SCIENCE

Regional Economics

Abstract. Regional economics investigates how productive activity, welfare, public institutions, labor, capital, housing, knowledge and environmental risks are distributed among interconnected territories and how these patterns evolve. This monograph develops a systematic exposition of regional economic thought, from nineteenth-century location theory and central-place analysis to regional science, cumulative causation, growth convergence, new economic geography and spatial equilibrium. It analyzes agglomeration economies, specialization, industrial transition, transport networks, interregional trade, regional fiscal systems and the measurement of regional performance. Dedicated methodological chapters establish the distinction between workplace production and resident welfare, administrative and functional geographies, accounting decomposition and causal inference, and spatial dependence and structural counterfactuals. Evidence from primary scholarly research, official statistical classifications, economic accounts and the 2026 OECD regional assessment informs the analysis. The synthesis finds that regional policy must address connected economies and identify beneficiaries, transfers, spillovers, displacement, housing capitalization and long-run territorial resilience. The discussion specifies research limits and an agenda for improved longitudinal and functional geographic data. This work synthesizes existing theories and evidence and advances an authorial interpretation without claiming a new independent economic theory.

Keywords: regional economics; regional development; spatial inequality; agglomeration economies; economic geography; core–periphery; regional convergence; local labor markets; spatial econometrics; interregional input-output; regional accounts; market access; place-based policies; regional productivity; housing and migration; territorial governance; climate resilience; functional economic regions

1. The Domain of Regional Economics

Regional economics studies how economic activity, resources, opportunities, institutions, and welfare are distributed among interconnected territories, why these distributions change, and how decisions in one place affect outcomes elsewhere. Its primary objects are regions understood as economic systems whose boundaries are specified for a research question. A region may be an administrative jurisdiction, a labor market, a metropolitan catchment, a rural production system, a border corridor, or a network of trading places. These objects overlap and evolve. A national average is therefore a weighted aggregate of distinct local economies; its trajectory can conceal simultaneous growth, stagnation, demographic decline, and displacement.

The central analytical questions are jointly spatial and economic: where should firms locate, how do households choose places to live and work, how do transport costs shape trade, when do dense clusters enhance productivity, why do some regions converge and others fall behind, and what happens to the rest of the economy when a particular region receives infrastructure or subsidy? Regional outcomes emerge from production, prices, amenities, labor mobility, land and housing markets, demographic processes, public institutions, and exchange networks. The synthesis in The Spatial Economy demonstrates why local, regional, and international locations belong to a connected analytical framework; the region-level welfare analysis in People, Places, and Public Policy shows why policies must be judged through their effects on workers, firms, landowners, and neighboring areas.

A region is specified by its unit of observation, connectivity, population at risk, outcome measures, and institutional authority. These requirements separate descriptive mapping from causal explanation. A prosperous city can occupy a declining administrative region, and a rapidly growing regional product can coexist with falling real income for incumbent residents. The research unit must fit the mechanism. A commuting labor-market area is useful for employment shocks; a watershed can be more suitable for environmental exposure; a jurisdiction may matter for taxation; an export corridor may cross national borders.

Regional economics thus provides an integrated explanation of spatial allocation and territorial development. Its scientific output includes accounting systems, structural models, estimates of spatial responses, identification of mechanisms, and evaluation of interventions. The analysis proceeds from accurately constructed places, through causal and equilibrium mechanisms, to carefully specified social outcomes.

2. Historical Foundations: Land, Markets, and Location

Early formal regional reasoning emerged from the observation that distance changes prices and production choices. Johann Heinrich von Thünen's 1826 original isolated-state study described an idealized central market surrounded by land on which producers choose agricultural uses under transport-cost and rent conditions. In the elementary version, for crop k at distance d, bid rent per unit of land can be represented as R_k(d) = (p_k - c_k)y_k - t_k(d)y_k, where p_k is market price, c_k unit production cost, y_k yield, and t_k(d) the cost of delivering output. The chosen use maximizes attainable rent given land quality and other constraints. The equation is a pedagogical representation of the location logic, rather than a literal transcription of the historical text.

The same logic extends to industrial production, retail markets, transport nodes, and land-intensive activities. Greater accessibility supports higher bid rents where firms or households obtain valuable market access; changes in freight costs, land policy, or infrastructure can alter the spatial ordering. Homogeneous soil, a single market and uniform transport conditions are restrictive assumptions, so the theoretical rings explain directional forces rather than predicted real-world maps.

Central-place theory then shifted attention from agricultural distance to service hierarchies and the market areas of settlements. Christaller's German work, made accessible through the 1966 English Central Places in Southern Germany, examined the spatial ordering of centers that offer goods with different market thresholds and ranges. Central place patterns illuminate why rare specialized services tend to locate in larger centers and ordinary services are distributed more widely. Their idealized geometric regularities depend on strong assumptions about settlement surfaces, demand, costs, and movement. In applied work, actual market areas must be derived from travel time, purchasing power and observed service use.

Regional science developed a more systematic language linking industrial location, spatial exchange, land use, and urban form. Walter Isard's Location and Space-Economy joined these questions into a broad economic framework. The enduring achievement of this tradition is the insistence that transportation costs and relative positions enter the choice and feasibility sets of firms and households. Its limitations are equally instructive: geographic coordinates do not themselves reveal institutions, endogenous innovation, social networks, or barriers to movement. Location analysis therefore becomes the first layer of a fuller regional science.

3. Regional Science, Unbalanced Development, and Cumulative Change

Regional economics acquired developmental depth when scholars treated spatial differences as processes capable of reinforcing themselves. Gunnar Myrdal's Economic Theory and Under-developed Regions advanced cumulative causation: an initial advantage in production, finance, infrastructure or skilled labor can attract additional resources, widening differences over time. Albert Hirschman's Strategy of Economic Development emphasized selective investment, forward and backward industrial linkages, and the dynamic tension between leading and lagging locations. Their approaches explain why the development path depends on where investments occur and how production systems are connected.

Cumulative causation has several concrete channels. A large customer base attracts suppliers, supplier variety lowers costs, and lower costs attract additional firms. A skilled labor pool draws employers that offer specialized jobs, while jobs attract skilled workers. The local fiscal base supports public services, which can reinforce attractiveness. Conversely, outmigration can reduce service viability, weaken local demand, and strain municipal revenue. None of these loops is mechanically permanent: congestion, high housing costs, local competition, institutional reform, or new trade connections can moderate or reverse accumulation.

Regional input-output analysis provides a complementary accounting perspective. If x denotes sectoral gross output and A the matrix of intermediate input requirements, the Leontief identity x = Ax + f yields x = (I - A)^{-1}f where the inverse exists and f is final demand. For regional applications, imports and leakage must be represented explicitly: purchases from outside the region do not all generate local production. An interregional input-output matrix provides flows among regions and sectors; it helps trace propagation but does not establish a causal effect or account automatically for capacity constraints and substitution.

The historical progression from location to linked development clarifies a methodological rule. Spatial differences are generated by both initial conditions and feedback. A region's observed specialization may be the outcome of comparative advantage, path dependence, policy, chance or their combination. An account of causation should identify which channel is operative and which counterfactual trajectory is implied. Regional science is strongest when historical accounts, formal equilibrium conditions and contemporary evidence are used together.

4. Core–Periphery Dynamics and New Economic Geography

Paul Krugman's 1991 Increasing Returns and Economic Geography provided a formal mechanism through which two otherwise similar places may develop different industrial concentrations. Manufacturing producers benefit from economies of scale and demand access; consumers and workers are attracted by varieties and opportunities in centers of production. Trade or transport frictions counterbalance concentration. The resulting core–periphery model captures the possibility that small initial advantages can be amplified and that shifts in transport costs can change the equilibrium distribution of industry.

One simplified market-access relation is MA_r = Σ_s E_s τ_rs^{1-σ} P_s^{σ-1}, where E_s is expenditure in destination s, τ_rs is iceberg trade cost, σ > 1 is the elasticity of substitution, and P_s is a destination price index. The expression illustrates how access depends simultaneously on demand and competitive conditions in other markets. It is a schematic component of a class of economic-geography models, not an empirical estimating equation without additional structure. The spatial outcome depends on firms' scale economies, consumer preferences, factor mobility, expenditure shares, and the geography of trade costs.

Fujita, Krugman, and Venables extend the argument across cities, regions, and cross-border networks in The Spatial Economy. Agglomeration is endogenous: market size both attracts and is increased by productive activity. A change in accessibility can distribute gains through cheaper inputs and new markets, while also relocate firms from previously self-sufficient peripheral communities. This mechanism explains why easier transportation need not generate equal growth everywhere.

A two-region illustration clarifies the distinction. Suppose a transport link lowers delivery costs to both regions. In the peripheral region, households enjoy lower delivered prices and perhaps a wider range of goods. Local firms simultaneously confront stronger competitors in the core. Local measured production may fall even while consumers benefit. Aggregate output can rise while a particular employment base contracts. Regional development is therefore evaluated with explicit outcome and welfare definitions, rather than by assuming that local GDP and residents' living standards move together.

New economic geography also faces identification challenges. Observed manufacturing concentration may reflect scale economies, historical infrastructure, geography, institutions, selection of productive firms, or an unobserved amenity. Structural models make these alternatives transparent but need credible parameterization and out-of-sample tests. Policy analysis must assess equilibrium responses in connected locations, not only the directly treated territory.

5. Agglomeration Economies: Sharing, Matching, Learning

Agglomeration raises productivity when spatial proximity enables economic interactions that would be more costly at a distance. Duranton and Puga's micro-foundations of agglomeration organize mechanisms into sharing, matching, and learning. Sharing includes common infrastructure, specialized suppliers and pooled indivisible facilities. Matching refers to improved search and pairing of firms, workers and tasks in thicker markets. Learning arises from interaction, imitation, movement among employers, and knowledge spillovers.

These mechanisms differ in their spatial reach and policy implications. A freight terminal can support firms over a transport corridor; a specialized labor pool may operate at the scale of a commuting zone; tacit knowledge diffusion can attenuate over shorter distances. Rosenthal and Strange's survey of agglomeration evidence emphasizes that industrial, geographic and temporal scope each matter. The statement that 'density causes productivity' is therefore incomplete without a causal design and a mechanism-specific spatial unit.

Ciccone and Hall's 1996 productivity-and-density analysis estimated that doubling employment density is associated with higher productivity in their U.S. framework, with an estimate around six percent under the studied specification. This result is historically influential, but its numerical magnitude is study-specific and should not be transferred unchanged to other countries, decades, or definitions of density. Density itself can be caused by prior productivity, and sorting of high-productivity firms or skilled workers can confound simple comparisons.

Agglomeration encounters rising costs. Scarce land, congestion, pollution, disease exposure, and infrastructure crowding can lower net benefits. The region's optimum density differs from the private optimum if agents do not internalize spillovers and costs. Housing regulation affects how production advantages translate into population, wages, housing prices and accessibility. An intervention that expands employment without elastic housing supply may transfer a significant share of its benefit into land rent.

The general analytical lesson is that concentration is neither universally beneficial nor universally harmful. The relevant object is the marginal social benefit of proximity relative to its marginal congestion, housing, environmental and distributional costs, evaluated through the full network of affected places.

6. Specialization, Diversity, Clusters, and Industrial Change

Regional specialization can reflect natural resources, labor skills, market access, historical networks, supplier linkages and strategic policy. A location quotient, LQ_{rk} = (E_{rk}/E_r)/(E_k/E), compares industry k's employment share in region r with the national share. A value above one indicates relative concentration, not necessarily competitive advantage, export success, local productivity or future growth. Employment shares may differ from value-added shares; the choice of variable matters.

The 1992 Growth in Cities study by Glaeser, Kallal, Scheinkman, and Shleifer examined large industries across 170 U.S. cities between 1956 and 1987. Its evidence associated local competition and urban diversity with industry employment growth more strongly than same-industry specialization in that historical sample. The finding informs debates between within-industry spillovers and cross-industry learning, but does not establish a universal preference for diversified local industrial structures. Industry composition, technological regimes, exposure to shocks and city size all mediate the relationship.

Michael Porter's cluster analysis stresses locally linked firms, institutions, specialized service providers, and knowledge relationships. Clusters may generate productivity advantages when complementary activities and organizational capacities reinforce one another. However, their definition often overlaps with ordinary sectoral concentration. Martin and Sunley's critique of cluster concepts cautions that flexible terminology can make almost any set of firms appear to be a cluster and weaken evaluation. A scientifically useful cluster claim specifies geographic reach, industry boundaries, actual transactions, innovation channels, governance and measurable counterfactual effects.

Structural transformation requires attention to the sequence of activities. A region can shift from agriculture to manufacturing, from manufacturing to business services, or toward tourism and care services; each transformation changes employment requirements and susceptibility to external demand. Policies should assess skills transfer, wage quality, productivity levels and transportability of services, not only firm counts. Supplier ties may produce multipliers, but leakages to other regions and displaced activities reduce simple headline effects.

The governing distinction is between observing co-location and proving an agglomeration externality. Co-location can be a private response to common advantages. An externality exists only where one agent's location or activity changes another's productivity, matching opportunities or learning beyond market-priced transactions. This criterion guides both research design and industrial policy.

7. Mobility, Spatial Equilibrium, and Regional Welfare

Regional prosperity is mediated by migration, commuting, housing, wages and the local cost of living. In a stylized open-city equilibrium, a mobile household compares indirect utility V_r = V(w_r, h_r, p_r, a_r, T_r), where w is earnings, h housing expenditure, p nonhousing prices, a amenities and T commuting or migration costs. If mobility were frictionless for identical households, utilities would tend toward equalization across populated locations. In practice, heterogeneity in skills, preferences, family commitments, legal constraints and financial resources permits persistent differences in living standards.

Jennifer Roback's wages-and-rents framework demonstrates how differences in amenities and productivity become capitalized into wage and land-rent gradients. A high nominal wage may compensate for expensive housing or an undesirable amenity, while low nominal wages can coexist with valued local quality of life. The sign of a wage difference alone therefore cannot rank places by welfare.

This logic transforms evaluation of local employment incentives. When a subsidy raises demand for local labor, wages may increase and workers may move in. Inelastic housing supply can raise rents; owners benefit through capitalization even as renters face higher costs. If productivity gains attract firms and housing expands, more residents can share the benefits. Kline and Moretti's spatial welfare treatment makes these distribution and general-equilibrium channels central to the social calculation.

Regional migration is selective. Young adults often respond more to career opportunities than older homeowners; credential recognition and child care affect movement; borders can restrict labor access. Regional labor-market estimates must distinguish workplace employment from resident employment, local residents from newcomers, and temporary commuting from permanent relocation. A growing employment count in a region can coexist with weak wage progression for incumbent low-skilled workers.

The proper welfare accounting combines earnings, local prices, housing, commuting, fiscal services, risk exposure, amenities and changes in access to opportunity. Regional economics therefore studies the trajectories of both territories and people. These perspectives interact through migration and capital flows, but a single place-level indicator never represents all affected households.

8. Regional Growth, Convergence, and Divergence

Growth analysis distinguishes changes in the scale of a regional economy from changes in living standards and productivity. Let Y_r be regional real output, N_r residents, E_r employed workers, and H_r hours. An accounting identity is Y_r/N_r = (Y_r/H_r)(H_r/E_r)(E_r/N_r). Growth in product per resident can therefore be decomposed into productivity per hour, hours per worker and employment per resident. Each component depends on demographic structure and boundary definitions. This accounting identity establishes sources of observed change; causal attribution requires separate evidence.

Barro and Sala-i-Martin's Convergence investigated whether poorer U.S. states grew faster than richer states and found convergence in the historical datasets they studied. β-convergence typically estimates whether initially lower log income predicts faster subsequent growth, conditional on structural characteristics. σ-convergence asks whether dispersion of outcomes declines. These are distinct propositions: poorer areas can grow faster on average while the distribution remains wide due to heterogeneous shocks, mobility, or instability.

A common descriptive regression is (ln y_{r,T} - ln y_{r,0})/T = α + β ln y_{r,0} + γ'X_r + ε_r. A negative β is conventionally interpreted as a conditional catch-up association under an identified steady-state framework. It is not by itself evidence that convergence is mechanically guaranteed or that a specific policy caused catch-up. Measurement error in initial output, reverse causality, regional sample selection and spatially correlated errors require diagnostic attention.

Myrdal's cumulative-causation argument and the new economic geography suggest mechanisms for divergence, while neoclassical growth and diffusion can produce catching up. These trajectories can coexist at different times and scales. Within-country metropolitan productivity can rise relative to peripheral areas even as national incomes converge internationally. A region may approach a lower long-run steady state and remain poorer than a leading city.

An adequate growth study states the convergence object—output per resident, consumption, real disposable income, productivity, employment, or access to essential services—and the population to which it applies. Changes in national accounts revisions, regional borders and industry classification must be reconciled before trend claims are accepted.

9. Territorial Measurement and Regional Accounting

Empirical regional economics starts by defining a territorial statistical object and a consistent accounting concept. Regional GDP measures gross value added generated by production located in a territory plus applicable taxes less subsidies on products. Household disposable income measures income accruing to residents after relevant transfers and taxes, while a regional employment count may be recorded by job location. These variables answer different questions. The U.S. regional economic accounts provide an official illustration: GDP by state and county, personal income, employment, regional price parities and real personal income are separate data products designed for distinct analytical tasks.

Consider a commuter who works in a productive metropolitan center and lives in an adjacent county. The worker contributes to product at the workplace, but receives personal income as a resident elsewhere. Housing expenditure and public service consumption may occur in both places. A comparison of county GDP per resident can therefore be distorted by cross-boundary commuting, particularly where an administrative county is a small component of a functional labor market.

OECD's 2026 territorial reader's guide distinguishes large and small administrative regions (TL2 and TL3), functional urban areas, labor market areas and grid-cell urbanization categories. Eurostat's NUTS classification similarly defines NUTS 1, 2 and 3, with the NUTS 2024 boundaries effective January 1, 2024. The 2027 version is identified separately and becomes valid January 1, 2027. These revisions are consequential when long panels join data from changing territorial vintages.

For a region r, total value added can be represented as GVA_r = Σ_k GVA_{rk}. An interregional sum agrees with national totals only under the accounting system's treatment of extra-regional territories, coverage, boundaries and allocation rules. Real regional growth often requires price indices, chain-linking and sectoral deflators. Applying a national deflator is a defensible convention when regional price data are limited, but can miss region-specific inflation and changing local cost structures.

The methodological requirement is a data dictionary recording spatial unit, boundary vintage, date, base period, current versus constant prices, place of work versus place of residence, population denominator, confidentiality suppression, seasonal adjustment and revision history. Comparisons without these controls can yield statistically precise answers to an incorrectly specified question.

10. Spatial Inequality and Distributional Decomposition

Regional inequality exists both between territories and within them. A population-weighted measure is needed when the question concerns the distribution of residents' economic opportunities; an equally weighted regional measure answers a different question about the dispersion of territorial units. Large cities and sparsely populated areas should not receive identical weights by default when estimating person-level inequality. Neither choice is intrinsically correct; the evaluative population determines the weights.

The OECD's Regions and Cities at a Glance 2026 documents divergent productivity, housing access, demographic conditions and climate exposure across territories. Its most recent comparison reports, for the 2018–2024 period, that among 13 member and accession countries where regional GDP-per-capita gaps narrowed, faster productivity in lagging regions was the main contribution in six and higher employment in lagging regions in seven. This is a decomposition of observed convergence in a defined sample, not a causal policy experiment or a claim covering every OECD country.

A useful generalized entropy measure is the Theil T index, T = Σ_r s_r ln(y_r/ȳ), where s_r is the income share of region r and ȳ is an appropriately defined average under consistent weights. For individual incomes, the standard formulation uses person-level observations; when applied to region means with income weights, it describes a between-region component. Total person-level inequality can be decomposed into within-region and between-region contributions under the compatible data construction. Because weights and units define the estimand, aggregated regional means cannot reconstruct within-region income distributions.

European territorial inequalities have been analyzed in Regional Inequality in Europe, which emphasizes geographically differentiated development paths and the importance of historically specific policy instruments. Regional income gaps may coexist with different patterns in life expectancy, housing affordability, employment access or educational outcomes. An income-based national equalization rule therefore requires complementary measures of social infrastructure and actual opportunity.

Evidence should be interpreted at three levels: the regional dispersion of means, the distribution of persons within places, and the mobility of people among places over time. This prevents an improvement in the region-level average from being automatically credited to improvements experienced by its incumbent population.

11. Spatial Econometrics, Dependence, and the Boundaries of Inference

Economic variables observed in neighboring regions often covary because of trade, shared infrastructures, commuting, environmental processes, policy spillovers or spatially patterned omitted variables. Statistical independence is consequently a substantive assumption, not a default feature of administrative data. Luc Anselin's Spatial Econometrics provides the classical analytical vocabulary for modeling spatial dependence and spatial heterogeneity.

For an outcome vector y and a prespecified weights matrix W, a spatial lag model can be written y = ρWy + Xβ + ε. Here Wy aggregates neighboring outcomes under a documented spatial relationship. In a spatial-error model ε = λWε + u, dependence operates through the disturbances. These models imply different causal stories; correlation in nearby outcomes is not enough to establish either model. A row-standardized W can be based on contiguity, distance, commuting, trade or network links, but each choice embeds an economic hypothesis and influences estimates.

Patrick Moran's foundational spatial autocorrelation work underlies the familiar Moran's I statistic, which compares cross-products of geographically linked deviations with overall dispersion. Significant spatial autocorrelation can indicate clustered outcomes or residual misspecification. It does not directly tell the researcher whether contagion, common exposure, selective migration or regional institutions are responsible.

The modifiable areal unit problem makes an additional warning unavoidable: statistical coefficients can change when a territory is partitioned differently, even if the underlying point-level phenomena are unchanged. A difference between districts might disappear when districts are merged, while a strong national trend can be misleading at a neighborhood scale. Aggregating sparse regions may protect confidentiality but obscure local deprivation and spatial exclusion.

Reliable regional inference therefore discloses the construction of W, spatial scale, geographic crosswalks, outcome transformation, estimation method, spatial standard errors and robustness to alternative boundary systems. In causal policy analyses, interference between treated and untreated regions must be addressed explicitly: the treatment in one jurisdiction may influence control outcomes in adjacent jurisdictions.

12. Identifying Regional Causal Effects

Regional associations frequently arise because policymakers select distressed or fast-growing territories for intervention. An investment zone's subsequent recovery cannot simply be compared with an untreated prosperous region; eligibility, anticipated change, and spillovers can confound the contrast. Credible empirical analysis specifies the treatment, assignment mechanism, counterfactual population, observation window and potential displacement channels.

Difference-in-differences methods estimate effects from relative changes over time under a parallel-trends assumption appropriate to the untreated counterfactual. Event studies assess pre-treatment trajectories and dynamic effects, but non-rejection of pre-trend tests does not prove identification. Geographic regression discontinuity can exploit boundaries when treatment changes discontinuously while relevant determinants remain locally smooth. Natural experiments in roads, migration, industrial closures and administrative rules can be informative when their exclusion restrictions and selection mechanisms are evaluated in the research setting.

Shift-share designs are common when local industry structures create different exposures to national or global shocks. A generic exposure measure is Z_r = Σ_k s_{rk,0}g_k, where s_{rk,0} is a predetermined local share and g_k a common industry shock. Goldsmith-Pinkham, Sorkin, and Swift's Bartik Instruments show how identification depends on exposure-share assumptions in typical applications. Borusyak, Hull, and Jaravel's shift-share research designs develop an alternative framework in which quasi-random shocks justify identification under explicit shock-level conditions. These papers establish distinct research designs; the mere arithmetic form of Z cannot validate an instrument.

Causal regional analysis also distinguishes local treatment effects from national general equilibrium. A subsidy may attract a new plant from a neighboring region; the treated locality gains jobs while aggregate output changes little. Transport corridors can expand demand access in several territories while changing their sectoral mix. Benefit-cost evaluation should examine the effect on all places touched through trade, mobility and public finances, not only the administrative recipient.

Good empirical practice includes preregistered mechanisms when feasible, transparent sample exclusions, uncertainty intervals, sensitivity to spillovers and alternative geographies, and disclosure of source data and methods. Claims of causal identification belong to particular estimands and institutional settings. Extrapolation to other regions is a separate question requiring structural comparability or additional evidence.

13. Infrastructure, Transport, and Market Access

Transport and digital networks change the economic distance between locations, affecting firm costs, consumer access, commuting, land prices and regional specialization. The transportation-costs synthesis by Redding and Turner emphasizes that infrastructure has multi-region general-equilibrium effects on wages, population, trade and industrial composition. An improved link can raise total market access without improving each locality's production or employment.

Evidence from the United States illustrates substitution across localities. Chandra and Thompson's rural interstate highway study reports different sectoral responses and reallocation of activity toward directly served counties and away from some neighboring counties. This is a warning against counting all local gains along a corridor as new national activity. A highway can reroute trade and commerce as well as create genuine net efficiencies.

Benjamin Faber's China trunk-highway analysis investigated peripheral counties connected indirectly by a network designed around major cities. His research reports reduced GDP growth in non-targeted peripheral counties, with an industrial-output mechanism, under the study's identification strategy. It offers a concrete counterexample to the assumption that connecting a peripheral area necessarily industrializes it. Better access can also expose its producers to stronger metropolitan competitors.

Market access must be operationalized through relevant costs. Road distance, actual travel time, reliability, border checks, congestion, tariffs, digital latency and multimodal transfer differ in economic meaning. A corridor's advantage may depend on ports, warehousing, safe transport, customs administration and last-mile access rather than highway length alone. Infrastructure appraisal should test realistic origin-destination demand matrices and distributional effects across households and firms.

The appropriate policy criterion combines travel-time benefits, safety, emissions, maintenance, housing responses, fiscal costs, opportunity costs, supply-chain resilience and redistribution. A project that improves aggregate efficiency may still require direct support for displaced workers or firms, and a local project's long-run benefits can be captured outside the investing jurisdiction.

14. Regional Labor Markets, Skills, and Employment Multipliers

Regional labor markets are shaped by employers' demand, labor supply, skill specificity, search networks, commuting and migration frictions. A workplace employment measure counts jobs physically located in a territory; a resident employment rate measures whether people living there work, including outside its boundary. The distinction is especially important for metropolitan peripheries, tourism areas and cross-border commuter regions. The OECD labor-market-area methodology delineates functioning labor markets primarily through patterns of commuting among smaller territorial units.

Local multiplier analysis asks how an initial demand shock affects employment outside the directly affected industry. Moretti's Local Multipliers models and estimates how growth of a city's tradable sector influences nontradable employment through local expenditures, endogenous wages and prices. A numerical multiplier is conditional on the geography, labor and housing supply, horizon, initial shock and classification of tradable activity. It should not be used as a fixed conversion factor in every grant application or public project appraisal.

A closure of a dominant plant can remove direct jobs, weaken supplier purchases, reduce retail demand and accelerate outmigration. The effect on residents depends on portable skills, alternative employers, asset ownership, unemployment support, commuting possibilities and new investment. A growing high-skill technology sector may increase local product while leaving less-skilled incumbent residents with weaker opportunities if occupational upgrading is limited.

Human capital policies have regional incidence. Training can raise individual earnings even when successful graduates leave the originating place; the destination may capture the tax base and knowledge spillovers. An evaluation should therefore identify whether the aim is to benefit residents, expand local production, strengthen a public institution, or provide national labor mobility. These goals can reinforce one another, but are measured differently.

Labor-market policy analysis is strongest when it follows people and firms through time, connects place of work with residence, and accounts for wage distribution, job quality, hours and the cost of housing. A headline increase in local employment is an intermediate indicator of development, not its complete measure.

15. Housing, Land Supply, and Metropolitan Networks

Housing is a central regional production constraint because workers must live within practical reach of employment. In regions with high productivity and limited housing supply, new demand can raise rents and house prices more rapidly than resident population. As a result, productive employment centers may become inaccessible to workers who would benefit from moving there, while owners capitalize part of the growth dividend. Spatial equilibrium links productivity, living costs, migration and neighborhood amenities rather than treating housing as a separate consumption category.

The 2026 OECD housing analysis finds that housing prices in urban centers were 27 percent above national averages across 23 countries in the relevant dataset, while areas outside functional urban areas were 34 percent below the national average. These are comparison statistics for specified territories and data availability, not causal estimates of zoning effects. The same study discusses supply constraints, urban concentration and substantial differences in metropolitan purchase affordability.

Land scarcity is physical in some centers and institutional in others. Planning rules, infrastructure networks, environmental hazards, finance, construction capacity and market expectations shape the effective supply curve. A transit investment can improve accessibility without proportionately increasing housing capacity; its gains may capitalize into station-area land values. Conversely, coordinated transport, zoning and utility provision can permit broader access to employment centers.

Regional housing analysis requires distinguishing prices from quantities, owners from renters, and current residents from movers. Housing affordability may be expressed as price-to-income, rent burden, residual income or access to appropriate dwelling quality; none captures all relevant costs. Homes can be cheap in a shrinking region because employment and amenities are scarce, while homes can be expensive in a growing center because access to productive jobs is valuable.

A credible development strategy therefore integrates housing construction, commuting networks, municipal finance, climate risk, and public-service capacity. The relevant regional unit often includes the commuting basin, not merely the city issuing construction permits.

16. Rural Regions, Peripheral Areas, and Territorial Access

Rural regions are heterogeneous economic systems. Some specialize in productive agriculture or natural resources; others provide recreation, environmental services, energy, logistics or specialized manufacturing. Many have small populations, thin markets for labor and services, long travel times and substantial exposure to demographic change. A single rural category can conceal accessible peri-urban areas, remote mountain settlements, islands and resource-dependent districts with fundamentally different development possibilities.

The World Bank's Reshaping Economic Geography organizes spatial constraints through density, distance and division. Density concerns concentration of activity, distance concerns economic separation from markets, and division concerns barriers between territories. For peripheral regions, improving access to education, health services, digital connectivity and employment can matter more to residents' welfare than attracting a large symbolic industrial project. Different distances—travel time, trade cost, service availability and social disconnection—have different remedies.

Peripheral economic change often reflects the interaction of tradable and nontradable sectors. A globally competitive agricultural producer may generate export income but employ relatively few workers as productivity increases. Local tourism can create labor demand but expose households to seasonality. Resource extraction can increase measured regional product while profits accrue to nonresident owners and environmental costs remain locally concentrated. Regional analysis must therefore keep ownership, distribution and sustainability visible.

The regional development approach studied by Barca, McCann, and Rodríguez-Pose emphasizes adapting institutions and interventions to specific territorial capabilities. A place-based measure can be justified by coordination failures, externalities, persistent exclusion or local public goods. It requires a measurable objective, credible local implementation capacity and an evaluation of its national opportunity cost.

Rural policy cannot be reduced to maximizing population or local factory counts. Access to public services, demographic security, sustainable livelihoods, mobility and territorial stewardship are substantive outcomes. Where population decline is persistent, maintaining high-quality services may call for shared provision, mobile delivery and cooperation across administrative jurisdictions.

17. Institutions, Fiscal Capacity, and Territorial Governance

Regions operate within overlapping systems of national law, regional government, municipal budgets and cross-border institutions. Fiscal transfers, regulatory rules, zoning authority, public procurement and service provision affect territorial development because they influence both incentives and local implementation capacity. Institutions alter the returns to infrastructure, education and firm investment. A public investment can fail when the financing authority, operating authority and beneficiary population fall in different jurisdictions.

Territorial public finance distinguishes revenue capacity from expenditure need. A wealthy employment center may collect substantial taxes from business activity while adjacent commuter jurisdictions supply housing, schools and social services. Equalization grants can reduce fiscal disparities but their formulas may also affect marginal incentives for local revenue effort. Funding design therefore requires transparent service responsibilities, tax bases, commuter flows, and redistributive goals.

The place-based policy debate includes the argument that institutional heterogeneity and coordination among complementary investments matter for results. Transport, workforce training, land regulation and university-industry links can produce larger combined effects when each removes a binding obstacle. However, policy bundling complicates causal evaluation: an observed local improvement cannot automatically be assigned to every element of the package.

A useful governance test is operational: which entity selects projects, which supplies recurrent funding, whose incentives are affected, what data will be reported, and who can correct implementation failures? Institutions are not abstract residuals. They determine permit timelines, contract enforcement, land assembly, intergovernmental cooperation and the credibility of program commitments.

Regional governance also requires accountability to the people affected by spillovers. A municipality's subsidy can relocate firms from another municipality, and an upstream jurisdiction's decisions can impose costs downstream. Regional compacts, shared labor-market planning and transparent impact assessments help align jurisdictional decisions with functional economic geographies.

18. Innovation, Knowledge Diffusion, and Regional Productivity

Innovation is spatial because research capability, occupational specialization, advanced services, networks and commercialization opportunities are unequally distributed. Universities, firms, laboratories, investors and public agencies may reinforce local learning when they can exchange personnel and ideas. Their effectiveness depends on institutional quality, firm absorptive capacity, entrepreneurial finance and access to larger markets. A high patent count can signal invention while revealing little about diffusion to smaller enterprises or average household prosperity.

The original Growth in Cities evidence on diversity and competition informs the hypothesis that cross-industry interaction can support learning; the sharing-matching-learning framework explains why labor mobility and specialized inputs matter. Yet observed co-location of innovative firms may be caused by selection into regions with valuable prior infrastructure or human capital. Innovation policy needs to identify whether universities, research collaborations, training, market integration or firm networks actually relax the regional constraint.

The OECD's 2026 regional comparison documents considerable concentration of knowledge-intensive services, patents and venture investment in large urban regions. Its 2024 sectoral decomposition associates knowledge-intensive business services with an important contribution to labor productivity growth in many observed countries, under the report's defined sector coverage. The linked industry productivity methodology uses real gross value added per worker, industry-specific national deflators, and a within-industry versus industry-reallocation decomposition. This helps describe where measured productivity growth occurs, while offering no automatic causal proof that a sectoral subsidy will reproduce the pattern.

Knowledge networks extend beyond administrative boundaries. Remote collaboration can improve information access while face-to-face matching, local labor markets and tacit learning continue to influence clustering. The decisive question is whether a peripheral firm can absorb and use knowledge generated elsewhere. This depends on digital access, managers, education, complementary infrastructure, contracts, suppliers and market incentives.

Regional innovation assessment should trace inputs, outputs and diffusion separately: research expenditure, workforce skills, patents, firm formation, commercialization, adoption, productivity and wages. The chain contains distinct failure points. Public policy can target demonstrable missing links without assuming that every territory should imitate the industrial composition of a leading technology hub.

19. Regional Trade, Global Value Chains, and Border Frictions

Regional economies are open systems. Their output depends on intermediate inputs, export demand, household purchasing power, capital investment and institutional access to external markets. A decline in trade cost can raise the competitiveness of some sectors while exposing others to stronger import competition. The spatial-economy approach connects firm geography with international trade; the transport-cost review emphasizes that network changes redistribute economic activity through prices and market access.

Cross-border regions encounter multiple forms of friction: customs procedures, regulatory incompatibility, currency risk, language, logistics, labor rules and data restrictions. Economic integration may create manufacturing corridors, cross-border commuter zones and shared service markets, but administrative indicators often truncate these systems at national boundaries. A subnational analysis that ignores flows across the border can mismeasure both demand and employment.

A regional export multiplier depends on where value is created and who receives it. Gross exports can include substantial imported inputs. Regional value added is the local contribution after intermediate purchases, while worker income and local tax revenue are further distinct components. Where a high-value plant purchases most inputs elsewhere and remits profits to distant owners, local economic benefits may be concentrated in labor income and a limited supplier network.

Infrastructure outcomes are particularly context dependent. The China trunk-highway findings show that peripheral industrial activity can be adversely affected by stronger integration with large centers. In other settings, improved links permit specialization that raises local productivity. The effect cannot be predicted from the direction of infrastructure expenditure alone; it depends on market structure, tradable composition, skills and existing competitive relationships.

A territorial trade strategy therefore combines logistics reliability, human capital, supplier upgrading, regulatory clarity and export market diversification. Its evaluation follows productivity, employment and real household welfare, accounting for activity displaced from neighboring territories and the resilience of the regional production network.

20. Regional Shocks, Resilience, and Industrial Transition

Regional economies face shocks that are asymmetric even when national aggregates remain stable. A change in global commodity prices can affect a mining district much more than a diversified service region. A major employer's closure can create long unemployment spells and reduce municipal revenue. Climate events, pandemics, migration, trade disruption and technology adoption similarly interact with inherited regional specialization and institutional capacity.

Resilience has several meanings: resistance to an initial shock, speed of recovery, the sustainability of adaptation, and the ability to change the local production structure. These should be measured independently. A region can restore employment by expanding low-wage services while failing to recover real productivity or household security. Equally, output can recover rapidly while the original workforce remains displaced by automation.

Industrial transition policies need a credible account of asset specificity. Skills learned in a declining sector may transfer partly to adjacent activities; industrial sites can be repurposed only when remediation, infrastructure and market access permit. Workforce programs are more successful when they connect training directly with observable employer demand. In resource-dependent regions, environmental liabilities and concentrated fiscal reliance can complicate recovery even when local product was previously high.

Rodríguez-Pose's study of declining places links persistent lack of opportunity and territorial discontent to broader political-economic consequences. The paper is an interpretive research contribution, not an assertion that every distressed region reacts identically. It identifies the need to examine regional exclusion as a durable political and social condition alongside productivity statistics.

An effective resilience evaluation follows the region before, during and after a shock; compares structurally similar exposures; and records the distribution of employment, real earnings, migration, public-service quality and investment. The appropriate policy response may differ across regional histories even when the immediate headline shock is similar.

21. Climate Change, Energy, and Environmental Regional Economics

Climate risks are geographically uneven because hazards, population exposure, economic activities, infrastructure and adaptive capacity vary across territory. Heat affects labor productivity and electricity demand; flooding can disrupt transport and housing; drought can affect water-intensive sectors and agriculture; coastal erosion influences land values and public liabilities. Regional climate policy consequently concerns production, migration, finance, social services and infrastructure planning together.

The OECD 2026 evidence reports that in nearly one in five large OECD regions, average temperatures in 2021–2025 exceeded the 1981–2010 reference period by more than 1.5 degrees Celsius. This is an observed local comparison under the report's method, not a statement about global mean warming or a causal estimate of economic damages. The report also analyzes the overlap of heat stress and population ageing and the uneven geography of the energy transition.

Regional decarbonization depends on starting structures. A manufacturing region with high process-heat demand faces different costs from an agricultural district or a service-oriented metropolis. Energy access, grid infrastructure, workforce transition, housing insulation and public transport have interconnected effects. The costs of mitigation and adaptation are allocated across taxpayers, ratepayers, workers and property owners, so equity analysis must trace incidence rather than presume that all local spending benefits all residents.

Risk appraisal should integrate hazard probabilities, asset exposure, vulnerability and recovery capacity. A common conceptual relation is expected loss = Σ_j probability_j × damage_j, with each damage term conditional on the exposed capital, population and adaptation. Under climate nonstationarity, historical frequencies may fail to describe future hazards; scenario analysis and transparent uncertainty ranges are necessary.

Sustainable regional development considers both the locational efficiencies of concentration and the local hazards that concentration creates. The relevant planning horizon extends beyond a budget cycle because land use and infrastructure can lock in exposure for decades. Effective regional science connects spatial economic models to climate scenarios, engineering feasibility and distributional welfare.

22. Regional Social Mobility, Amenities, and Public Services

The place where a person grows up can influence later economic opportunity through school quality, local networks, safety, health, peer environments and access to employment. The regional scale relevant for these mechanisms may be smaller than an administrative region, but the consequences can accumulate into regional differences in lifetime income and population mobility.

Chetty, Hendren, and Katz's Moving to Opportunity analysis links a randomized housing-voucher experiment to long-term tax records. The study finds beneficial adult outcomes for children who moved to lower-poverty neighborhoods when younger than thirteen, with effects that varied by age at move. This evidence concerns a specific U.S. program, population and cohort; it does not give a universal numeric effect for all forms of regional relocation. It establishes that exposure to local environments can matter causally for children's later outcomes.

Public services influence mobility constraints. Reliable education, health care, child care, transit and broadband can expand people's feasible opportunity sets without immediately relocating large firms. Conversely, a new industrial investment can fail to attract workers if housing and schools cannot accommodate additional families. Evaluating opportunity requires measuring access in actual travel time and cost, not merely the count of facilities within a jurisdiction.

Spatially targeted public-service investments can improve residents' welfare even when they do not induce measurable regional GDP growth. A repaired hospital, stable drinking-water network or accessible transit line can reduce risk and time burdens. The welfare case may be strongest for groups with limited mobility, including children, older residents, people with disabilities and households with low liquid wealth.

The distribution of opportunities among persons and across places are related yet distinct. A regional policy may improve the circumstances of incumbent residents without narrowing the region's national GDP gap; a migration policy may improve movers' lifetime income while the source territory loses population. A rigorous evaluation documents which objective the intervention serves and who receives the benefit.

23. Evaluating Place-Based Policies

Place-based policy directs resources or changes rules for a geographically specified beneficiary area. Instruments include infrastructure, tax relief, public services, land development, training, innovation support and intergovernmental coordination. Its scientific evaluation begins with a market or institutional failure: underprovided local public goods, agglomeration externalities, financial constraints, coordination problems or persistent exclusion. A spatially targeted expenditure that merely transfers commercial activity from one region to another can impose fiscal costs without creating equivalent national benefits.

Kline and Moretti's people-and-place welfare framework requires tracking local prices, household location choices, landownership and interregional spillovers. Barca, McCann, and Rodríguez-Pose's place-based versus place-neutral analysis emphasizes institutional context and complementarities. These approaches can be reconciled through empirical counterfactual evaluation: a policy may be locally tailored while its benefits and costs are judged from the perspective of the full affected economy.

A transparent policy appraisal estimates ΔW = benefits to affected households and firms - resource and fiscal opportunity costs - external damages, applying a specified social welfare function and discounting convention where appropriate. Cash transfers, higher land prices and fiscal receipts require careful treatment: some are redistributions rather than new aggregate surplus. A policymaker should report both aggregate net effects and their distribution across residents, movers, landlords, businesses and neighboring territories.

A practical evaluation record states the objective, target geography, treatment mechanism, comparison strategy, implementation fidelity, outcome horizon and displacement assumptions. Programs with several components need evidence on each component's contribution when feasible. Statistical impact estimates should be integrated with administrative feasibility, maintenance costs and the durability of the local economic response.

A policy should not be called successful simply because investment was spent, buildings were constructed or jobs appeared inside a boundary. Success is a defensible improvement in a defined welfare or development objective relative to a credible counterfactual, including effects outside the designated area.

24. Contemporary Evidence: Regional Economics in 2026

The most recent source reviewed for this monograph is the OECD Regions and Cities at a Glance 2026 report released on November 4, 2026. The report gives a contemporary cross-national picture of productivity, employment, innovation, housing, demographic change and heat exposure. It documents substantial variation among regions and emphasizes the declining demographic scope for catching up through employment growth alone. These are comparable indicators across member and accession countries with available data, not a single worldwide panel with uniform coverage.

One result is that metropolitan regions increased their share of national population in 35 of 37 countries with data between 2005 and 2024. The report projects metropolitan regions to grow fastest or shrink least in 27 of 33 countries by 2045. Projections are conditional on demographic assumptions and should be distinguished from observations. It also notes that the urban geography of knowledge-intensive work and investment can reinforce housing and labor-market pressures.

In the 2025 occupational analysis reported by the OECD, employment highly exposed to generative artificial intelligence accounts for 52 percent in regions with the highest tertiary-educated workforce shares, compared with 42 percent elsewhere. The indicator concerns estimated occupational exposure, which is neither a realized employment loss nor a predicted net productivity gain. Regional outcomes depend on adoption, complementary skills, firm organization, regulation and changes in the location of demand.

The same 2026 housing chapter measures urban affordability and regional contrasts, while the methodologies appendix defines labor market areas through official national approaches to commuting and local labor-market integration. This expanded attention to functional geographies improves diagnosis beyond broad administrative averages. The 2026 evidence supports a general interpretation that productivity potential, housing constraints, population change and climate risk increasingly interact; it does not establish the universal optimal program for any particular territory.

Regional science in 2026 must therefore treat economic geography as a joint system of technology, demography, settlement, mobility and environmental exposure. Spatial relationships require current data and historically consistent definitions. The fact that a report is recent improves the relevance of its observation window, while its estimates remain subject to sampling coverage, measurement definitions and subsequent revision.

25. Structural Models, Counterfactuals, and Regional Forecasting

Structural regional models describe how households, firms and governments respond to prices and constraints across places. Their purpose is to evaluate counterfactual changes in accessibility, productivity, amenities, tariffs, housing supply or transfers under explicit assumptions. A comparative-static counterfactual asks where the system would settle after a change; a dynamic model adds adjustment costs, capital accumulation, expectations, migration persistence and the sequence of transitions.

A stylized multi-region model specifies firms' cost and demand schedules, trade frictions τ_rs, worker mobility costs m_rs, housing supply, local amenities, national resource constraints and the market-clearing conditions that join wages and prices. Krugman's core–periphery mechanism identifies endogenous concentration forces, while the transportation costs synthesis shows how transport shocks can propagate across regions. The assumptions governing substitution, mobility and land supply determine whether benefits appear as higher real wages, housing costs, population or output.

Input-output analysis traces accounting linkages. Spatial econometric analysis describes statistical dependence. Reduced-form designs estimate effects of past treatments under identification conditions. Structural estimation makes more ambitious out-of-sample counterfactuals but relies on the model and identifying restrictions. These are complementary tools, with different evidentiary burdens. None should be used to claim that a precise regional forecast is certain.

Regional forecasts should disclose baseline population, projected industry mix, productivity assumptions, housing construction, fiscal conditions, energy costs and climate exposure. Alternative paths can include a central scenario, a positive investment or diffusion scenario, and a stress scenario with migration and production disruption. Forecast uncertainty should include both parameter uncertainty and uncertainty about the institutional and technological mechanisms themselves.

A credible regional counterfactual also has an aggregation check. If one region's employment grows because workers relocate from another, the model's national totals should reflect the transfer rather than count both as new employment. The possibility of congestion and price capitalization requires consistency between real and nominal outcomes.

26. Interdisciplinary Connections and Geographic Scale

Regional economics draws on microeconomics for choice under constraints, macroeconomics for growth and structural change, trade theory for market access, public economics for taxation and public goods, labor economics for employment and mobility, and economic history for path dependence. It engages geography to define landscapes and spatial networks, urban studies to analyze land use and metropolitan form, political science to assess territorial representation, demography to study migration and aging, and environmental science to quantify hazards.

The value of these connections is methodological specificity. Regional health economics may require medical service catchments rather than legislative districts. Energy economics may require a transmission network. A supply-chain investigation may need firm-to-firm links across countries. Demographic service planning may focus on travel times to schools and clinics. A regional employment study generally benefits from commuting-defined labor market areas such as those documented in the OECD geographic classification.

Scale creates both empirical opportunities and risks. Metropolitan growth may be beneficial within a national account while contributing to increasing inequality among peripheral communities. A transport corridor can generate a high average return while harming firms in an intermediate town. Neighborhood-level effects can differ from city-level estimates because housing and labor-market responses operate over distinct spatial ranges. The modifiable areal unit literature explains why analysts must audit the sensitivity of statistical patterns to geographic partitioning.

Interdisciplinary analysis is most rigorous when each theory is attached to the object it explains. Spatial econometrics models dependence in data; planning law determines what can legally be built; engineering establishes infrastructure feasibility; social policy assesses people who face movement barriers. Combining their insights does not remove the need for separately validated data and causal claims.

27. Analytical Synthesis: Economic Places as Interdependent Systems

The evidence reviewed supports an integrated interpretation of regional development: a territory's trajectory is shaped by market access, productive capabilities, institutional arrangements, mobility constraints, housing capacity, and interregional feedback. Each dimension has observable measures, established theoretical foundations and distinct policy channels. This synthesis brings the literature together without introducing a new named theory or claiming unverified novelty.

First, regional growth and regional welfare are different empirical objects. Growth accounting tracks production and productivity; resident welfare includes real income, amenities, public goods, security and access. The distinction is established by workplace-residence accounting, spatial compensating-differential theory, and experimental neighborhood research. The analyst specifies the object before selecting measures and interventions.

Second, the regional benefit of connectivity is conditional. Lower trade costs can promote market access, diffusion and consumer gains while restructuring local industry. Core–periphery models and the trunk-highway evidence describe the tension. A place may lose production even when the economic system gains. Full regional policy evaluation must therefore separate efficiency, spatial incidence and transitional loss.

Third, initial advantages and current institutions interact. The circular processes emphasized by Myrdal's regional analysis can magnify inherited specialization; local institutions govern which investments are feasible and whether complementary public goods are supplied. Regional divergence may arise even if fundamental technologies are available nationally. Policies must identify the binding local bottleneck and any national constraints.

Fourth, spatial scale is part of theory rather than clerical geography. Commuting, supplier transactions, environmental exposure and political authority operate over different boundaries. A single regional polygon cannot serve all questions. OECD functional geographies and spatial econometric methods provide tools for specifying and testing these relationships.

Fifth, evaluation should link empirical identification with equilibrium reasoning. An experiment can establish a local causal effect, while a structural model interprets resulting wages, migration and spillovers. A model can propose a mechanism, while direct evidence determines its empirical relevance. This reciprocal relationship is a sound basis for public decision-making and future regional research.

28. Research Frontiers and Methodological Agenda

Several fronts are especially important for future regional economics. The first is microgeographic longitudinal measurement: firm, worker and household panels linked to reliable territorial crosswalks can distinguish productive gains from population selection and reallocation. Privacy-preserving aggregation and reproducible metadata are essential for scientific use of these data. The second is functional-network measurement, where commuting, supplier trade, digital connections and disaster exposure describe regions more accurately than single administrative borders.

The third frontier is the joint analysis of artificial intelligence, knowledge-intensive services and territorial labor demand. The OECD 2026 report documents geographically concentrated exposure to generative AI and innovation. Future causal research must distinguish occupation-based exposure indices from actual adoption, wage effects, business creation, work reorganization and productivity changes. A spatial technology model needs to specify what tasks can be delivered remotely and which interactions remain local.

The fourth is housing and climate interaction: high-opportunity areas may be constrained by housing supply while other places are vulnerable to heat, flooding or infrastructure decay. Policy evaluation should combine land use, adaptation spending, insurance costs, migration and long-run fiscal sustainability. The fifth is the study of declining regions where demographic contraction, local tax base erosion and limited service access reinforce one another.

Researchers should improve disclosure of weighting schemes, price indices, geographic boundary versions, causal identification, spatial spillovers and external validity. The share-based and shock-based shift-share approaches exemplify how similar-looking estimators can rely on distinct identification conditions. Such distinctions must be reported rather than hidden behind a common instrument label.

The immediate empirical program is to combine richer subnational data with sharper counterfactual evaluation, including systematic scrutiny of displacement and distribution. Better regional policy follows from resolving mechanisms, costs and beneficiaries in specified settings, not from projecting a standard intervention onto every territory.

29. Conclusion

Regional economics explains the spatial organization of production, employment, income, innovation and welfare in an economy of connected and changing places. Its intellectual development runs from transport costs and market areas through regional science, cumulative causation, convergence, core–periphery dynamics, agglomeration micro-foundations, spatial econometrics and contemporary causal policy evaluation. These traditions provide mutually informative accounts of location choices, growth paths, spatial interactions and distributional consequences.

The strongest empirical conclusion is that spatial concentration, regional growth and human well-being must be measured and evaluated separately and jointly. Local growth may arise from national reallocation; improvements in market access may help consumers while displacing local firms; rising productivity may be capitalized into housing prices; and resident opportunity can improve through services or mobility without a large increase in local output. These mechanisms follow from established theoretical and empirical work, and their relative importance varies by region and policy.

The appropriate scientific method therefore begins by defining a functional or administrative territory that fits the question. It constructs consistent accounts of output, residents and costs; identifies channels through trade, labor, housing, institutions and environmental exposure; selects causal or structural tools suitable to the evidence; and evaluates benefits and costs across connected territories. Current comparative sources, including the OECD's 2026 regional evidence, show the importance of integrating demographic shifts, technology, housing and climate exposure into that analysis.

Regional economics will continue to develop through improved spatial data, transparent identification and models that connect geographic change to people and institutions. Its lasting contribution is a rigorous explanation of why location matters for economic opportunity and how regional interventions alter an interdependent economy.

References


Published by Aisentica Development.