The quality of life in Czech rural and urban spaces

The paper addresses the multidimensional and dual nature (objective versus subjective) of quality of life (QoL) and its strong geographical expression. Drawing on literature across sociology, psychology, medicine, economics and geography, the authors emphasize that objective QoL reflects external environmental and socio-economic conditions (e.g., health, education, environment), often operationalized through domains and indicators, while subjective QoL captures individuals’ perceptions. The concept of livability is discussed as an objective assessment of place conditions that shape QoL. Given spatial variability, the study focuses on the urban–rural continuum, noting definitional ambiguities and overlaps between rural and urban spaces due to processes like suburbanization. Traditional dichotomous delimitations (based on population and density) may oversimplify; therefore, the authors adopt a fuzzy approach to quantify municipal membership along a rural–intermediate–urban continuum. The study aims to objectively assess relationships between a municipal QoL index and degree of membership in rural/urban space, answering: (1) to what extent QoL differs across rural and urban spaces; (2) whether QoL is higher in urban or rural spaces; and (3) the role of intermediate (suburban) spaces.

The review distinguishes studies using subjective versus objective QoL measures across urban and rural contexts and methods ranging from regression models to spatial-analytic approaches. Many subjective studies report higher QoL in rural areas (e.g., Bernini & Tampieri 2017; Knight & Gunatilaka 2010; Sørensen 2014; Winters & Li 2017). Objective assessments are fewer and mixed: some find higher QoL in rural areas (Bertolini & Pagliacci 2017; Campanera & Higgins 2011), others find comparable levels (Ma et al. 2020), with results sensitive to indicator choice and territorial context. Approaches to defining rural/urban space vary: population/density thresholds (OECD, Eurostat), multivariate classifications (PCA-based), and fuzzy continuum methods (Pagliacci 2017; Pászto et al. 2015, 2016). The review motivates adopting a fuzzy urban–rural membership and an objective, indicator-based QoL index to examine spatial differentiation and local (non-stationary) relationships.

Study area: municipalities of Czechia. Data included: (1) a QoL index constructed following Murgaš & Klobučník’s (2016) Gold Standard of Quality of Life; and (2) municipal degrees of membership in rural/urban spaces from Pászto et al. (2015, 2016), representing a fuzzy rural–intermediate–urban continuum on a 0–1 scale (0 fully urban, 1 fully rural). QoL indicators (10 total) and sources/periods: (1) suicides (2014–2018, districts, CZSO); (2) male life expectancy (2014–2018, districts, CZSO); (3) female life expectancy (2014–2018, districts, CZSO); (4) mortality (2014–2018, municipalities, CZSO); (5) birth rate (2014–2018, municipalities, CZSO); (6) divorce rate (2014–2018, municipalities, CZSO); (7) population with completed tertiary education (2011, municipalities, CZSO, census); (8) unemployment rate (2014–2018, municipalities, CZSO); (9) emissions of solid pollutants, SO2, NOx, CO, VOC, NH3 (2013–2015, districts, CHMI); (10) generativity (proportion of blood donors, 2015, regions, IHIS CR). Most indicators use 5-year averages where possible; some are constrained by data availability (e.g., census year 2011 for education, 2013–2015 for emissions, 2015 for generativity). Pre-processing and normalization: The QoL index was min–max normalized to 0–10 (worst–best). Urban and rural membership variables were represented on 0–1 scales, with urban membership complementary to rural membership. Statistical analyses: Distributions tested non-normal (Kolmogorov–Smirnov), so non-parametric methods were used. Spatial autocorrelation was assessed via global Moran’s I and local Moran’s I (LISA) using k-nearest neighbors with multiple bandwidths (k = 50–1000), with α = 0.05 and False Discovery Rate correction; LISA clusters categorized as high–high, low–low, high–low, and low–high. Interrelationships were examined using Spearman’s rank correlation globally, and geographically weighted Spearman’s correlation (GWC) to assess spatial non-stationarity (tricube kernel, adaptive bandwidths k = 25–1000 neighbors; primary interpretations at k = 100 and k = 400). Correlations with absolute value >0.20 and α = 0.05 were considered meaningful. Typology: A bivariate typology combined three QoL classes (low/medium/high via natural breaks) with three space-membership classes: rather rural to rural (urban membership 0.0–0.4), intermediate (0.4–0.6), and rather urban to urban (0.6–1.0). Frequencies were computed by municipalities, population, and area.

- Global interrelationship: No significant global correlation between the QoL index and urban/rural membership (Spearman’s ρ ≈ 0.04), suggesting local/non-stationary relationships. - Spatial autocorrelation: QoL index exhibited strong positive global spatial autocorrelation (Moran’s I = 0.78, α = 0.01), indicating clustering of similar values. Urban/rural membership showed weaker clustering (Moran’s I = 0.25). LISA patterns for QoL were robust across bandwidths. Low QoL clusters occurred in the northwest (Karlovy Vary, Ústí nad Labem, Liberec regions), around Prague in a surrounding belt, and in the northeast (Moravia-Silesia). High QoL clusters included Prague and much of Hradec Králové, Pardubice, Vysočina, South Moravia, Zlín, Pilsen, and Olomouc regions. Spatial outliers often appeared at the boundaries of high/low clusters. - Geographically weighted correlation (k = 100 neighbors): Predominantly positive local correlation between QoL and urban membership in and around core regional cities (Prague, Brno, Plzeň, Olomouc, České Budějovice, Pardubice, Zlín), with more continuous positive zones in the southeast and island-like positives elsewhere. Notable negative relationships occurred around Ostrava (northeast), parts of Karlovy Vary (southwest of the region), and smaller areas in Ústí nad Labem and Liberec (northwest). Negative correlations diminished as bandwidth increased, revealing a general positive trend. Maximal absolute correlations tended not to lie at the city cores but in surrounding areas (suburbs/intermediate spaces). - Typology and population shares (Table-based highlights): The largest population group lived in high QoL and rather urban to urban space (3,246,894 people; 30.6% of population). Next were low QoL and rather urban to urban (2,204,362; 20.8%) and medium QoL and rather urban to urban (1,835,175; 17.3%). Medium QoL and rather rural to rural included 1,331,781 people (12.6%). Intermediate-space groups had the smallest populations: low QoL (~227,021; 2.1%), medium QoL (~363,833; 3.4%), high QoL (~244,589; 2.3%). By proportions within each space type, medium QoL predominated in rather urban to urban spaces (~42%) and in intermediate spaces; rather rural to rural spaces also had a majority medium QoL (over half), with high and low each around 20%. - Synthesis: Results indicate that QoL generally increases with increasing urban membership locally, yet the highest potential for high QoL often appears in intermediate (suburban) spaces combining favorable urban and rural characteristics. Exceptions correspond to structurally deprived regions (e.g., around Ostrava, parts of northwest Czechia).

The study updated an objective QoL index (2014–2018 reference) and combined it with a fuzzy urban–rural membership to examine spatial differentiation across Czech municipalities. Data constraints affected indicator temporal and spatial resolution (e.g., education from 2011 census; emissions only 2013–2015 at district level; generativity for 2015), yet the index met methodological criteria and allowed nationwide analysis. Method sensitivity analyses showed that spatial parameterization (neighborhood size) influences LISA and GWC details; the chosen k = 100 neighbors balanced local detail and interpretability, and overall spatial trends were robust to bandwidth changes. The fuzzy approach avoided a strict urban–rural dichotomy, enabling analysis of intermediate spaces that proved crucial: strongest local QoL–urban membership associations frequently occurred in suburban belts rather than city cores, and typology results pointed to the potential for high QoL in intermediate spaces. While objective indicators facilitate comprehensive spatial coverage, they omit subjective well-being. Differences with prior studies likely reflect national contexts and indicator sets. The approach is generalizable: alternative QoL indices could be substituted while maintaining the analytical framework. Continuous updating of QoL datasets would enhance longitudinal insights, which are currently limited by one-off indices in the Czech context.

The paper presents an objective, spatially explicit methodology to assess relationships between quality of life and degree of municipal membership along a rural–intermediate–urban continuum using fuzzy membership, spatial autocorrelation, geographically weighted correlation, and a bivariate typology. Findings show no significant global association, but locally QoL generally increases with urban membership, especially around city regions, with maximal associations and a high QoL potential often located in intermediate (suburban) spaces. The analysis answers: (1) QoL levels differ spatially across the urban–rural continuum; (2) on balance, higher QoL is more associated with more urban spaces locally; and (3) intermediate spaces show notable potential for high QoL. The framework is transferable to other territories and can incorporate different objective QoL constructs. Future work should: (a) update and expand indicators (including emissions and generativity at finer resolutions); (b) test alternative QoL indices and incorporate subjective measures where feasible; and (c) conduct longitudinal analyses to track shifts of high QoL potential from city cores toward suburban and peri-urban areas.

- Indicator availability and resolution: Education derived from the 2011 census (due to decennial collection); generativity (blood donors) available only for 2015; emissions at district level only for 2013–2015 (post-2016 only regional). - Mixed spatial scales (municipality, district, region) introduce potential MAUP and aggregation issues. - Objective-only QoL may not capture subjective well-being; large-scale subjective data are difficult to obtain. - Sensitivity to spatial parameterization (neighborhood size) for LISA and GWC, though main spatial trends remained robust. - Cross-sectional design limits causal inference and temporal dynamics; many QoL indices in Czechia are not regularly updated.