The Samarkand region is located in the central part of Uzbekistan, within the Zarafshan River valley. The region covers approximately 16,400 km² and borders the Jizzakh, Navoi, Bukhara, Kashkadarya and Surkhandarya regions, as well as Tajikistan to the south-east. Its geographical position has historically supported the development of settlement networks, trade routes, agriculture, tourism and regional economic connections. A map of the study area illustrating the administrative boundaries, urban settlements, and key spatial characteristics is presented in Figure 1.

Administratively, the region includes 14 districts and major urban settlements, including Samarkand city and Kattakurgan city (Table 1). Samarkand city is the regional centre and one of the largest and most historically significant cities in Uzbekistan. By 2025, the population of the region had reached approximately 4.3 million people, making it one of the most densely populated regions of the country.

Source: Compiled by the authors based on statistical data.

The economy of the region is diversified and includes agriculture, industry, construction, tourism, trade, education and services. The central and south-eastern districts, especially Samarkand, Bulungur, Jambay, Tayloq and Urgut, demonstrate higher population concentration and stronger urban functions. In contrast, the northern and western districts remain more rural and agriculture-oriented. This spatial heterogeneity makes the Samarkand region a suitable case for analysing urbanisation transformation and socio-economic links.

The empirical foundation of the study was based on official statistical data published by the State Committee of the Republic of Uzbekistan on Statistics for the period 2010–2025 (State Committee of the Republic of Uzbekistan on Statistics, 2010–2025). This time span was selected for two reasons: it captures the full cycle of institutional and economic reforms initiated after 2016, and it is sufficiently long to identify persistent urbanisation patterns and compare development trajectories across different reform phases.

The dataset was compiled to reflect the multidimensional character of urbanisation and was organised into three thematic blocks corresponding to the principal dimensions of analysis: demographic, socio-economic and production–spatial (Table 2). All the indicators were obtained from the State Committee of the Republic of Uzbekistan on Statistics and reflect annual data for the Samarkand region over the period 2010–2025.

Note: The demographic block captures population dynamics and settlement structure; the socio-economic block reflects labour market conditions and consumer activity; the production–spatial block assesses economic output, investment flows and land-use transformation.

The selection of indicators was informed by contemporary approaches in urban economics and regional studies (Duranton & Puga, 2023; Koster & Thisse, 2024). Those chosen satisfy three practical criteria: they reflect structural changes across demographic, social and economic spheres; they are available as consistent annual time series; and they are widely applied in comparable regional urbanisation studies. In particular, natural population growth, employment, retail turnover, industrial output, investment and agricultural land area have been used as core urbanisation indicators in studies examining post-Soviet and Central Asian contexts (Zhang et al., 2020; Ma & Sun, 2020; Turemuratov et al., 2024), as well as in broader comparative research on urbanisation and socio-economic development (Ciommi et al., 2018; Polinesi et al., 2020). The use of the urban population share as the key urbanisation indicator follows the approach recommended by UN-Habitat and Eurostat (UN-Habitat, 2020; OECD, 2019) and has been adopted in analogous regional-level analyses (Norimov, 2024).

Before analytical procedures were applied, the dataset underwent multi-stage preprocessing. First, missing observations in the time series were addressed using linear interpolation in order to preserve the continuity of temporal trends. Second, year-to-year changes were examined for each indicator to evaluate temporal stability and identify potential structural breaks or random fluctuations. Third, since the indicators were measured in different units, all the variables were standardised to a comparable scale using z-score transformation. For each indicator x, the z-score was calculated as Equation 1.

where is the observed value in year I; x̄ is the mean value over the study period; and is the standard deviation. This procedure eliminates the effect of differing measurement scales and allows direct comparison of indicators across groups. Fourth, logarithmic transformation was applied to indicators exhibiting high year-to-year variability – most notably industrial production, construction volume and investment in fixed capital – in order to reduce heteroscedasticity and enhance the stability of temporal dynamics prior to correlation analysis.

Correlation analysis. To assess the relationships between urbanisation level and key demographic, socio-economic and spatial indicators, the Pearson correlation coefficient was employed. For the two variables X and Y observed over n time periods, the coefficient was calculated as Equation 2.

where and are the respective sample means. The coefficient r ranges from −1 to +1, withvalues close to ±1 indicating a strong linear relationship, and those close to 0 indicating the absence of a linear association. The results are presented in the form of correlation matrices for each analytical block (demographic, socio-economic and production–spatial). Following the approach recommended in recent methodological literature (Caminiti et al., 2026; Espey et al., 2024), only the lower triangle of the symmetric matrix is reported in order to avoid redundancy and improve visual clarity. Statistical computations were performed using STATA 17.0.

Spatial analysis. To examine territorial differences in urbanisation and socio-economic development across the 14 districts of the Samarkand region, spatial analysis was conducted using ArcGIS 10.6. The spatial dataset included administrative boundaries, the distribution of urban and rural populations, population density, indicators of natural population change and migration balance, and selected socio-economic characteristics at the district level.

Thematic choropleth maps were produced for the following indicators: urbanisation level (share of urban population in total district population), population density, natural population increase rate, net migration balance, and projected urbanisation trajectories to 2040. For each mapped indicator, values were classified into five intervals using the natural breaks (Jenks) method, which minimises within-class variance and maximises between-class differences, thereby ensuring that district boundaries reflect meaningful spatial discontinuities rather than arbitrary quantile thresholds. To identify the spatial typologies of districts, a classification was applied based on two dimensions: the level of urbanisation and the dominant demographic dynamic (natural increase versus migration-driven growth). This produced four district types: (1) highly urbanised with sustained natural growth; (2) moderately urbanised with mixed demographic drivers; (3) predominantly rural with positive natural population growth; and (4) peripheral rural with negative migration balance. The resulting typology is presented cartographically in Figures 3 and 4, and serves as the spatial framework for interpreting the correlation results reported in Section 4. The overall research framework and analytical workflow are summarised in Figure 2.

Population and urbanisation forecasting. To project urbanisation trajectories at the district level up to 2040, the trend extrapolation method was applied based on the annual time series of urban population share for each of the 14 districts over the period 2010–2025. For each district, a linear regression model was fitted to the observed urbanisation trend, and the resulting slope coefficient was used to extrapolate the urban population share forward to 2040 under a baseline scenario assuming continuation of current demographic and economic trends.

The forecasting procedure consisted of three steps. First, the annual rate of change in urban population share was calculated for each district individually in order to account for district-level heterogeneity in urbanisation dynamics, rather than applying a uniform regional rate. Second, linear extrapolation was applied separately for each district using the equation 3.

where Ût is the projected urban population share in year t; Û0 is the observed share in the base year 202; b is the estimated annual rate of change derived from the 2010–2025 trend; and t is the number of years beyond 2025. Third, the resulting district-level projections were mapped using ArcGIS 10.6 and classified into four urbanisation trajectory groups: accelerated urbanisation (projected share exceeding 70% by 2040); moderate urbanisation (50–70%); stable mixed (30–50%); and predominantly rural (below 30%). It should be noted that these projections represent a baseline scenario and do not account for potential policy interventions, economic shocks or structural breaks in demographic trends. They are intended to illustrate spatial tendencies rather than to provide precise quantitative forecasts.

The demographic analysis shows that urbanisation in the Samarkand region is strongly associated with natural population growth. During the period 2010–2025, the total population of the region increased from 3.119 million to 4.315 million. This growth was accompanied by an increase in both urban and rural populations, indicating that urbanisation in the region is not based on rural depopulation, but on simultaneous population growth across settlement types.

All the key demographic indicators showed very high positive correlations with urban population size, as detailed in the time series presented in Table 3. The demographic results confirm that the urbanisation process in the Samarkand region follows an endogenous model. Urban population growth is primarily driven by natural population increase rather than large-scale migration from rural to urban areas. Figure 3 illustrates the district-level distribution of urban and rural populations, providing a spatial reference framework for comparing urbanisation patterns with the statistical relationships identified through the correlation analysis.

In many districts of the Samarkand region, the fertility rate exceeds the mortality rate, resulting in a positive natural population increase. The intensity of natural population change varies across the region and is generally higher in areas with a younger population age structure (Figure 4). The positive natural population growth is associated with several factors: (1) the predominance of a young population; (2) the persistence of traditional reproductive norms, including early marriage; (3) the preservation of rural lifestyles despite the rapid pace of urbanisation; and (4) the fact that urban population growth in the Samarkand region is not accompanied by a sharp decline in the rural population, which helps to maintain an overall positive demographic balance.

Figure 4 shows the spatial distribution of demographic indicators in the Samarkand region for 2025, indicating that demographic processes were developing unevenly. In the central and south-eastern districts, population concentration was higher and a positive natural population increase was being maintained. In contrast, the northern and south-western peripheral districts were characterised by lower fertility rates and were experiencing population outflows, which is reflected in a negative migration balance.

The correlation matrix presented in Figure 5 clearly illustrates the nature of the statistical relationships between the size of the urban population and key demographic indicators in the region over the period 2010–2025. These indicate the presence of very strong associations between the variables, confirming the close interdependence between demographic change and urbanisation processes. The correlation matrix shows that very strong statistical relationships exist between the urban population size and the main demographic indicators. Nearly all correlation coefficients exceed 0.9, indicating a synchronous pattern of demographic change in the region during the period under review.

Particularly strong relationships can be seen between urban population growth and total population size, together with population density. This suggests that urbanisation processes in the Samarkand region are mainly developing not through a simple redistribution of population between urban and rural areas, but under conditions of simultaneous demographic growth and increasing settlement density. High correlation values with fertility indicators confirm the key role of natural population increase as one of the principal drivers of urbanisation dynamics.

At the same time, a stable positive correlation between urban and rural population sizes reflects the specific demographic development pattern of the region, in which urban growth is not accompanied by a sharp decline in the rural population. This points to the persistence of close functional linkages between urban and rural territories and to the gradual, evolutionary nature of urbanisation, rather than its abrupt or discontinuous manifestation.

On the basis of the time series presented in Table 3, the relationship between urban population growth and key demographic indicators – total population size, rural population size, numbers of births and deaths, and population density – was analysed. The assessment was made using the Pearson correlation coefficient.

Source: Compiled by the Authors Based on Statistical Data.

The relationship between urban population size and population density was found to be almost maximal (r ≈ +0.97). This indicates that urban growth in the Samarkand region occurs primarily through the intensification of already developed areas, rather than through active territorial expansion. In other words, cities are “growing inward,” concentrating their population within existing settlements. Such a model is consistent with contemporary global trends, in which compact and higher-density urban development is playing an increasingly important role. Based on the provided data, Pearson correlation coefficients (r) were calculated between the key indicators. All correlations were positive and very high (r > 0.90), confirming a strong linear interrelationship between the processes. Between urban population size and total population of the region, the correlation coefficient was r = +0.998 (r² ≈ 0.996, meaning that 99.6% of the variance is mutually explained). This indicates that growth in the region’s total population is almost entirely reflected in the growth of the urban population. Between the urban and rural population sizes, the correlation is also very high (r = +0.992, r² ≈ 0.984, or 98.4% of the variance explained). Unlike the classical urbanisation model associated with rural depopulation, the Samarkand region demonstrates parallel population growth in both urban and rural areas.

High fertility rates and sustained natural population growth have generated significant demographic pressure. In cities, these processes are more intense due to increasing population density. In the urban areas of the region, demand for social and economic infrastructure has risen, leading to emerging challenges in the provision of schools, hospitals, kindergartens and employment opportunities. The correlation coefficient between the size of the urban population and the number of births equals r = +0.967 (r² ≈ 0.935, i.e. 93.5% of the variance explained). The relationship between the urban population size and number of deaths in the Samarkand region is also very strong (r = +0.964, r² ≈ 0.929, or 92.9% of the variance explained), although it is slightly weaker than that observed for births. The absolute number of deaths increased in parallel with population growth; however, relative mortality rates (per 1,000 inhabitants) have declined in recent years, reflecting improvements in living conditions and healthcare quality. The demographic dynamics of the region are characterised by exceptionally strong interrelationships between key indicators, with r values ranging from +0.964 to +0.998. This implies that urban development planning should take into account ongoing natural population increases, rising settlement density, and the need to maintain balanced linkages between urban and rural areas.

These relationships suggest a clear causal chain: sustained high fertility — rooted in the region's young age structure and traditional reproductive norms — generates continuous natural population growth, which in turn increases the absolute size of the urban population without requiring large-scale rural-to-urban migration. Rising settlement density follows as a consequence, as population growth is concentrated within existing urban boundaries, rather than triggering territorial expansion.

The interaction between social factors and urban population growth plays an important role in understanding contemporary urbanisation processes. This subsection examines how the social characteristics of the region are associated with changes in the size of the urban population. The main social indicators used in the analysis are presented in Table 4.

Source: Compiled by the Authors Based on Statistical Data.

The strongest relationships were observed between urban population size and indicators such as the volume of services, retail trade turnover, the size of the working-age population, and retail turnover per capita. This suggests that urban growth is accompanied by the active expansion of the service sector, trade and consumer activity, as well as by the concentration of the economically active population.

Figure 6 presents a correlation matrix illustrating the relationships between urban population size and the principal social indicators in the Samarkand region. This indicates that most social characteristics are closely associated with urban population growth, pointing to the integrated and multifaceted nature of urbanisation processes.

A relatively weaker, yet statistically significant, correlation was observed between the number of higher education graduates and level of urbanisation (r ≈ +0.75), reflecting the growing importance of educational potential under urban population expansion conditions. The relationship between urban population size and employment remains positive, although somewhat less pronounced (r ≈ +0.88), which may indicate more complex and heterogeneous processes in the labour market.

The results demonstrate that growth in the urban population of the region is closely linked to socio-economic development and to increases in the quality and intensity of urban life. The correlation matrix demonstrates a strong relationship between the level of urbanisation and key socio-economic indicators. In particular, the growth of the urban population is closely associated with an increase in retail trade turnover (r ≈ +0.90), indicating a direct link between urban expansion and the development of the consumer market. The correlation analysis indicates that urbanisation processes in the Samarkand region are closely associated with key socio-economic indicators. A strong positive correlation was observed between per capita retail trade turnover and employment levels (r ≈ +0.92), confirming the role of cities as the main centres of economic activity in the region. Urban population size is strongly correlated with the number of people of working age (r ≈ +0.95) and with employment levels (r ≈ +0.88), reflecting the concentration of labour resources in urban areas. Population growth in the cities of the Samarkand region is positively correlated with increasing housing demand (r ≈ +0.86).

From a cause-and-effect perspective, the direction of these relationships is largely bidirectional. Urban population growth generates demand for retail goods, services and labour, thereby stimulating the expansion of the service sector and the working-age population concentration in cities. At the same time, a growing service economy and rising per capita incomes attract further population inflows and sustain the natural increase, creating a self-reinforcing cycle of urban demographic and economic growth.

The correlation matrix presented in Figure 7 demonstrates that the growth of the urban population in the Samarkand region is closely associated with the expansion of industrial production, increased investment and construction activity, together with the development of the service sector and small-scale entrepreneurship. Correlation analysis revealed stable and statistically significant relationships between the size of the urban population and key economic indicators, confirming the pronounced city-centred nature of regional development.

Figure7. Correlation Matrix Between City Population and Economic Factors.

The strongest relationship was identified between the size of the urban population and industrial output, with a correlation coefficient of r = 0.91. This indicates that urban population growth is primarily accompanied by the expansion of industrial activity in manufacturing sectors oriented towards urban labour markets and infrastructure. A strong positive correlation was also observed between urban population size and the volume of services provided (r = 0.80). This reflects the service-oriented nature of urbanisation and the growing demand for trade, transport, social and household services as the urban population increases. The relationship between urban population size and investment in fixed capital is characterised by a high correlation coefficient (r = 0.88), demonstrating the concentration of investment flows in urban areas. Investments are mainly directed towards industry, construction and infrastructure, thereby strengthening the role of cities as centres of economic growth. Urban population growth is closely linked to an increase in construction activity (r = 0.89). This indicates a direct response by the construction sector to rising demand for housing, production facilities and infrastructure, particularly within cities and their suburban areas. The positive correlation between urban population size and the number of small business entities (r = 0.89) highlights cities as favourable environments for the development of small-scale entrepreneurship. The concentration of population and consumer demand stimulates entrepreneurial activity and contributes to the expansion of the service sector. At the same time, a stable inverse correlation was identified between urban population size and the area of agricultural land (r = –0.69), reflecting structural changes in land use under conditions of urbanisation, whereby agricultural land is gradually replaced by residential development, industrial zones and infrastructure facilities, especially in suburban areas.

Overall, the values of the correlation coefficients indicate that urbanisation in the Samarkand region is accompanied by the concentration of industrial production, investment, construction, services and small businesses in urban areas, while simultaneously leading to a reduction in agricultural land use. This process signifies a profound structural transformation of the regional economy, in which cities are emerging as the primary centres of economic activity and spatial development. The economic indicators used in the study and the corresponding statistical data are presented in Table 5 for the period 2010–2024, as complete data for 2025 were not available at the time of the research.

Note: Data on investments in fixed capital and agricultural land area for 2025 were not available from official sources at the time the study was conducted. Remaining 2025 indicators are reported in Tables 3 and 4.

The table outlines the key characteristics of regional economic development, which are employed to analyse the relationship between urban population size and economic factors over the study period. The economic development of the Samarkand region during the 2010–2024 period is characterised by growth in industry and the service sector, increased investment and construction activity, together with a gradual transformation of land use. These trends demonstrate a close relationship between urban population growth and the region's key economic factors.

The causal structure underlying these correlations reflects the dual role of urbanisation as both a driver and an outcome of economic concentration. Investment in fixed capital and industrial expansion is partly attracted to cities because of their existing labour supply and infrastructure, making urbanisation a precondition for economic activity. At the same time, economic growth generates employment and income, which sustains population growth and accelerates urbanisation. The inverse relationship with agricultural land (r = −0.69) represents an irreversible structural consequence: as urban and peri-urban areas expand, agricultural land conversion follows, particularly in districts adjacent to Samarkand city.

GIS-based analysis shows that urbanisation in the Samarkand region is spatially uneven. The highest levels are observed in central and south-eastern districts, especially around Samarkand city and its neighbouring territories. These areas have higher population density, stronger transport accessibility, more developed services, and greater economic activity. This spatial concentration reflects several reinforcing factors: the historical role of Samarkand city as an administrative, cultural and economic centre; superior transport connectivity along the Zarafshan valley corridor; and the agglomeration of educational, healthcare and industrial infrastructure, which attracts both labour and investment flows to the core districts. In contrast, northern and western districts remain more rural and agriculture-oriented. These areas have lower urban population shares, weaker industrial development, and stronger dependence on agricultural land use. This spatial pattern confirms the presence of a centre–periphery structure in the region. The lower fertility rates and negative migration balances observed in these two districts suggest that population outflow is driven by limited local employment opportunities and the gravitational pull of Samarkand city's labour market, rather than by purely demographic factors.

The spatial distribution of urbanisation suggests that Samarkand city functions as the dominant regional growth pole. Surrounding districts are increasingly influenced by suburbanisation, infrastructure development, labour-market integration, and land-use transformation. Peripheral districts, however, remain less integrated into the urban economic system. The spatial pattern also directly reflects the correlation results reported in Section 4.1: the inverse relationship between urbanisation and agricultural land area (r = −0.69) is most pronounced in peri-urban districts surrounding Samarkand city, where residential and industrial expansion has systematically displaced agricultural land since 2010, while peripheral districts have maintained larger agricultural areas precisely because urban expansion pressure there remains limited.