Balanced and imbalanced: global population spatial mobility and economic patterns in coastal and interior areas

The study addresses whether global coastal and interior areas are moving toward balanced development as population and economic factors relocate in the context of globalization and the SDGs (particularly SDG10 on reducing inequalities). It motivates a re-examination of coastal–interior disparities given coastal advantages (industry, capital, infrastructure) and historically lagging interiors. It posits that economic factors have begun shifting inland and asks three questions: (1) Is there a new trend in population mobility across coastal and interior blocks? (2) How do dynamic spatial distributions of economy and population shape coastal–interior economic patterns? (3) Is the coastal–interior gap narrowing or widening? Grounded in core–periphery theory, the study uses population as an entry point to analyze 21st-century spatial dynamics and the interactive relationship between population and economic factors across the two regional systems to inform sustainable development and policy.

Past work shows coastal concentration of population and output (e.g., ~50% of global population within 100 km of coasts; high shares of output) and coastal advantages in transport, communication, and infrastructure. Macro- and development economics link population mobility to economic growth and regional disparities; population moves toward higher income and opportunities, reflecting imbalanced wealth and capital mobility. Two debates persist: (1) whether population and economic spatial changes are temporally misaligned (population lags) or (2) whether balanced development implies consistent spatial distributions of economy and population. Classic and neoclassical migration theories (Ravenstein; Schultz; Todaro) emphasize economic motives; agglomeration and market access theories (Marshall; Krugman; Redding & Venables) highlight spatial inequalities. Empirically, studies document both co-movement and misalignment of population and economic factors. This study extends the literature by providing a dynamic, long-period, global multi-scale analysis of coastal–interior population–economy interactions and testing nonlinear threshold effects of per capita income and population–economic density on factor mobility relationships.

Study design and spatial division: The global land area was partitioned into two sub-regions using GIS: "near regions" (coastal areas within 100 km of the sea or ocean-navigable river) and "far regions" (interior areas beyond 100 km). A new GIS database covering 172 countries compiled attributes for GDP, population, and area for 2000–2018. Near regions approximate coastal areas; far regions approximate interiors. Data sources: Population counts come from WorldPop Hub (1 km resolution, WGS84, people per pixel; random forest-based dasymetric redistribution integrating census, survey, satellite, mobile data). Economic data use constant GDP (2010 USD). Additional spatial layers include global coastlines and water vectors (Natural Earth) and a world vector map (World Bank). Nighttime lights were used in inspection steps (details in Supplementary). Variables: - Regional population proportion (RPP): share of population in a region (near/far, global and continental) relative to total. - Regional economic proportion (REP): analogous share for GDP. - Per capita GDP (PGDP): GDP divided by population within near/far regions. - Population density (PD): population per unit land area. - Economic density (ED): GDP per unit land area. - Population–economic density (PED): population relative to GDP level (using GDP as the economic development indicator). Analytical approach: 1) Descriptive spatial statistics and GIS mapping to profile global and intercontinental distributions (including latitudinal bands), PD and ED comparisons, RPP and REP trajectories, and PGDP gaps (ratios between near and far). 2) Correlation and regression by income group (World Bank classification of 172 countries into high-, upper-middle-, lower-middle-, low-income), fitting relationships between GDP and population in near and far regions and computing Pearson correlations and best-fit regression models (R² > 0.8 overall). 3) Nonlinear threshold panel regression (Hansen, 1999) to test whether the effect of population (POP) on GDP (G) depends on: (a) per capita income level (pg = PGDP) and (b) population–economic density (PED). Models specify single and multiple thresholds with indicator functions splitting the sample into regimes, include control variables ED and PD, and allow for region-specific effects. Estimation and inference: Stata 14.0, bootstrap (1,000 replications) to obtain F statistics and p-values for threshold significance. For pg as threshold: single thresholds in near and far regions significant at 1% (double/triple not significant at 10%). For PED as threshold: single and double thresholds significant (triple not). Threshold values and regime-specific coefficients were estimated for global near and far regions. Timeframe: 2000–2018.

- Global distribution and densities: Coastal near regions comprise 18.43% of land but host 52.8% of population; far regions (81.57% land) host 47.2%. PD and ED are much higher along coasts: near-region PD rose 121→152 people/km² (2000→2018) vs far-region PD 28→36; near-region ED rose 1.378M→2.143M USD/km² vs far-region ED 151,600→284,100 USD/km². - Latitudinal concentration: About 23% of the global population is in coastal near regions of the 10°–30° N/S subtropics (5.9% of land), ~19% in 30°–60° temperate near regions (8.5% land), ~7% in 0°–10° tropical near regions (3.2% land), indicating uneven climatic–latitudinal distribution. - Shifts in shares (2000–2018): Global near-region population share decreased slightly 49.71%→48.97% (−0.74 pp); far-region population share increased 50.29%→51.03%. Global near-region GDP share decreased 67.25%→63.02% (−4.23 pp), showing stronger economic than population inland shift and an asynchronous evolution (population lagging economy). - Global PGDP gap narrowed: Near vs far per capita GDP ratio fell from 2.08× to 1.78×, implying a trend toward balanced development despite asynchrony. - Intercontinental patterns and PGDP gaps: • North America (slight convergence): near-region GDP share 55.28%→50.87%; population share 59.19%→57.32%; PGDP gap 0.85×→0.77× (interior higher PGDP overall). • South America (slight convergence): GDP 58.75%→52.43%; population 51.98%→51.89%; PGDP gap 1.32×→1.02×. • Oceania (slight convergence yet still coastal polarization): GDP 91.30%→84.24%; population 86.14%→82.56%; PGDP gap 1.69×→1.13×. • Europe (slight convergence with reverse mobility): GDP 69.74%→68.41% (coastal share down); population 51.28%→53.64% (coastal share up); PGDP gap 2.19×→1.87×. • Africa (expansion of disparities with reverse mobility): GDP coastal share 47.55%→48.88%; interior population share 72.95%→74.16%; PGDP gap increased 2.44×→2.74×. • Asia (significant contraction with reverse mobility): interior GDP share 20.42%→30.75%; coastal population share 52.45%→52.97%; PGDP gap 3.53×→2.00×. Three mobility trajectories identified: (1) population and economy landward (North America, South America, Oceania); (2) population landward–economy seaward (Africa); (3) economy landward–population seaward (Europe, Asia). - Income-group correlations: Strong positive population–economy correlations globally, decreasing with development level; highest in high-income, lowest in low-income groups. Fitted models achieved high R² (>0.8). - Threshold results (global): • pg (PGDP) thresholds: Near-region single threshold at ~USD 3,942. When pg ≤ 3,942, population inflow hinders wealth accumulation (coefficient smaller and implies pressure from surplus labor); when pg > 3,942, population inflow promotes wealth accumulation (coefficient positive and significant). Far-region pg threshold slightly higher (≈ USD 4,335). Population inflow consistently increases wealth, with stronger effects at higher pg (both regimes significant at 1%). • PED thresholds: Near-region double thresholds at PED = 0.2323 and 1.8345. Effects of population mobility on wealth: PED ≤ 0.2323: +0.5621 (5% sig); 0.2323 < PED ≤ 1.8345: +0.6046 (5%); PED > 1.8345: −7.1630 (1%). Far-region double thresholds at PED = 0.2938 and 1.5764: PED ≤ 0.2938: +0.7577 (1%); 0.2938 < PED ≤ 1.5764: +0.5010 (5%); PED > 1.5764: −7.508 (1%). Thus, high PED (indicative of high population relative to economic level) turns population inflows into a drag on wealth accumulation in both sub-regions.

Findings indicate a global trend toward balanced development between coastal and interior areas despite asynchronous factor mobility, with economic factors moving inland more strongly and population following more slowly. Reverse trajectories between population and economy at continental scales do not necessarily widen gaps; instead, the development stage largely conditions outcomes. Post-industrial regions (e.g., North America, Europe) exhibit relatively balanced coastal–interior development and small PGDP gaps, with some interiors surpassing coasts in PGDP due to mature manufacturing, infrastructure, and environmental risks along coasts. Early-stage regions (e.g., Africa) face limited mobility and widening gaps due to weak endowments, market access, and infrastructure. Middle-to-late industrializing regions (e.g., Asia) experience rapid factor reallocation, narrowing gaps, aided by integration initiatives (e.g., Belt and Road, China–Europe rail) improving interior connectivity. The observed population lag relative to economy reflects differing labor needs across industrial stages, adjustment frictions in household location decisions, and settlement stickiness. Policy implications include prioritizing per capita balance over uniform factor distributions; supporting interior growth poles, logistics, and infrastructure; managing industrial–labor gradient transfers; avoiding excessive factor concentration in low-development coastal pockets; and leveraging thresholds: raising regional pg and keeping PED within ranges where population inflows maximize wealth accumulation. Global and continental strategies should align with agglomeration–diffusion dynamics to ensure that no regions are left behind.

The study constructs a global coastal–interior panel (172 countries, 2000–2018) to jointly analyze population–economy interactions. Main contributions: (1) It documents stark spatial disparities: coasts (18.43% of land) host 52.8% of population; coastal subtropics (10°–30°) alone host ~23% of global population within near regions. (2) It shows factor mobility is asynchronous—economic inland shifts outpace population—but both trend landward, and the global PGDP gap narrows from 2.08× to 1.78×, signaling emerging balance. (3) It identifies three intercontinental mobility trajectories and three economic patterns: slight convergence (North America, South America, Oceania, Europe), expansion (Africa), and significant contraction (Asia), driven chiefly by industrialization stage. (4) It reveals nonlinear threshold effects: in coasts, population inflows hinder wealth when pg ≤ USD 3,942 but help when pg is higher; wealth gains from population inflow are maximized at intermediate PED (0.2323–1.8345) and turn negative at high PED. In interiors, higher pg strengthens positive impacts of population inflow (threshold ≈ USD 4,335), with benefits greatest at low-to-mid PED (≤ 1.5764) and negative at high PED. Overall, improving per capita income and maintaining population–economy densities within favorable ranges in both coasts and interiors can narrow gaps and foster balanced development. The work reframes balanced development toward per capita convergence, informs monitoring and prediction of global spatial balance, and offers actionable thresholds for policy design.

The analysis focuses on population proportion changes as a proxy for population mobility and does not explicitly model or quantify key drivers such as climate change, social welfare systems, transportation costs and distance, housing/real estate prices, and policy-induced migration frictions. Data constraints (e.g., reliance on modeled gridded populations and national GDP allocations) and aggregation at near/far regional scales may mask subregional heterogeneity. These omitted factors and finer spatial dynamics are proposed for future research.