1 Introduction

Effective rural spatial planning and development policies necessitate a robust comprehension of the rural context, along with a precise categorization of rurality based on key defining attributes (Agarwal et al., 2009; Argent, 2008; Beynon et al., 2016; Cloke, 1977; Gallent & Robinson, 2011; Li et al., 2015; Waldorf, 2006). Defining rurality and assessing its levels within rural development offer numerous benefits including the mitigation of spatial and social marginalization, facilitation of equitable socio-spatial advancement in rural regions, and enhancement of living standards for marginalized populations (Humphreys, 1998; Li et al., 2015; Ocana Riola & Sanchez Cantalejo, 2005; Watt & Sheldon, 1993). The dynamics of rural living are intricately tied to concerns of community sustainability, well-being, equity, and accessibility to services (Nelson et al., 2021). As contemporary lifestyles and expectations evolve, these challenges will increasingly hold greater significance for rural communities (Nelson et al., 2021). While these challenges are not exclusive to rural areas, there has been ongoing debate about how the underlying factors of these issues manifest distinctively in rural settings compared to the others (Berenguer et al., 2005; Ellis & Biggs, 2001). Rural development and rural regions were previously inextricably linked to non-urbanization and agriculture, but this no longer adequately describes today’s complicated reality (Isserman, 2005; Li et al., 2015) instigated by the changing climate and growing socio-spatial complexity (Rahman et al., 2023). Scrutiny through economic lens educes that rural region faces heightened disaster vulnerability due to limited socio-economic diversity, overly dependence on functions that are highly susceptible to climate change impacts and hazards (Bonfiglio et al., 2021; Freshwater, 2015). Characterizing such areas as rural and the extent of rurality are necessary to devise fit-to-context strategies and action plans to abate the loss and damages incurred by climate change and disastrous events. Academics have therefore emphasized the necessity of refining rural definitions to enhance policy focus and development objectives (Agarwal et al., 2009; Argent, 2008; Beynon et al., 2016; Gallent & Robinson, 2011). The inadequate comprehension of rurality in rural development leads to missed opportunities in effectively leveraging targeted policies that could otherwise capitalize on understanding dynamics and fostering a sense of identity (Li et al., 2015). Gaining comprehensive recognition and deeper understanding into the characteristics of diverse rural areas can offer vital insights for rural developer or planners, serving as valuable references to reshape the structure of rural policies.

Due to globalization impacts and consequential transformation experienced in Bangladesh, the rural areas are increasingly becoming subservient to urban priorities, as government policies are focusing urban centers more and more. Since the independence, rural Bangladesh has been experiencing a significant outflow of working age (15 to 64 years) population due to constrained job prospects, low income, and inadequate socio-cultural amenities, resulting in critical social and demographic concerns (Khan, 1982) including increased dependency and mortality rates, declining fertility, further livelihood vulnerability, and discouragement of public and private sectors investments in rural regions. Such issues also exacerbate regional imbalance, which is detrimental to the national development (Khan, 1982). Effectively addressing these concerns is essential for achieving balanced development and promoting sustainability, which requires operational concepts of rurality and objective measurements to elucidate functional variations within rural areas (Yang & Li, 2020). At policy level, knowledge regarding degree of rurality allows for inclusive planning and addressing spatial inequalities well. A concrete definition of rurality is therefore instrumental to generate effective rural development policies and planning that address well-being and sustainability concerns of rural communities. A distinct framework for assessing degree of rurality and categorizing rural areas is imperative for planners and policy makers to apprehend rural dynamics. Thus, this article intends to evaluate degree of rurality of the Southwestern region in Bangladesh from functional perspective.

The concept of rural or rurality has been theorized differently by scholars from different perspectives. Being rural as opposed to urban is an attribute that people easily attach to a place based on their own perceptions, which include low population density, remoteness from urban areas, extensive landscape, lower degree of physical infrastructures and limited public amenities and utilities etc. (Casey et al., 2001; Waldorf, 2006; Wallace et al., 2010; Zhang et al., 2000). However, researchers and policymakers must embrace an empirical understanding of rurality, departing from conventional definitions. Otherwise, the outcomes of research, policies, and strategies rooted in these notions may not attain their intended objectives (Agarwal et al., 2009; Argent, 2008; Gallent & Robinson, 2011; Isserman, 2005; Waldorf, 2006). From a demographic viewpoint, differentiating rurality from urbanity is traditionally drawn from absolute population size, faces criticism for its limitations to apprehend rural dynamics properly. Alternatively, population density has been recommended for a more precise assessment of rurality (Harrington & O’Donoghue, 1998; Hugo et al., 2003; Ocana Riola & Sanchez Cantalejo, 2005). However, there exists no definitive threshold for either population density or population size that can distinctly demarcate rural from urban areas (Beynon et al., 2016; Martin et al., 2000). Furthermore, this perception of rurality overlooks various dimensions encompassing socio-cultural, economic, spatial, and political aspects. This raises concerns about the reliability of population-based rural definitions (Hugo et al., 2003; Martin et al., 2000; Waldorf, 2006). In the 1940s, Redfield (1941) introduced the concept of a rural–urban continuum, suggesting rural area’s transition into urban ones through gradual spatiotemporal changes and socio-economic functions, indicating a lack of distinct boundaries and instead a gradient connecting rural and urban (Entrena-Durán, 1998; Hewitt, 1989; Ocana Riola & Sanchez Cantalejo, 2005). The rural–urban continuum theory recognizes the multidimensionality of rurality, proposing a conceptual framework where rural and urban represent opposite points on a continuum that defines spatial variations within a country or region (Hewitt, 1989; Ocana Riola & Sanchez Cantalejo, 2005; Sanz, 1994; Waldorf, 2006). From a political-economic standpoint, rurality is characterized by unique practices favoring private and voluntary initiatives over government interventions, distinct production systems, and disparities in delivering public amenities and utilities (Redfield, 1941). Alternatively, functionality perspective of rurality defines rural areas based on agrarian land-use mix, low-density small-scale infrastructure with widespread landscape and homogenous sociocultural identity (ibid). The demarcation between rurality and urbanity has become more complicated due to shared socio-economic attributes and political frameworks between rural and urban areas. The discrepancies in conceptualizing rurality underscore the crucial need for in-depth research to enhance methodological rigor, facilitating the establishment of a dependable typology for characterizing rural areas.

Rural studies have frequently highlighted the necessity for conducting dedicated research on rural areas, driven by the understanding that the rural context differs inherently from the urban context (Osmani & Sen, 2011; Redfield, 1941; Watkins & Champion, 1991; Yang & Li, 2020). Given this context, defining and measuring rurality has emerged as a specific focal point for researchers, yielding significant contributions over the past four decades (Beynon et al., 2016). For instance, Redfield's rural–urban continuum model (Redfield, 1941) offered a conceptual groundwork for constructing a rurality index, although its development predates the contemporary times. Rurality indices prioritize discerning locales through rural functionality, moving beyond a binary division between rural and urban (Ocana Riola & Sanchez Cantalejo, 2005). The presence of conflicting and diverse theoretical viewpoints regarding rurality poses epistemological challenges in shaping rurality indices (Kaneko et al., 2021; Waldorf, 2006). The first rurality index was developed in England by the Department of the Environment for investigating rural areas and small towns of England and Wales using only three variables (Harrington & O’Donoghue, 1998). However, Clokes’ rurality index (Cloke, 1977) was a significant and innovative contribution to the measures of rurality (Beynon et al., 2016; Galluzzo, 2019). Utilizing principal component analysis (PCA), this quantitative measurement employed a quartile classification system, spanning from 'extreme' rural to 'extreme non-rural' categories (Cloke, 1977; Cloke & Edwards, 1986; Harrington & O’Donoghue, 1998). Employing PCA to construct the rurality index offers a key advantage by quantifying the distinct influence of each indicator on the level of rurality, allowing researchers to proceed their studies without anticipating the outcomes (Beynon et al., 2016; Harrington & O’Donoghue, 1998). However, scholars have affirmed the inherent methodological drawbacks of indexing approach in measuring the extent of rurality (Beynon et al., 2016; Cloke, 1977; Cloke & Edwards, 1986; Harrington & O’Donoghue, 1998). Although indexing method is technically sophisticated, selection of rurality indicators in the analysis is itself a subjective approach and thence, the characterization of rural areas is influenced by the adopted variables (Harrington & O’Donoghue, 1998). Furthermore, rurality indices offer a relative, rather than absolute characterization of investigated area (Blunden et al., 1998; Waldorf, 2006). Although expecting an absolute measurement of rurality is irrational without a theoretical foundation of absolute rurality. A school of thought has further criticized the indexing approach arguing that indexing methods fail to explain more of critical variation that exists in the data (Beynon et al., 2016). For example, an area with significantly low population density can have a higher physical infrastructure, non-agricultural socio-economy. There are no cut-off values proposed in the rurality indexing approach for dealing with such issues. Notwithstanding these limitations, the indexing approach has emerged as a reliable application for investigating rurality and measuring degree of rurality more consistently. Notable number of studies (see for example, Banister, 1980; Best, 1981; Beynon et al., 2016; Harrington & O’Donoghue, 1998; Moseley, 1979; Ocana Riola & Sanchez Cantalejo, 2005; Pacione, 1984; Phillips & Williams, 1984) have used rurality indexing methodology and have advanced it further to increase applicability and reliability.

Assessing the extent of rurality and characterizing rural areas based on that holds practical significance in formulating fit-to-context development strategies, addressing rural–urban disparity, distributing resources and services effectively, devising evidence-based rural development plans and policies. Despite such significance, a little has been done for measuring rurality and characterizing rural areas in Bangladesh. This study is thus intended to contribute in this segment of knowledge to enrich rurality literature. Besides, rural areas in Bangladesh is often defined arbitrarily, raising complexities regarding the status of an area that impede critical decision-making processes such as regional development initiatives, rural infrastructure development programs, rural welfare projects etc. To address these challenges, insights derived from this research can be useful for defining rurality and classifying the region according to the extent of rurality or non-rurality. Unlike the traditional approach, this study has considered multidimensional functions to conceptualize rurality, acknowledging the distinctiveness of rural areas (Beynon et al., 2016; Coombes & Raybould, 2001; Harrington & O’Donoghue, 1998; Waldorf, 2006). Furthermore, beyond employing a single-factor model of PCA, this research employs a multi-factor model to formulate the rurality index, as this approach enhances precision and dependability of the index (ibid). This study offers valuable insights into varying levels of rurality across the Unions in the southwestern region of Bangladesh, which will be instrumental for policy makers and local development authorities to generate evidence-based rural development plans and actions at the local level. The findings also provide initial glimpse of the rural to urban transitional status of different locales of the investigated area. Additionally, this article prompts scholars to consider the value of dedicated exploration into various aspects of rurality and underscores the necessity for advancing rurality indices.

2 Materials and methods

2.1 Study area

We have selected the southwestern region of Bangladesh—Khulna, Satkhira, and Bagerhat districts—as the case study areas (see Fig. 1) for this research due to their unique socio-economic traits, elevated climate change vulnerability, and significant potential for swift socio-economic transformation in the coming decades. This potential is driven by anticipated impacts from major infrastructure projects, including the Padma Bridge, Ruppur Nuclear Power Plant, and the proposed Southwest Bangladesh Economic Corridor (SWBEC). We anticipate a series of pivotal development initiatives and investment projects aimed at the region in the coming years. This underscores the need for clear demarcation between rural and urban areas based on empirical studies and operational definition, as such status impact decision-making, investment categories, and project types. The coastal region is located in between 21°60´ and 24°13´ north latitudes and in between 88°34´ and 89°58´ east longitudes (BBS, 2011a, 2011b, 2011c). The world's largest mangrove forest, the Sundarbans, spans the southern portions of Satkhira, Khulna, and Bagerhat districts within this region. Within the three case areas, Satkhira exhibits the highest household size, followed by Khulna and Bagerhat districts (BBS, 2011a, 2011b, 2011c). Population trends since 2000 reveal a decline in total population for Khulna and Bagerhat districts, with annual growth rates of -0.25% and -0.47% respectively. In contrast, Satkhira district has witnessed an increase in population, with an annual growth rate of 0.62% (ibid). However, Khulna district has undergone more pronounced urbanization (33.54%) compared to Satkhira (9.95%) and Bagerhat (13.23%) (ibid). According to BBS (2011a, 2011b, 2011c), Khulna district records 58.6% of housing as non-structured, with 3.6% of households lacking toilet facilities. In contrast, Bagerhat and Satkhira districts exhibit 83.1% and 57.2% non-structured housing, respectively. Khulna's economy is anchored in agriculture, supplemented by its dependence on the Sundarbans and Mongla port. Notably, around 41.31% of Khulna's households are engaged in agricultural activities (BBS, 2011a). Bagerhat and Satkhira exhibit comparable socio-economic circumstances, with approximately 68.75% and 57.78% of households engaged in farming, respectively (BBS, 2011b, 2011c). Khulna stands out among the districts as the most developed district, displaying greater potential for rapid advancement and the capacity to drive progress in the southwestern region of the country.

Fig. 1
figure 1

Map of the study area. Source: Developed by the authors utilizing spatial data derived from the GIS unit of Local Government and Engineering Department (LGED) of the government of Bangladesh (2021)

Full size image

2.2 Developing a framework for measuring rurality

Our study followed four key stages: (i) conceptualizing the issue and identifying research gaps, (ii) choosing a suitable scientific methodology, (iii) gathering and processing relevant data for analysis, and (iv) summarizing and analyzing the collected data (see Fig. 2).

Fig. 2
figure 2

Schematic diagram depicting the steps followed in this research

Full size image

Drawing from an extensive literature review, we have selected twenty indicators to assess the level of rurality in the Southwestern region of Bangladesh. Table 1 displays the chosen variables and their functional relationship with rurality. It is crucial to acknowledge that the selection of indicators in the rurality indexing approach is inherently subjective, raising questions about accuracy and validity. However, expecting an entirely objective method for indicator selection for the indexing is unrealistic, given the absence of an absolute concept of rurality. We conducted the study using our refined version of Cloke's rurality index methodology. Our enhancements include a slightly expanded set of indicators, utilization of a PCA n-factor model (n = 6) as opposed to Cloke's one-factor model (Cloke, 1977; Cloke & Edwards, 1986), and enhanced interpretability of the final outcomes. Furthermore, the selection of the approach has also taken into account the availability of suitable indicators and corresponding data.

Table 1 Dimensions and indicators for measuring rurality index
Full size table

2.3 Data collection and processing

We have considered 'Union' level (i.e., the smallest strata of local government of Bangladesh) as the unit of analysis for measuring degree of rurality and characterizing rural areas based on rurality score. It's worth noting that the spatial data for Bangladesh's administrative boundaries, sourced from the LGED, encompasses Wards within municipalities and Unions at administrative boundary level-4. We have utilized census data sourced from the Bangladesh Population and Housing Census 2011 (community reports), published by the BBS. It's important to mention that the raw data from BBS underwent refinement and processing using data manipulation techniques within MS Excel and IBM SPSS applications. Subsequently, we have conducted PCA analysis, rurality indexing, and GIS-based data visualization using this dataset. We obtained the geospatial database of Bangladesh's administrative boundaries from LGED and customized it as required. This database was utilized to map rurality levels, facilitated by GIS techniques.

2.4 Generating indices and summarizing outcomes

We have adopted an indexing methodology to understand degree of rurality. At the beginning of the process, all selected indicator data for rurality measurement have been normalized through the application of normalization index, using the Eq. (1) (for positive functional relationship) and Eq. (2) (for negative functional relationship). These equations were initially formulated for HDI (Human Development Index) to assess life expectancy and subsequently gained widespread adoption for computing other indices like LVI (Livelihood Vulnerability Index) (Antwi-Agyei et al., 2013; Hahn et al., 2009), VI (Vulnerability Index) (Ha-Mim et al., 2022), and RI (Resilience Index) (Asmamaw et al., 2019). Normalization indexing is essential in this context to standardize indicators with varying measurement scales into a unified index for meaningful analysis.

$$Index\;{X}_{s}= \frac{{X}_{s}-{X}_{min}}{{X}_{max}-{X}_{min}}$$
(1)
$$Index\;{X}_{S}= \frac{{X}_{max }-{X}_{s}}{{X}_{max}-{X}_{min}}$$
(2)

Here, Index Xs is the normalized index value and Xs is the original value of the indicator X for Union S, Xmin and Xmax are the minimum and maximum values of the indicator respectively.

In the next step, we have conducted a PCA using the twenty selected indicators, employing oblique rotation (oblimin with Kaiser normalization). This analysis generated loading scores for each variable (refer to Table 2) utilized as their respective weight in the rurality index.

Table 2 Loadings of the indicators produced from the six-factor model of PCA
Full size table

The PCA analysis has been verified as statistically valid through the Kaiser–Meyer–Olkin (KMO) and Bartlett’s test (Antwi-Agyei et al., 2013; Hahn et al., 2009). The sampling adequacy for PCA analysis has been verified through the KMO measure (0.568), surpassing the accepted threshold. Furthermore, the Bartlett's test of sphericity (p < 0.05) indicates that the correlations between the variables were statistically significant (see Table 3). The average communality of the PCA (0.71) was also found to be above the recommended minimum of 0.500.

Table 3 KMO and Bartlett's Test of the PCA
Full size table

Figure 3 illustrates the extraction of six components in the PCA based on the Kaiser criterion (Eigenvalues greater than 1) and the 'point of inflection' as per Tabachnick and Fidell (Asmamaw et al., 2019), so defined as the six-factor model (Beynon et al., 2016). These six components collectively explain 70.91% of the total variance in the underlying variables used in this study. It is crucial to highlight that Cloke (1977) used one-factor model (i.e., considered only the first factor) in his rurality index methodology. If we follow the same approach, the first factor would explain only 29.48% of the total variability of the indicators included in this analysis. Hence, we adopted a six-factor model for the rurality index, enhancing measurement accuracy and reliability compared to Cloke's one-factor model. In this model, each variable of the analysis is loaded onto the factor it is most associated with (based on largest loading value). It's worth noting that in his rurality index methodology, Cloke employed loading scores from a one-factor PCA model as weights for the indicators in his study. He did not standardize the loadings, resulting in indicators with negative loadings closely aligning with rural characteristics in his research (Cloke, 1977). In contrast, we standardized the obtained variable loadings using a normalization index following Eq. (2) for enhanced interpretability. This approach is consistent with findings in scholarly literature (Beynon et al., 2016; Harrington & O’Donoghue, 1998). After normalizing the indicators and calculating their corresponding normalized loadings, the rurality index for each Union within the three case study districts was computed using Eq. (3).

$${I}_{j}={a}_{1}{X}_{1}j+{a}_{2}{X}_{2}j+\cdots +{a}_{20}{X}_{20}j$$
(3)

where, \({I}_{j}\) = rurality index for union j

Fig. 3
figure 3

Scree plot showing the numbers of components retained in the PCA

Full size image

\({a}_{n}\) = normalized loading score corresponding to variable \({X}_{n}\)

\(X_nj\) = normalized value of variable Xn for union j

Thus, the normalized value of each indicator in every Union is multiplied by their respective normalized loadings. The resulting values are then aggregated to determine the rurality score for each Union in the study area. The Unions of the individual districts are later categorized into four clusters such as Extremely Rural, Intermediate Rural, Intermediate Non-Rural and Extremely Non-Rural, following an equal interval or quartile classification method (Beynon et al., 2016; Cloke, 1977; Ocana Riola & Sanchez Cantalejo, 2005). Besides, the average rurality index score for all the three districts has been calculated for characterizing them based on the relative rurality index score. Finally, the outputs derived from this rurality index have been linked with their corresponding spatial data sets in the GIS interface (i.e., ArcGIS Pro) with a view to develop rurality map characterizing rural areas into four clusters within the study area as mentioned above.

3 Results

3.1 Degree of rurality and characterization of rural areas

Figure 4 depicts the study generated relative rurality index score of the case study sites (i.e., Satkhira, Khulna, and Bagerhat). The figure illustrates that Satkhira district exhibits a greater degree of rurality (RI = 4.23) in comparison to Khulna (RI = 3.58) and Bagerhat (RI = 4.14) districts. The Unions across the three districts are classified intro four classes based on their respective rurality index score following a quartile classification method: extremely non-rural (RI <  = 2.662), intermediate non-rural (2.662 < RI <  = 3.538), intermediate rural (3.538 < RI <  = 4.414), and extremely rural (RI > 4.414). This functional characterization offers a distinct insight into the rurality level of each Union within the districts. The map depicts a highly rural southern region of the study area, encompassing around 32.45% of all Unions. Conversely, the northeastern and northwestern parts exhibit a lower degree of rural functionality indicating more urban characteristics. Only 9.8% of the Unions, including Bagerhat municipality, Satkhira municipality, and Khulna City Corporation, are registered as extremely non-rural. About 12.1% of Unions are categorized as intermediate non-rural, while the majority (45.65%) fall under the intermediate rural classification. Table 4 shows the percentile distribution the rurality clusters across the three districts. We have rendered the status of each Union across the study area in terms of the degree of rurality through Table 5.

Fig. 4
figure 4

Rurality map of the study area

Full size image
Table 4 Percentile distribution the rurality clusters across the three districts
Full size table
Table 5 Rurality status of individual Unions across the study area
Full size table

3.2 Dimensional contribution to rurality

Figure 5 portrays the respective contributions to each examined dimension of rurality, offering a holistic insight into how various sectors influence rural–urban functionality. The study's findings suggest that the education sector has limited influence on defining rurality levels in the investigated area, indicating comparable education quality and efficiency between urban and rural areas. This suggests that either urban areas lack advanced educational facilities or rural areas are provided with quality education services. It elicits that the socio-economic dimension is the dominant sector driving rural–urban disparities in the southwestern region, contributing 59.09% to the rurality index score. This suggests that socio-economic factors play a crucial role in defining rurality levels in the area. The findings indicate that demographic characteristics are the second most significant aspect shaping rural–urban disparities in the coastal region, accounting for 21.75% of the overall rurality extent. The research found that the infrastructural dimension, encompassing housing, water, and sanitation facilities, is a crucial determinant of rurality levels in the study area, contributing 16.57% to the overall rurality level. Nevertheless, the figure might not encompass all pertinent aspects of rurality crucial for policymaking, warranting potential further research to delve into supplementary dimensions.

Fig. 5
figure 5

Contribution of the dimensions to the level of rurality

Full size image

4 Discussion

This study informs that Satkhira district exhibits the highest level of rurality, closely followed by Bagerhat, whereas Khulna demonstrates the lowest degree of rurality. The data uncovers signs of spatial inequality among the districts, and delving into spatial inequality could provide insight into the reasons behind varying degrees of rurality among them. The analysis uncovers disparities within the non-rural (urban) areas of the districts, akin to those observed in the rural areas. The relative rurality index score indicates that non-rural areas in Khulna district exhibit relatively lower rural functions and a higher degree of urban functionality compared to Bagerhat and Satkhira districts (see Table 6). Almost a similar scenario can be seen for rural areas too. Table 6 elicits that the rural areas in Satkhira district are characterized with comparably higher degree of rurality than that of Khulna and Bagerhat districts. An intriguing observation emerges: rural areas in Bagerhat district display slightly lesser rurality than those in Khulna district, despite Khulna district exhibiting the lowest rurality level among the three case study areas. This phenomenon is influenced by factors like higher female literacy rates, female employment in the service sector, and improved access to sanitation facilities in Bagerhat's rural areas compared to those in Khulna and Satkhira. Another point can be approximated from the analysis that rural–urban disparity within Khulna district is significantly higher (see Table 6). This signifies that resource allocation, infrastructure advancement, employment prospects, public utility access, gender parity, educational provisions etc. are disproportionately concentrated in urban areas, potentially resulting in a significant developmental imbalance across the district.

Table 6 Rural–urban disparity among the districts based on rurality index
Full size table

An evident trend identified in the study reveals the concentration of urban functions along major rivers, historically pivotal in the urbanization process (Phong, 2015). For instance, Mongla Port Paurashava and Chalna Paurashava situated on the banks of the Pashur River, with various other urban areas in the three districts similarly evolving alongside such waterways. Hence, key industrial, commercial, and service sectors in Bagerhat district, including Mongla sea port, Mongla EPZ, cement factories, banks, power stations, and industrial complexes, have thrived along the Pashur River's banks. Khulna district's industrialization stemmed from the growth of Khulna Shipyard and industries such as steel, rubber, cement, seafood, and jute along the Rupsha River. Infrastructure development alongside major rivers fosters employment, drawing people, investors, and government attention. This transformation shifts agrarian economies into urbanized one with increased non-agricultural employment, literacy rates, population density, and female empowerment. In contrast, the extremely rural areas in these three districts feature agriculture-focused employment, loss of labor force, less robust housing, restricted sanitation and water access, along with elevated proportions dependent populations. Intermediate areas represent a transitional phase between rural and urban, where the attributes of rurality are gradually yielding to urbanization, transforming these areas from rural to urban.

An unforeseen concern emerged in Bagerhat district when a disparity surfaced between the statistical data provided by BBS for Sarankhola Range and the actual spatiotemporal characteristics of the Union. Sarankhola Range lies within the forested expanse of the Sundarbans mangrove forest, sparsely populated. However, official data from BBS presents urban traits, including non-agricultural employment, tube well water access, and elevated literacy rates. This overestimation incorrectly classifies Sarankhola Range as an extremely non-rural area rather than an extremely rural one. It's worth noting that the loadings of certain variables used to measure rurality do not align with the assumed (based on extensive literature review) functional relationship with rurality. To resolve this concern, we applied a normalized scoring method for the loadings, as suggested by Beynon et al. (2016) and Harrington and O’Donoghue (1998), in calculating the rurality index.

Although globalization has shifted the balance in favor of urban areas, rural development remains crucial. Sustainable rural development is instrumental for achieving balanced growth and greater sustainability in the southwestern region. Socio-economic transformation of an area is imperative to address increasing spatiotemporal uncertainties and climate change-induced risks (Rahman et al., 2023). Ensuring active public participation in decision-making and development processes is critical to harness such transformative potential of communities. Requiring community consensus for decisions concerning rural areas is a robust strategy to ensure the inclusion of residents in the development process. It will also enhance local government autonomy, accountability, and transparency, promoting good governance in the region. Implementing an ‘economic diversification’ policy can mitigate economic vulnerability in rural areas by integrating compatible sectors (for example, promoting pisciculture, apiculture with farming through a smart management system; forestry and haor-based tourism; farming and food processing industries). Complementing these policies, incentivizing diverse income sources for rural residents, involving unemployed youth and students in freelancing and rural innovation through specialized training can effectively diversify rural economies. Furthermore, this approach can mitigate rural-to-urban migration by retaining the working-age population in rural regions. Establishing cottage industries in rural areas with appropriate institutional arrangements can empower women and stimulate local economic growth effectively. Moreover, these industries offer valuable export opportunities. Pastoral farming is another promising option for promoting rural economic growth, particularly in providing alternative employment for women. Creating a direct cooperative partnership between the private sector and local communities, facilitated by local governments, involves communities supplying raw materials (e.g., milk, meat, poultry, eggs) to industries in exchange for financial and technical support ─ can effectively direct investments to rural areas and create employment opportunities. Incorporating local government here is essential to safeguard against potential exploitation of local communities by capitalist industries. These policies and plans, once realized, will lead to the improvement of infrastructure and education sectors, thereby contributing to the overall development of rural areas. To address uncertainties like pandemics (e.g., COVID-19) and natural disasters (e.g., floods, cyclones), establishing a local government-managed ‘uncertainty and climate change fund’ (Rahman et al., 2023) for community support in times of crisis can be considered. However, implementing these strategies hinges on central and local governments reshaping their rural development priorities and agendas to create a conducive environment. The governments must recognize the importance of rural areas for national growth and establish well-defined development agenda for them.

5 Conclusion

This research has assessed the degree of rurality of the southwestern region of Bangladesh using a rurality indexing approach from on functional perspective. The study has also categorized the region into four clusters (i.e., extremely rural, intermediate rural, intermediate non-rural, and extremely non-rural) using a quartile classification method based on the rural index scores, measured at Union level. The study has utilized primarily population census data deriving from the Bangladesh Bureau of Statistics (BBS). Local Government Engineering Department (LGED) has provided the required geospatial datasets.

The results indicate that Satkhira district has the highest degree of rurality (RI = 4.23), followed by Bagerhat (RI = 4.14) and Khulna (RI = 3.58) districts. The study reveals a highly rural southern part (32.45% of Unions) and lower rural functionality in the northeastern and northwestern sections, with 9.8% of Unions being extremely non-rural, 12.1% intermediate non-rural, and 45.65% intermediate rural. Sectoral analysis highlights socio-economic factors as the primary drivers of rural–urban disparities in the southwestern region, followed by demographic characteristics and infrastructure. Additionally, the study indicates that education plays a limited role in shaping rurality levels in the area. Analysis reveals that Khulna district has notably higher rural–urban disparities compared to the other two. We have also observed a concentration of higher non-rurality along major rivers like Rupsha, Mongla, and Bhairab, suggesting a potential link between geographic features and rurality.

This research will aid rural planners and policymakers in comprehending the area's unique rural dynamics, crucial for crafting context-specific development strategies. Moreover, precise delineation of administrative units, like Unions and Wards, is vital for effective implementation of evidence-based development initiatives. Additionally, the article offers an adaptable methodology for assessing rurality and defining rural areas in diverse contexts. The study highlights the necessity for additional research on rural–urban disparities in southwestern Bangladesh, considering factors like spatiality, political dynamics, and cultural aspects. This article calls for a robust methodology in rurality research, incorporating geographical, political, adaptation, and vulnerability factors alongside other dimensions.