts and household mobility”, Journal of Urban Economics, Vol. 68/1, pp. 34-45, https://doi.org/10.1016/j.jue.2009.10.007.

[4] Ganong, P. and D. Shoag (2017), “Why has regional income convergence in the U.S. declined?”, Journal of Urban Economics, Vol. 102, pp. 76-90, https://doi.org/10.1016/j.jue.2017.07.002.

[60] Gil-Alana, L., Z. Ozdemir and A. Tansel (2018), “Long Memory in Turkish Unemployment Rates”, Emerging Markets Finance and Trade, Vol. 55/1, pp. 201-217, https://doi.org/10.1080/1540496x.2018.1425837.

[65] Glaeser, E., M. Kahn and J. Rappaport (2008), “Why Do the Poor live in Cities? The role of public transportation”, Journal of Urban Economics, Vol. 63/1, pp. 1-24, https://doi.org/10.1016/j.jue.2006.12.004.

[6] Gobillon, L. and H. Selod (2021), “Spatial Mismatch, Poverty, and Vulnerable Populations”, in Handbook of Regional Science, Springer Berlin Heidelberg, Berlin, Heidelberg, https://doi.org/10.1007/978-3-662-60723-7_7.

[38] Heckman, J. (2006), “Skill Formation and the Economics of Investing in Disadvantaged Children”, Science, Vol. 312/5782, pp. 1900-1902, https://doi.org/10.1126/science.1128898.

[59] 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.

[52] İlkkaracan, İ. (2012), “Why so Few Women in the Labor Market in Turkey?”, Feminist Economics, Vol. 18/1, pp. 1-37, https://doi.org/10.1080/13545701.2011.649358.

[48] Jelnov, P. (2021), “The marriage age U-shape”, Journal of Demographic Economics, Vol. 89/2, pp. 211-252, https://doi.org/10.1017/dem.2021.27.

[58] Kondo, K. (2015), “Spatial persistence of Japanese unemployment rates”, Japan and the World Economy, Vol. 36, pp. 113-122, https://doi.org/10.1016/j.japwor.2015.11.001.

[31] Königs, S. and J. Terrero-Dávila (2025), “Who climbs the income ladder?: Cross-country evidence on income mobility from tax record data”, OECD Social, Employment and Migration Working Papers, No. 322, OECD Publishing, Paris, https://doi.org/10.1787/eadbbcbf-en.

[54] Lindenlaub, I., R. Oh and M. Peters (2022), Firm Sorting and Spatial Inequality, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w30637.

[13] McNeil, A., D. Luca and N. Lee (2023), “The long shadow of local decline: Birthplace economic adversity and long-term individual outcomes in the UK”, Journal of Urban Economics, Vol. 136, p. 103571, https://doi.org/10.1016/j.jue.2023.103571.

[35] Mian, A., A. Sufi and E. Verner (2017), “Household Debt and Business Cycles Worldwide*”, The Quarterly Journal of Economics, Vol. 132/4, pp. 1755-1817, https://doi.org/10.1093/qje/qjx017.

[43] Moreno Monroy, A. et al. (2025, forthcoming), “Do neighbourhoods influence the likelihood of receiving social assistance benefits? Evidence from the Netherlands”, OECD Publishing, Paris.

[23] Nie, Y. et al. (2024), “Disparities in public transport accessibility in London from 2011 to 2021”, Computers, Environment and Urban Systems, Vol. 113, p. 102169, https://doi.org/10.1016/j.compenvurbsys.2024.102169.

[3] O’Connell, M. (2019), “Is the impact of SES on educational performance overestimated? Evidence from the PISA survey”, Intelligence, Vol. 75, pp. 41-47, https://doi.org/10.1016/j.intell.2019.04.005.

[39] OECD (2024), Education at a Glance 2024: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/c00cad36-en.

[15] OECD (2023), “Hard work, privilege or luck? Exploring people’s views of what matters most to get ahead in life”, OECD Policy Insights on Well-being, Inclusion and Equal Opportunity, No. 12, OECD Publishing, Paris, https://doi.org/10.1787/e1903ab2-en.

[17] OECD (2023), OECD Regional Outlook 2023: The Longstanding Geography of Inequalities, OECD Publishing, Paris, https://doi.org/10.1787/92cd40a0-en.

[36] OECD (2023), On Shaky Ground? Income Instability and Economic Insecurity in Europe, OECD Publishing, https://doi.org/10.1787/9bffeba6-en.

[69] OECD (2023), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, PISA, OECD Publishing, Paris, https://doi.org/10.1787/53f23881-en.

[41] OECD (2022), Delivering Quality Services to All in Alentejo: Preparing Regions for Demographic Change, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/63ffb4d7-en.

[66] OECD (2022), OECD Territorial Grids, http://stats.oecd.org/wbos/fileview2.aspx?IDFile=cebce94d-9474-4ffc-b72ad731fbdb75b9.

[42] OECD (2022), Shrinking Smartly in Estonia: Preparing Regions for Demographic Change, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/77cfe25e-en.

[32] OECD (2021), Delivering Quality Education and Health Care to All: Preparing Regions for Demographic Change, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/83025c02-en.

[33] OECD (2021), “Inequalities in household wealth and financial insecurity of households”, OECD Policy Insights on Well-being, Inclusion and Equal Opportunity, No. 2, OECD Publishing, Paris, https://doi.org/10.1787/b60226a0-en.

[64] OECD (2021), The Digital Transformation of SMEs, OECD Studies on SMEs and Entrepreneurship, OECD Publishing, Paris, https://doi.org/10.1787/bdb9256a-en.

[29] OECD (2018), Divided Cities: Understanding Intra-urban Inequalities, OECD Publishing, Paris, https://doi.org/10.1787/9789264300385-en.

[40] OECD (2017), The Funding of School Education: Connecting Resources and Learning, OECD Reviews of School Resources, OECD Publishing, Paris, https://doi.org/10.1787/9789264276147-en.

[63] OECD/European Union (2020), Health at a Glance: Europe 2020: State of Health in the EU Cycle, OECD Publishing, Paris, https://doi.org/10.1787/82129230-en.

[27] OECD et al. (2021), Applying the Degree of Urbanisation: A Methodological Manual to Define Cities, Towns and Rural Areas for International Comparisons, OECD Regional Development Studies, OECD Publishing, Paris/European Union, Brussels, https://doi.org/10.1787/4bc1c502-en.

[18] Ormerod, P. (2013), “The persistence of unemployment at the local area level: evidence from the US and the UK”, Applied Economics Letters, Vol. 21/1, pp. 28-30, https://doi.org/10.1080/13504851.2013.835472.

[61] Overman, H. and X. Xu (2024), “Spatial disparities across labour markets”, Oxford Open Economics, Vol. 3/Supplement_1, pp. i585-i610, https://doi.org/10.1093/ooec/odae005.

[22] Owens, A. and J. Candipan (2019), “Social and spatial inequalities of educational opportunity: A portrait of schools serving high- and low-income neighbourhoods in US metropolitan areas”, Urban Studies, Vol. 56/15, pp. 3178-3197, https://doi.org/10.1177/0042098018815049.

[20] Paunov, C. et al. (2019), “On the concentration of innovation in top cities in the digital age”, OECD Science, Technology and Industry Policy Papers, No. 85, OECD Publishing, Paris, https://doi.org/10.1787/f184732a-en.

[19] Roca, J. and D. Puga (2016), “Learning by Working in Big Cities”, The Review of Economic Studies, Vol. 84/1, pp. 106-142, https://doi.org/10.1093/restud/rdw031.

[70] Schiavina, M. et al. (2023), GHS-POP R2023A - GHS population grid multitemporal (1975-2030), https://doi.org/10.2905/2FF68A52-5B5B-4A22-8F40-C41DA8332CFE.

[72] Schiavina, M. et al. (2023), “GHSL Data Package 2023”, Publications Office of the European Union, Luxembourg.

[71] Schiavina, M., M. Melchiorri and M. Peseresi (2023), GHS-SMOD R2023A - GHS settlement layers, application of the Degree of Urbanisation methodology (stage I) to GHS-POP R2023A and GHS-BUILT-S R2023A, multitemporal (1975-2030), https://doi.org/10.2905/A0DF7A6F-49DE-46EA-9BDE-563437A6E2BA.

[44] Schmillen, A. and M. Umkehrer (2017), “The scars of youth: Effects of early‐career unemployment on future unemployment experience”, International Labour Review, Vol. 156/3-4, pp. 465-494, https://doi.org/10.1111/ilr.12079.

[7] Spring, A., B. Gillespie and C. Mulder (2023), “Internal migration following adverse life events: Assessing the likelihood of return migration and migration toward family”, Population, Space and Place, Vol. 30/3, https://doi.org/10.1002/psp.2711.

[9] Thomassen, J., E. Lundholm and G. Malmberg (2023), “Who stays in their birthplace? The role of multigenerational local ties in young adults’ staying behaviour”, Population, Space and Place, Vol. 30/3, https://doi.org/10.1002/psp.2710.

[28] van Ham, M., D. Manley and T. Tammaru (2024), “Geographies of socio-economic inequality”, Oxford Open Economics, Vol. 3/Supplement_1, pp. i634-i641, https://doi.org/10.1093/ooec/odad045.

[51] Villalobos-Hernández, A. et al. (2015), “Embarazo adolescente y rezago educativo: análisis de una encuesta nacional en México.”, Salud pública de México, Vol. 57, pp. 135-143.

Annex 3.A. Typologies for regions, cities, and other areas

Copy link to Annex 3.A. Typologies for regions, cities, and other areas

TL2 and TL3 regions

Copy link to TL2 and TL3 regions

Within the 38 OECD countries, regions are classified into two territorial levels mirroring countries’ administrative structure (OECD, 2022[66]). The 433 OECD "Territorial Level 2" (TL2) regions represent the uppermost subnational administrative tier, such as federal states in Germany. The 2 414 OECD "Territorial Level 3" (TL3) regions denote lower administrative divisions, except in Australia, Canada and the United States.

Degree of urbanisation

Copy link to Degree of urbanisation

The degree of urbanisation classification defines territorial units on an urban-rural continuum, as cities, towns and semi-dense areas, and rural areas. This methodology was jointly developed by six organisations – the European Commission, the Food and Agriculture Organization of the United Nations (FAO), the United Nations Human Settlements Programme (UN-Habitat), the International Labour Organization (ILO), the Organisation for Economic Co-operation and Development (OECD) and The World Bank. For further explanations, see Dijkstra et al. (2021[67]).

Functional urban areas composed of local administrative units

Copy link to Functional urban areas composed of local administrative units

People’s daily lives often span multiple locations that do not match administrative boundaries. They may live in one region, commute to work in another and spend leisure time elsewhere. Flows of people, goods and services – through commuting, business networks, and production linkages – create functional connections between regions. These interactions frequently cross administrative borders and reflect the real geography of economic and social activity.

To reflect these functional connections, the European Commission and the OECD jointly developed a harmonised definition of functional urban areas (FUAs) (Dijkstra, Poelman and Veneri, 2019[68]). FUAs are defined for nearly all OECD countries and consist of a city and its surrounding commuting zone. This approach captures the true economic and functional footprint of cities, based on daily movements of people. For urban areas, FUAs offer a more accurate basis for planning infrastructure, transport, housing, education and recreational spaces. They support better policy design by aligning public investment and services with how people live and move, rather than with administrative borders.

Settlement sizes in the OECD PISA survey

Copy link to Settlement sizes in the OECD PISA survey

The OECD PISA survey classifies schools across six territorial units based on their population size (OECD, 2023[69]). School principals are asked to fill out a questionnaire where they also indicate the size of the settlement where their school is located. The spatial units are as follows:

Fewer than 3 000 people: Village, hamlet or rural area
3 000 to about 15 000 people: Small town
15 000 to about 100 000 people: Town
100 000 to about 1 000 000 people: City
1 000 000 to about 10 000 000 people: Large city
More than 10 000 000 people: Megacity.

Annex 3.B. Measuring the healthcare and essential infrastructure

Copy link to Annex 3.B. Measuring the healthcare and essential infrastructure

Life Expectancy	Figure 3.17 shows relative average life expectancy at birth across small regions (TL3) in 30 countries, based on the most recent data available. Life expectancy at birth represents the average number of years a newborn can expect to live if current age-specific mortality rates persist throughout their lifetime. Data refer to 2022 for AUS, CZE, DNK, ESP, EST, FIN, FRA, HUN, LTU, LVA, PRT; 2021 for GBR, ITA, NOR, SWE; 2020 for DEU, JPN, KOR, TUR; and 2018 for NZL. Source: OECD calculations based on OECD (2024), Life Expectancy – Regions database (accessed in June 2025).
Mortality	Figure 3.17 shows relative crude mortality rates across small regions (TL3) in 34 countries, based on the most recent data available. Crude mortality rate is defined as the number of deaths per 1 000 inhabitants in a given year. Source: OECD calculations based on the OECD Database on Regions, Cities and Local Areas, http://oe.cd/geostats (accessed in June 2025).
Health access	Figure 3.17 shows relative access to hospitals across small regions (TL3) in 16 countries, based on the most recent data available. Values represent the share of the population within a 30-minute drive of the nearest hospital. Access was estimated using point of interest (POI) data, 1-kilometre resolution population grids (Schiavina et al., 2023[70]) combined with high-resolution settlement grids (Schiavina, Melchiorri and Peseresi, 2023[71]) and urbanisation levels, applying the Mapbox Isochrone API. For countries where hospital data was available only as postal addresses, these were converted into geographic coordinates using the geocoder Python package (ArcGIS provider https://developers.arcgis.com/rest/geocode/api-reference/overview-world-geocoding-service.htm). Data refer to 2022 for CHE, CZE, FRA, HUN, JPN, KOR, LTU, LVA, MEX, PRT, SVK, SVN; 2021 for AUS, DEU, EST, NOR, SWE, TUR; 2020 for FIN; 2019 for NZL; and 2011 for ESP. Source: OECD calculations using geospatial methods based on Schiavina et al. (2023[72]) (accessed in June 2025).
Doctors	Figure 3.17 shows the number of physicians across TL3 regions, based on the most recent data available. Physicians include generalists, who provide continuing care to individuals and families, and specialists such as paediatricians, obstetricians/gynaecologists, psychiatrists, medical specialists, and surgical specialists. Data refer to 2022 for CHE, CZE, FRA, HUN, JPN, KOR, LTU, LVA, MEX, PRT, SVK, SVN; 2021 for AUS, DEU, EST, NOR, SWE, TUR; 2020 for FIN; 2019 for NZL; and 2011 for ESP. Source: OECD calculations based on the OECD Database on Regions, Cities and Local Areas http://oe.cd/geostats (accessed in June 2025).
Broadband internet	Figure 3.18 shows the most recent data on the share of households with access to broadband internet, defined as a download speed of at least 256 kilobits per second. Data refer to 2023 for most countries, 2022 for COL, 2021 for USA and ISL, 2020 for GBR, and 2017 for CHL. Source: OECD calculations based on national household survey data and communications regulators.
Fixed internet speed	Figure 3.18 shows fixed download speed for 32 countries. Fixed download speed estimates are measured in megabits per second (Mbps) and are based on user-performed tests from Speedtest by Ookla between 2019 Q1 and 2023 Q2. Data may be subject to testing biases (e.g., faster connections being tested more frequently) or strategic testing by internet service providers in specific markets. As speed-testing methodologies can vary across providers, regional indicators are presented as deviations from the national average (in %). Source: OECD calculations based on Speedtest® by Ookla® Global Fixed and Mobile Network Performance Maps. Based on analysis by Ookla of Speedtest Intelligence® data for 2019Q1-2023Q2. Provided by Ookla (accessed August 2023).

Notes

Copy link to Notes

← 1. These poverty rates have been calculated based on nominal incomes, i.e., they do not consider regional differences in price levels. Gaps in living standards may be narrower than the differences in poverty rates suggest to the extent that the cost of living, and notably housing, is lower in lower-income regions.

← 2. The analysis calculates poverty rates as the share of adult women and men living in households with incomes below the poverty line, hence mirroring the approach used for the child poverty indicator. This way of calculating poverty rates implicitly assumes the equal sharing of resources within households, i.e., it does not try to attribute the various components of household income to different household members.

← 3. For more results on financial fragility and asset poverty at the regional level, see Espasa Reig et al. (2025[73]).

← 4. PISA assessments ask school principals to identify the type of settlement where their school is located. The analysis presented calculates average PISA scores across settlements using this information. PISA does not report data on the location of students’ homes or on the size or geographical location of the settlement where the school is located, so the analysis assumes that students live in the same type of settlement where their school is located, and relevant factors to differentiate settlements, such as their proximity, to cities are ignored. This means that the role of location is only roughly controlled for in this analysis, and further analysis using better proxies for location may result in different findings with respect to the role of place on educational outcomes.

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Countries

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Publications

Publications

Featured publications

Data

Data

Featured data

News & events

News & events

Featured events

About OECD

About

Engage with us

Work with us

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment