1 Introduction

The COVID-19 pandemic has underscored the significance of quality of life (QoL) in one’s place of residence. Just prior to and during lockdowns, many countries grappled with the phenomenon of reverse migration (FAO 2021). QoL considerations and subsequent population movements have brought the future of urban and rural areas, along with their sustainability, to the forefront of the debate within the European Union (EU 2021; Le Roy and Ottaviani 2021). Examining the factors contributing to life satisfaction in urban and rural areas plays a pivotal role in formulating tailored location-specific policies to mitigate QoL disparities among regions and counteract the negative repercussions of urban density.

Individuals often ‘vote with their feet’ (Tiebout 1956), meaning they choose to reside in places where the benefits outweigh the costs. Changes in personal or external circumstances can prompt individuals to consider relocating. However, when the costs of relocation become prohibitively high, individuals may express their dissatisfaction through voting behaviour and a surge in populist sentiment (Koeppen et al. 2021).

According to statistical data from the Organisation for Economic Co-operation and Development (OECD 2020), approximately half of the world’s population prefers to live in large or medium-sized cities, while the remaining half is evenly divided between rural and semi-dense/intermediate areas. Based on these location choices, one would expect individuals living in big and medium-sized cities to, on average, report higher levels of happiness and greater life satisfaction. However, studies have identified the emergence of the so-called ‘urban paradox’ (Morrison 2021), which refers to the lower life satisfaction reported by individuals living in urban areas compared to their rural counterparts. Agglomeration diseconomies, such as crime, congestion, pollution, and other factors, are likely significant contributors to this discrepancy in self-reported life satisfaction.

In Europe, several studies have presented evidence supporting the urban–rural divide (Sørensen 2014; Lenzi and Perucca 2016a, b, c; Lenzi and Perucca 2019). In this line of research, it is crucial to consider the heterogeneity both within and between areas (Brereton et al. 2008). Simply residing in a rural area may not be the sole determinant of higher life satisfaction within a country. Various other factors, including economic disparities between affluent and less affluent regions, the accessibility and quality of transportation and digital infrastructure, and the geographical remoteness or centrality of regions, all significantly shape overall quality of life and life satisfaction. A similar consideration applies to amenities, whether natural or human-made. Other factors include socioeconomic features, social relationships, and institutional and infrastructure provisions (Fantechi et al. 2020; Faggian et al. 2018a, 2018b; Brereton et al. 2011).

Recent research on the urban–rural dichotomy emphasises the importance of investigating the association between life satisfaction and location-specific characteristics (Lenzi and Perucca 2016a). Until recently, studies addressing the urban–rural paradox have often overlooked systematically considering the role of amenities and social interactions (Meijers et al. 2016; Sørensen 2014; Lenzi and Perucca 2021; Hand 2020; Hoogerbrugge et al. 2022). In contrast, the urban and regional economics literature on quality of life in cities, which emerged in the 1970s and 1980s, recognised the influence of amenities (for a literature review, see Lambiri et al. 2007). Seminal works by Rosen (1979) and Roback (1982) employed hedonic models to estimate implicit values of amenities across US cities, providing a ranking of cities based on the revealed ‘price’ of quality of life. Recent contributions by scholars like Albouy (2016) and Albouy et al. (2016) have revitalised the study of the relationship between amenities and urban well-being. Furthermore, the incorporation of amenities into interregional migration studies can be traced back to the pioneering work of Graves (1980), and subsequent research has recognised the pivotal role of human-made amenities in attracting high-skilled individuals to cities (Roback 1982; Glaeser et al. 2016).

Another body of literature on the determinants of life satisfaction underscores the role of individual characteristics, external conditions, the fulfilment of human needs (Sirgy and Cornwell 2002), opportunities, accessibility, and social relationships, particularly in relation to amenities (Costanza et al. 2007).

Within the context of the urban–rural debate, this study contributes to two key areas. Firstly, it aims to bridge the gap between urban studies, which emphasise the significance of amenities in quality of life, and research on life satisfaction within the urban–rural debate. By doing so, it establishes connections between these two bodies of work and provides valuable insights into the relationship between amenities and life satisfaction in urban and rural contexts. Secondly, the study employs the capability approach developed by Sen and adopts a comprehensive definition of ‘well-being’ that incorporates both the availability (i.e., capability) and accessibility (i.e., functioning) of amenities and social interactions. While the availability of amenities is essential, their accessibility (i.e., individuals’ actual possibilities to experience them) is even more critical. Opportunities vary based on individual constraints, such as free time, age, health status, family commitments, employment, and income status, as well as external conditions, such as the actual presence of and ease of access to infrastructure and amenities. Therefore, the presence of a park at a reasonable distance from one’s home, without the real possibility of enjoying it, might have a similar effect on life satisfaction as not having such a park at all.

By integrating these perspectives and drawing upon insights from various streams of literature, this study presents a comprehensive theoretical framework that encompasses the availability of amenities, social interactions, and their functioning, which refers to their actual accessibility. This approach offers a lens to explore the intricate dynamics between amenities and well-being in urban and rural contexts. Furthermore, the gradient of life satisfaction between urban and rural areas may also depend on the strength of the relationships between these areas (Lenzi and Perucca 2021). Indeed, following the concept of borrowed size proposed by Alonso (1973), one can expect higher levels of life satisfaction in rural areas with good networks to urban centres compared to rural areas without such connections. Similarly, life satisfaction in urban areas that are well-connected to rural areas is likely to be higher than in urban areas that are less connected to rural areas.

Empirically, this paper aims to investigate three key aspects: (1) confirming the positive impact of amenities on life satisfaction, irrespective of the urban–rural context; (2) assessing the influence of accessibility to amenities and social relations on the disparity in urban–rural life satisfaction; and (3) examining whether residing in a rural area alone guarantees a higher likelihood of experiencing greater satisfaction.

The analysis focuses on the case of Italian regions, characterised by notable heterogeneity primarily attributable to the enduring North–South divide (González 2011; Capello 2016; Cannari et al. 2019). A period of regional convergence marked the early 1970s, coinciding with the first oil crisis; however, since then, inequalities across Italian regions have been on the rise (Gagliardi and Percoco 2011). Additionally, Italy faces with complex issues of depopulation, population aging, and out-migration from rural areas (Reynaud et al. 2020). During the decade spanning from 2001 to 2011, nearly 40% of Italian municipalities experienced a decline in their population, with the phenomenon being particularly pronounced in the southern regions. Furthermore, additional data suggests that, at least during the period from 1991 to 2001, depopulation was a prevalent demographic trend, predominantly affecting small towns and rural villages (Reynaud et al. 2020).

The study utilizes individual data from a multipurpose household survey called ‘Aspects of Daily Life,’ conducted by the Italian Institute of Statistics (ISTAT). The questionnaire investigates the socioeconomic factors affecting individual life and other factors related to the area of residence. The dataset covers the years 2013–2018 and includes 204,892 observations stratified by gender and age at the regional level (NUTS 2). The life satisfaction question used is as follows: ‘Generally speaking, how satisfied are you with your life?’ The econometric strategy employs ordinal regression models, taking into account the ordinal nature of the dependent variable (i.e., a five-point Likert scale), and treating the differences between the levels as rankings.

Regarding the measures of urbanity or rurality, there is no universally accepted definition of urban–rural classification. To classify regions as either urban or rural, this paper adopts the categorisation of territorial units provided by the Organisation for Economic Co-operation and Development (OECD). This categorisation allows for an assessment of the urban–rural composition of regions by comprehensively evaluating provinces (NUTS 3) within each region (NUTS 2). The classification considers two measures: first, a count of provinces classified as urban, rural, or intermediate within each region, and second, an analysis of the share of the population residing in provinces categorised as urban, rural, or intermediate within each region. The main analysis employs the initial urban–rural indicators, while the robustness check section considers the latter measure. Results confirm that people in more urbanised regions report lower life satisfaction. Furthermore, the presence of and accessibility to amenities significantly influence reporting higher satisfaction. Findings suggest that living in rural areas alone does not guarantee higher life satisfaction per se, revealing a gradient of life satisfaction in favour of economically richer rural areas with higher quantitative and qualitative levels of infrastructural services.

The remainder of the article is organised into sections. Section 2 discusses the literature on the urban–rural dichotomy. Section 3 presents the theoretical model, whilst Sect. 4 illustrates the data, the empirical model, and the estimation strategy. Section 5 discusses the results, and Sect. 6 offers the main conclusions of the research.

2 Literature Review

The well-being of individuals within their living environment has gained significant attention in academic literature. How individuals experience life in cities is crucial not only for their own well-being but also for the vitality and sustainability of urban areas. Enhancing well-being in cities is essential for attracting new residents and retaining existing ones. The literature on individual well-being in urban settings encompasses various research streams. Of relevance to this paper is the urban and regional economics literature that emerged in the 1970s and 1980s (Lambiri et al. 2007) that recognize the impact of amenities on the overall quality of life in cities. Notably, the seminal works of Rosen (1979) and Roback (1982) employ hedonic models using wage and rent data to estimate the implicit values of amenities across different US cities, thus providing a ranking of cities based on the “price” of quality of life. A fundamental concept underlying the research of these scholars is the notion of equilibrium in different urban locations. According to their perspective, wages and rents serve as compensatory mechanisms that account for the presence of both amenities and disamenities in each area. This research stream has experienced a recent resurgence in academic studies, thanks to the contributions of scholars like Albouy (2016) and Albouy et al. (2016), among others. Their work has revitalized the exploration of the relationship between amenities and urban well-being. The incorporation of amenities into interregional migration studies can be traced back to the pioneering work of Graves (1980), who introduced the concept of pure amenities and their influence on migration patterns. This line of research, along with the literature on urban growth, has increasingly recognized the pivotal role of certain human-made amenities in attracting high-skills individuals to cities (Roback 1982; Glaeser et al. 2016). Scholars have broadened their investigation to encompass various human-produced amenities, including public services and cultural establishments (Blomquist et al. 1988; Gyourko and Tracy 1991; Glaeser et al. 2001; Florida 2002; Boschma and Fritsch 2009).

Overall, the majority of individuals prefer to reside in locations where economic growth and living standards are higher (Glaeser et al. 2016). Thus, it follows that life satisfaction should be higher in denser places. However, population density is commonly associated with agglomeration diseconomies, negative externalities, and lower levels of satisfaction. Indeed, an additional strand of literature examining the well-being associated with different living environments, particularly in terms of life satisfaction, has identified a “paradox” whereby individuals residing in cities tend to report lower levels of satisfaction compared to their rural counterparts (Morrison 2011, 2021). In this stream of research, a recent work by Lenzi and Perucca (2021) disentangles the effect of urbanisation on life satisfaction, finding a positive effect of the proximity to higher rank cities in the urban hierarchy in 21 European member countries. The authors explain the findings using the ‘borrowed size’ effect à la Alonso (1973), for which the proximity to large urban centres generates positive externalities for closer areas (Meijers et al. 2016). Indeed, the spatial structure of the cities within regions affects life satisfaction: in polycentric cities of North-West Europe life satisfaction is lower respect to the centralised counterparts and respect to rural areas (Hoogerbrugge et al. 2022).

Furthermore, this literature explains the urban–rural paradox in different ways. A possible explanation refers to income in urban areas that do not adequately compensate for the presence of disamenities (Okulicz-Kozaryn 2015; Sørensen 2014; Hand 2020). Another explanation is that life satisfaction (and happiness) decreases due to socioeconomic inequalities that are usually higher in urban areas (Ballas 2013; Helliwell et al. 2020). Other studies point out the presence of lower social capital in cities due to a lower sense of belonging and community feeling (Sørensen 2012; Helliwell et al. 2020). Or the need to live in cities to match education and skill with jobs when the urban settlements are not the preferred location (Sørensen 2014; Okulicz-Kozaryn 2015). Similarly, a selective migration of “less satisfied individuals” could exist in cities (Hoogerbrugge and Burger 2021). According to recent literature, life satisfaction in urban places depends on frustration about unmet personal aspirations (Hanell 2022). As Lenzi and Perucca (2020, p.22) explain, the urban–rural debate “hides a deeper level of complexity”. Table 1 reports the main research published in the last two decades on urban–rural life satisfaction at the regional level. Some studies have focused on Europe (Sørensen 2014; Lenzi and Perucca 2016a, b, c; Lenzi and Perucca 2019; Hanell 2022), the US (Berry and Okulicz-Kozaryn 2011), China (Knight and Gunatilaka 2010; Clark et al. 2019; Dang et al. 2020) and Kazakhstan (Kalyuzhnova and Kambhampati 2008). Overall, previous findings support the urban–rural dichotomy and confirm the paradox for developed economies but not for Eastern economies. In the case of Romania, studies have reported that life satisfaction relates to better economic conditions and opportunities provided by cities (Lenzi and Perucca 2016b, 2016c). Meanwhile, some studies on Europe (Gerdtham and Johannesson 2001; Shucksmith et al. 2009) and Kazakhstan (Kalyuzhnova and Kambhampati 2008) did not find a substantial difference in urban–rural satisfaction. Lenzi and Perruca (2019) analysed the spatial differences in life satisfaction in Italy from 1980 to 2012 and confirmed the negative effect of urbanisation. Furthermore, the effect they found was not so robust and the effect was found only for a subsample of years (1991–2002) and regions (1980–1991 for Northern regions and 1991–2002 for Southern regions).

Table 1 Review of studies on the urban–rural paradox in life satisfaction and happiness.
Full size table

This literature review highlights as in urban settlements of more developed countries, lower levels of life satisfaction have been recognized as a stylized fact. Furthermore, a more in-depth analysis reveals a lack of a standard definition of urban–rural classification that is substantially data- and country-driven. This issue is not necessarily a drawback, as geographic and demographic characteristics vary across countries. Moreover, many works define ‘well-being’ using the concepts of ‘happiness’ and ‘life satisfaction’ interchangeably. However, as Sørensen (2014) explains, the two terms do not have the same meaning: ‘happiness’ reflects a contingent and short-term emotional status, whilst ‘life satisfaction’ refers to a broader and long-term spectrum of factors influencing personal well-being. In other words, life satisfaction is the ‘degree to which an individual judges the overall quality of life as-a-whole favourably’ (Veenhoven 1991: 10), whilst happiness is a measure of people’s moods and emotions (Glatzer 1991).

The existing research on the urban–rural dichotomy emphasizes the significance of investigating the association between life satisfaction and location-specific characteristics (Lenzi and Perucca 2016a). However, studies addressing the urban–rural paradox and taking into account amenities (Meijers et al. 2016; Lenzi and Perucca 2021), and social interaction (Sørensen 2014; Hand 2020; Hoogerbrugge and Burger 2021) do not explicitly differentiate between the availability and frequency of use.

Another stream of the literature on life satisfaction not related to the urban–rural debate nor to the amenity-related literature, pointed out the role of individual characteristics, external conditions, and the fulfillment of human needs (Sirgy and Cornwell 2002). A theoretical study conducted by Costanza et al. (2007) highlight the importance of opportunities, the availability and accessibility of various natural and human-made amenities, and the presence of social relations.

Respect to the reviewed literature, the current study makes contributions in two key areas. Firstly, it aims to bridge the gap between urban studies focusing on the significance of amenities in quality of life and the research on life satisfaction within the urban–rural debate. By doing so, it connects these two bodies of work and provides valuable insights into the relationship between amenities, social relations, and life satisfaction in urban and rural contexts. Secondly, the study introduces the capability approach developed by Sen and a comprehensive definition of ‘well-being’ that incorporates both the availability (i.e. capability) and accessibility (i.e. functioning) of amenities and social interactions. As far as the author’s knowledge is concerned, in urban quality of life literature, only a prior study (Biagi et al. 2018) has explicitly employed Sen’s approach to analysing urban quality of life.

By integrating these perspectives and incorporating insights from the three aforementioned streams of literature, the study presents a comprehensive theoretical framework that includes the availability of amenities, social interactions, and opportunities for individuals to access and derive benefits from these elements. In the following section, utilizing the Sen framework, the underlying mechanism is elucidated, which links amenities and social relations to regional characteristics while considering the distinctive urban–rural factors.

3 The Underlying Mechanism of Life Satisfaction in the Place of Living

Following Veenhoven (1991: 10), life satisfaction can be defined as the ‘degree to which an individual judges the overall quality of life as-a-whole favourably’. When modelling life satisfaction related to the place of living, the analysis becomes more complicated because it implies including external elements. The main question then becomes, ‘By which mechanism does the external environment affect this judgement?’. Life satisfaction is shaped by a multitude of factors, including both individual characteristics and external conditions (Sirgy and Cornwell 2002). Moreover, the theoretical studies conducted by Costanza et al. (2007) underscore the significance of opportunities and the fulfilment of human needs, as well as the availability and accessibility of diverse natural and human-made amenities, alongside social relations. Together, these elements contribute to the overall individual assessment of life satisfaction. The capability approach developed by Sen provides a theoretical framework for integrating the notions of opportunities and the satisfaction of needs into empirical analyses of life satisfaction within urban and rural contexts. Building upon the capability approach and the prior conceptual work of Costanza et al. (2007), as well as its operationalization for an Italian city by Biagi et al. (2018), this paper explicitly distinguishes between the quantity (capability) and accessibility (functioning) of amenities and social relations within an individual’s place of living. The capability approach emphasizes the importance of considering people’s abilities and opportunities to achieve desired outcomes. In the context of this research, it entails examining not only the availability of various amenities and social relations but also the extent to which individuals can effectively access and utilize them in their everyday lives. By differentiating between quantity and accessibility, this analysis acknowledges that mere availability of amenities is not sufficient to determine their impact on life satisfaction. Instead, it recognizes that the ability to access and fully engage with these resources plays a crucial role in shaping individuals’ life satisfaction. This approach allows for a more nuanced understanding of the relationship between the urban environment, the availability of amenities and social relations, and their influence on life satisfaction. The presence of amenities measures capability, whilst the ‘frequency of use’ or ‘time spent’ measures the functioning, which refers to their actual accessibility. Figure 1 visually illustrates the domains that influence life satisfaction in the place of living, as indicated in the literature (for a detailed literature review Lambiri et al. 2007; Costanza et al. 2007; Biagi et al. 2018). These domains encompass personal characteristics such as for instance gender and work (Della Giusta et al. 2011), age (Alesina et al. 2004; Dolan et al. 2008), health status (Anand et al. 2011) as well as external factors such as the availability of amenities, both natural and human-made (Sirgy and Cornwell 2002; Costanza et al. 2007), and social relations (Costanza et al. 2007; Dolan et al. 2008; Sørensen 2012; Helliwell et al. 2020; Hand 2020; Hoogerbrugge et al. 2022). While the presence of these factors is crucial for reporting high life satisfaction, it is not solely sufficient. The arrows in Fig. 1 indicate that opportunities to enjoy these amenities and social connections are equally important for fostering life satisfaction.

Fig. 1
figure 1

Source: author’s elaboration

The underlying mechanism of life satisfaction in the place of living.

Full size image

The average level of individual life satisfaction is additionally influenced by the specific characteristics of the urban or rural settlement in which it is situated (for a review see Lenzi and Perucca 2020). The previous section highlights that existing literature suggests that lower life satisfaction in urban settlements can be attributed to various factors, such as the absence of income compensation for the presence of disamenities, high levels of inequality, limited social capital and sense of belonging, a mismatch between education and job opportunities, and the selective migration of less satisfied individuals. This paper asserts that the dynamic interaction between individuals and their living environment holds substantial influence over life satisfaction. Specifically, when examining the disparities in life satisfaction between urban and rural contexts, it is crucial to acknowledge the roles of both amenities and social interaction. However, in order to gain a deeper understanding of the impact of these drivers on individual life satisfaction, it is imperative to consider the presence of amenities as well as the frequency of their utilization as a measure of accessibility. For example, the presence of an easily accessible park within a reasonable walking or transportation distance would have a greater positive impact on life satisfaction compared to simply being aware of its existence without the practical ability to reach it due to limited connections or restricted opening hours. This principle also applies to other public and private amenities available in different urban and rural settings. In urban areas, individuals often experience lower life satisfaction due to the challenges they face in accessing and enjoying amenities and social relations. These challenges also encompass issues such as congestion, limited accessibility, and time constraints. In selecting the place of living, rational agents trade off the benefits and costs of each location and choose the optimum location based on their personal situations and constraints. The burden of relocating from cities due to dissatisfaction with urban living can often be prohibitively high, trapping individuals in their current locations, resulting in low levels of satisfaction (Sørensen 2014; Okulicz-Kozaryn 2015). While rural areas offer abundant natural amenities and opportunities to enjoy them, they still face challenges such as limited access to health services, education, infrastructure, and employment opportunities. However, as also highlighted by very recent works (Lenzi and Perucca 2020; 2022), the same apply to rural place; it is expected that rural areas situated in more prosperous regions with good infrastructure will demonstrate higher levels of life satisfaction compared to those located in economically disadvantaged areas. In rural settlements, individuals are supposed to enjoy more natural amenities and lower disamenities (e.g. crime, pollution, noise and congestion). However, they are also supposed to experience a lower level of human-made amenities (e.g. cinema, theatres, museums, public services and infrastructures). Residing in rural areas that lack adequate provision of human-made amenities, such as healthcare services, educational facilities, and transportation infrastructure, does not contribute to higher individual satisfaction. Conversely, when these rural settlements are in close proximity and/or well-connected to well-equipped urban areas, it is likely that life satisfaction would be higher. Therefore, the relationship between life satisfaction and the urban–rural gradient could also be influenced by the interconnectedness between different areas (as in Lenzi and Perucca 2021). By integrating the concept of borrowed size, introduced by Alonso (1973), rural areas situated in wealthier regions with well-developed infrastructures, which facilitate strong networks and connections with urban centers, have the potential to experience the positive impacts of human-made amenities while mitigating the burdens associated with disamenities (Okulicz-Kozaryn 2015; Sørensen 2014; Hand 2020). As a result, these regions are anticipated to exhibit higher levels of life satisfaction in comparison to geographically peripheral and economically disadvantaged rural areas.

According to the proposed theoretical framework, Eq. 1 formalizes the general model, outlining the key components that influence an individual’s life satisfaction i in place j at time t.

$$ \begin{aligned} Life \,Satisfaction_{ijt} & = f_{ijt} \left( Personal\, Characteristics_{it} , Amenities_{jt} ,\right.\\ & \quad \left. Social\, Relations_{ijt} ,Regional\, Characteristics_{jt} \right) + \varepsilon_{ij}\end{aligned} $$
(1)

In the empirical application, each domain will incorporate measures encompassing both the quantity (capability) and accessibility (functioning) aspects. The following section explains the data and the variables employed to control for each type of component in the empirical model.

4 Data, Empirical Model and Estimation Strategy

4.1 Data

The analysis is centered on the case of Italian regions, which serves as a compelling study due to its notable heterogeneity, primarily attributed to the enduring North–South divide (González 2011; Capello 2016; Cannari et al. 2019). Moreover, Italy confronts the challenges of depopulation, population aging, and out-migration from rural areas (Reynaud et al. 2020).

The data used in this study came from a multipurpose survey on households called ‘Aspects of Daily Life’ conducted by the ISTAT. This survey is administered annually and is stratified by gender and age at the regional level (NUTS 2). The survey offers information on the everyday lives of residents. In particular, the questionnaire investigates individual socioeconomic aspects affecting life satisfaction and other aspects related to the place of living. The final dataset refers to 2013–2018 and includes 204,892 individuals aged 18 years and older. The question on life satisfaction is presented as follows: ‘Generally speaking, how satisfied are you with your life?’ The response options are organised on a Likert scale ranging from 0 to 10, where 0 = ‘not at all satisfied’ and 10 = ‘very satisfied’. Considering the frequency and the normal distribution of the responses, the answers were recoded into five categories, as shown in Fig. 2: 1—very low satisfaction (0–5 in the original coding, 17.05%), 2—low satisfaction (6 in the original coding, 17.73%), 3—medium satisfaction (7 in the original coding, 25.09%), 4—high satisfaction (8 in the original coding, 25.16%) and 5—very high satisfaction (9 and 10 in the original coding, 14.96%).

Fig. 2
figure 2

Source: author’s elaboration from Survey Aspects of Daily Life of the Italian Institute of Statistic (ISTAT)

Distribution of life satisfaction.

Full size image

Two measures were used to obtain the first hints about possible differences in life satisfaction according to the respondents’ urban–rural places of living: the share of urban counties (provinces; share_p_urban) and the share of rural counties (share_p_rural). The methodology used to build the indicators follows the OECD definition of territorial units wherein Italian provinces corresponding to TL3 (NUTS3 regions in EU classification) are categorised as follows: ‘predominantly urban (PU)’ if the share of population living in rural local units is below 15%, combined with the existence of urban centres where at least one-quarter of the regional population resides; ‘predominantly rural (PR)’ if the share of population living in rural local units is higher than 50%; and ‘intermediate (I)’ if the share of population living in rural local units is between 15 and 50%. Following this classification, a figure indicates how many provinces in the region are categorised as PU, PR and I. Therefore, each indicator goes from 0 to 1, where 0 denotes a region in which none of the provinces is rural/urban/intermediate, and 1 represents a region with rural/urban/intermediate provinces.

To ensure the robustness of the indicators used, supplementary calculations have been introduced to account for the spatial distribution of the population within each NUTS2. To this purpose, additional measures have been incorporated considering the share of the population living in PU/PR/I NUTS3 (i.e., Share_PU, Share_PR, Share_I) regions within each NUTS2. These indicators provide a further measurement of urban agglomeration economies and test the reliability and validity of the analysis.

The majority of Italian regions have a certain degree of urbanity or rurality; for instance, the intermediate regions are 3 out of 20. The first column in Table 2 shows the ranking of regions by life satisfaction (as the average of the years 2013–2018). The fourth column presents the macro areas to which each region belongs. The fifth and sixth columns indicate the shares of PU and PR provinces in each region, respectively. Given that this work aims to investigate the differences in the urban/rural gradient in life satisfaction, the table only shows these two indicators. It is worth noting that when the sum of the two indicators is lower than 1, the complement to 1 represents provinces classified as intermediate. Marche, Umbria and Valle d’Aosta are thus identified as the intermediate regions; for this reason, the variables report a zero value for both urban and rural classification.

Table 2 Ranking of life satisfaction by regions (2013–2018), macro area and urban–rural shares.
Full size table

Individuals residing in regions situated in the northern part of Italy (emphasized in bold within the table), including Trentino-Alto Adige, Valle d’Aosta, and Lombardia, manifest notably higher levels of life satisfaction that surpass the national average (3.03). It is worth noting that a significant proportion of these regions mainly consist of urban provinces. Interestingly, the southern regions, characterized by lower mean life satisfaction compared to the national average, exhibit a mix of mainly rural provinces (such as in the case of Calabria and Basilicata) and mainly urbanized provinces (such as in the case of Campania, Puglia, and Sicilia). Sardegna stands as the sole exception, where the majority of provinces are rural, yet the mean average of life satisfaction approximates the national average. Thus, the gradient of life satisfaction does not seem to reflect the urban–rural dichotomy, but rather the spatial differences in economic performances. Indeed, after plotting the average income level and the average life satisfaction in the regions, as shown in Fig. 3, it can be clearly observed a positive relation between the two variables. The observed trend reveals that individuals residing in the affluent northern regions, characterized by their relative prosperity, such as Trentino Alto-Adige, Lombardia, and Emilia-Romagna, consistently report higher levels of life satisfaction compared to their counterparts inhabiting the economically disadvantaged southern regions, including Campania, Calabria, and Sicilia. This observation provides a first hint of a relation between socioeconomic disparities across Italian regions and the subjective well-being of their residents. The higher life satisfaction reported by individuals in the wealthier northern regions could be attributed to various factors, including greater access to economic opportunities, higher quality public services, improved infrastructure, and overall better living conditions. Conversely, individuals in the southern regions, facing persistent economic challenges and structural limitations, may experience lower life satisfaction due to limited resources, higher unemployment rates, and less favorable socioeconomic conditions.

Fig. 3
figure 3

Source: author’s elaboration from the Aspects of Daily Life survey of the Italian Institute of Statistic (ISTAT)

Average life satisfaction and average income by region (2013–2018).

Full size image

4.2 The Empirical Model and Estimation Strategy

Equation 2 shows the empirical model, whose variables are explained in depth in Table 3.

$$ \begin{aligned} & Life\, Satisfaction_{ijt} \\ & = \beta_{1} Male_{it} + \beta_{2} Age_{it} + \beta_{3} Children_{it} + \beta_{4} Employed_{it} + \beta_{5} Education_{it} \\ & + \,\beta_{6} No\, Chronic\, Disease_{it} + \beta_{7} Low\, Health\, Limitations_{ it } + \beta_{8} Foreign\, Citizenship_{it} \\ & + \beta_{9} Working\, Hours_{it} + \beta_{10} Family\, Hours_{it} + \,\beta_{11} Green\, Amenities_{itj} + \beta_{12} Green\, Zero_{itj} \\ & + \beta_{13} No\, Recreation_{itj } + \beta_{14} No\, Restaurants_{itj} + \,\beta_{15} Friends\, zero_{itj} + { }\beta_{16} Friends\, once_{itj} \\ & + \,\beta_{17} Income_{tj} + \beta_{18} North_{j} + \beta_{19} South_{j} + \beta_{20} Second\, Tier_{j} + \beta_{21} Second\, Tier_{j} *Income \\ & + \beta_{22} PU_{j} + \beta_{23} PR_{j} + \beta_{24} PR\_North_{j} + \beta_{25} PR\_South_{j} + \varepsilon_{ij} \\ \end{aligned} $$
(2)
Table 3 Descriptions of variables.
Full size table

The variables from male to family hours control for personal characteristics; amongst them, working hours and family hours represent the individual-related functionings that might affect the time dedicated to other activities. The square of age and working hours and family hours were added for further check. Green Amenities measures the presence of natural amenities within walking distance (15 min), whilst green zero represents the functioning. As the survey does not ask for recreation activities and restaurants in the place of living, the empirical model controls only for the related functioning: no recreation and no restaurants. The survey also allows for social relations to control only for the associated functionings: friends zero and friends once. Amongst the regional characteristics, apart from the urban–rural dichotomy controlled by the indicators PU and PR, the model includes the average per capita income (Income), the macro region of the place of living (dummies North and South with the reference category the Centre), and a variable named Second Tier that measures the presence of the capital city (Rome) or a second-tier city (Milan, Naples and Turin) in the region. This last variable follows Lenzi and Perucca (2019). To further account for urban–rural effects, the analysis incorporates a series of interaction dummies. Specifically, the variable Second_tier_Income considers the regional per capita income within a NUTS2 region hosting either the capital city (Rome) or a second-tier city (Milan, Naples, Turin). By including this interaction term, we control for the urban paradox, wherein highly urbanized regions often exhibit lower levels of reported satisfaction.

To disentangle the complexities of life satisfaction in rural areas, the empirical model includes additional controls, PR_North and PR_South. The variable PR_North captures the proportion of rural provinces (TL3/NUTS3 regions) within the northern region (TL2/NUTS2) where the respondent resides. Similarly, the variable PR_South represents the share of rural provinces (TL3/NUTS3 regions) within the southern region (TL2/NUTS2) where the respondent resides. It is crucial to recognize that rural areas, especially in the case of Italy, are characterized by significant heterogeneity, primarily due to the enduring North–South divide (Capello 2016; Cannari et al. 2019). As a result, life satisfaction in rural areas is likely to vary substantially based on their location within wealthier and well-connected regions, as well as more economically disadvantaged and peripheral regions. Consequently, while rural areas may exhibit higher life satisfaction compared to their urban counterparts in the former scenario, the same positive trend may not be observed in the latter scenario.

Given the characteristics of the dependent variable, the empirical strategy implies the use of a model for categorical variables. Grilli and Rampichini (2014) demonstrated that ordered logit is the most appropriate empirical method for studies on life satisfaction based on ordinal response variables. In the ordered logit model, the observed ordinal variable Y is a function of a continuous latent variable \({Y}^{*}\) that has various threshold points. Y depends on whether a particular threshold is crossed (Menard 2002). We compute \({Y}^{*}\) using the following equation:

$$ Y^{*} = \mathop \sum \limits_{k = 1}^{K} \beta_{k} X_{ki} + \varepsilon_{i} = Z_{i} + \varepsilon_{i} , $$
(3)

where \({Z}_{i}=E\left({Y}^{*}\right)\), and \({\varepsilon }_{i}\) is the random disturbance term. This class of models estimates the probability that the unobserved variable Y* falls within the various threshold limits and assumes that the relationship between all pairs of groups is the same (parallel regression assumption). If this assumption holds, then a single set of coefficients is estimated. Otherwise, the literature suggests using an alternative estimator: the generalised ordered logit (gologit; Williams and Quiroz 2019). The tests reported in Table 4 agree that the proportional odds assumption does not hold. In this case, the empirical literature suggests using gologit estimation, which is presented in the next section.

Table 4 Tests for the parallel regression assumption.
Full size table

5 Results

From an empirical point of view, this paper aims at: (1) confirming the positive impact of amenities on life satisfaction, regardless of the urban–rural context; (2) assessing the influence of accessibility to amenities and social relations on the disparity in urban–rural life satisfaction; and (3) examining whether residing in a rural area alone guarantees a higher likelihood of experiencing greater satisfaction. Table 5 shows the marginal effects of the gologit model. In line with previous studies, being men, educated, having children and not suffering from chronic diseases increase the probability of having higher life satisfaction (Biagi et al. 2018). Older people are less satisfied than younger individuals, thus confirming the non-linear impact of age, as reported in seminal studies (Alesina et al. 2004; Dolan et al. 2008). At the same line, increasing the time dedicated to work and family positively affects satisfaction, even though this effect declines beyond certain thresholds. Indeed, Della Giusta et al. (2011) and Biagi et al. (2018) suggested that it is the possibility of spending time with one’s children—and not having children per se—that affects life satisfaction. Furthermore, having health limitations and being foreign may also reduce satisfaction (Gerdtham and Johannesson 2001).

Table 5 Gologit model: drivers of life satisfaction.
Full size table

Recent studies on immigration in Italy signal integration issues at the socio-cultural and economic levels (Frattini and Sartori 2021; Iannelli et al. 2021). Studies on other European countries have reported similar findings (Angelini et al. 2015). For instance, a previous work in the UK revealed that individual characteristics have a higher impact than place characteristics (Ballas and Tranmer 2012). However, Koeppen et al. (2022) find that place-related features also matter. When considering amenities (either natural or human-made) and the frequency of use, the picture changes: their presence does matter in improving life satisfaction and QoL (Roșu et al. 2015; Biagi et al. 2018; Carvalho et al. 2019); however, not enjoying them at all negatively affects one’s life satisfaction (Green_zero, No_recreation, No_restaurant). Furthermore, social interactions are essential: seeing friends at least once a week makes a difference in the level of life satisfaction. Social interactions play a pivotal role in shaping individual well-being. Specifically, regular social engagement, such as meeting with friends at least once a week, has been shown to exert a discernible impact on the overall level of life satisfaction (Hoogerbrugge and Burger 2021).

Overall, the findings raise questions about the relative importance of accessibility compared to the presence of amenities in influencing life satisfaction.Footnote 1

At the same time, one study argued that income matters (Pittau et al. 2010), and indeed, individuals living in the northern part of Italy—where GDP per capita is the highest, on average—are the most satisfied. As in Lenzi and Perucca (2019), living in a region filled with second-tier cities (Milan, Turin, Rome and Naples) improves life satisfaction. However, in the wealthier metropolitan regions, the effect is negative (Second_tier_Income), indicating a sort of urban paradox. In line with previous studies, people in higher urbanised regions reported less satisfaction (PU), whilst those living in rural regions were more satisfied (PR). Looking deeper at the urban–rural dichotomy, the findings suggest that living in rural areas does not guarantee higher life satisfaction, but rather living in those located within the northern regions (PR_North). In these regions, income per capita and access to public and private amenities are higher in terms of availability and quality. Therefore, this result supports the conclusion that living in rural areas has no positive effect on life satisfaction, regardless of location.

Therefore, this work extends to the entire country the findings reported by Lenzi and Perruca (2019) only for few years (1980–1991 and 1991–2002) and macro-areas. Furthermore, these results can be linked to the borrow size concept (Alonso, 1973) according to which smaller places well connected with neighbouring major cities can experience urbanisation benefits without having the related drawbacks (Meijers et al. 2016; Lenzi and Perucca 2022; Koeppen et al. 2022).

Overall, our results confirm that individuals living in more urbanised regions report less satisfaction and that the presence of, and especially accessibility to, amenities matter in reporting higher levels of life satisfaction. However, contrary to previous studies, the findings suggest that living in rural areas does not guarantee higher life satisfaction per se; thus, a gradient of life satisfaction in rural areas emerges. This finding partially confirms the results that Lenzi and Perruca (2021, 2022) obtain studying life satisfaction in higher rank cities in the urban hierarchy showing that the higher the distance from a larger city negatively affect life satisfaction.

5.1 Robustness Check

The results in this paper are robust and consistent. Indeed, using different estimators and specifications led to stable results (see the Appendix). As seen in the previous methodological section, the proportional probability hypothesis does not hold in the case of these data, which is why gologit is applied. However, the results of the ordered logit estimator (ologit) are in line with the main results in Table 5. A second check involves the application of the ordinary least square (OLS) estimator, confirming the direction and significance of the results. A third check involves categorising the dependent variable into a dummy variable that takes the value one if respondents report a medium to high level of life satisfaction (i.e. 3 = medium satisfaction, 4 = high satisfaction and 5 = very high satisfaction) and zero otherwise (i.e. 1 = very low satisfaction and 2 = low satisfaction). This transformation allows the use of a logit estimator, which further confirms the robustness of the results.

The results shown in the previous section indicate a substantial difference in life satisfaction in rural areas between individuals living in the northern and southern regions of the country. To further verify these important findings, the sample was divided into two groups: the first group included only respondents living in the north, whilst the second included only respondents living in the south. The analysis for the South confirms the previous results: it is not going to rural areas per se that guarantees a positive and higher probability of being more satisfied, but rather the act of living in rural areas in the northern part of the country.

To ensure the robustness of the PU/PR indicators, a further check considers the share of the population living in PU/PR/I NUTS3 regions within each NUTS2. These modifications contribute to a nuanced interpretation of the findings and increase confidence in the implications for borrowed size. As Table A4 shows, the inclusion of the new measures (i.e., Share_PU, Share_PR, Share_PR_North and Share_PR_South) confirm and strengthen the validity of the findings.

6 Conclusions

People living in urban areas report having less life satisfaction than those living in rural areas in developed countries. However, data indicate that half of the world’s population prefers to live in large or medium-sized cities (OECD 2020). This indicates a paradox (Morrison 2011, 2021) and highlights a dichotomy in life satisfaction between urban and rural areas. The paper investigates the urban–rural paradox by using the capability approach à la Sen. This work specifically applies a more comprehensive definition of regional well-being by distinguishing the roles of the availability (i.e. capability) and accessibility (i.e. functioning) of amenities, as well as the connections and interactions within the place of living. This, in conjunction with regional characteristics, ultimately affects the gradient of life satisfaction. By incorporating urban–rural literature into amenity-related literature on life satisfaction and quality of life, drawing inspiration from the work of Sen, this study presents a comprehensive theoretical framework that explores the urban–rural paradox. From an empirical standpoint, this paper aims to investigate three key aspects: (1) confirming the positive impact of amenities on life satisfaction, regardless of the urban–rural context; (2) assessing the influence of accessibility to amenities and social relations on the disparity in urban–rural life satisfaction; and (3) examining whether residing in a rural area alone guarantees a higher likelihood of experiencing greater satisfaction. The empirical application involved a set of data extracted from the multipurpose survey on households called ‘Aspects of Daily Life’ for the years 2013–2018, which was conducted by the ISTAT at a regional level (NUTS2). The data set contained information on the self-reported life satisfaction of individuals living in Italy. Overall results confirm the stylised fact that people living in urban areas are less satisfied than those living in rural areas as underlined by previous empirical studies. However, the picture changes when considering life satisfaction in rural areas located in economically advantaged regions. Results highlight that it is not living in rural areas per se that guarantees a higher life satisfaction but actually living in rural areas where labour, leisure, social relations and access to amenities are more elevated in terms of availability and quality compared to other areas. As far as social relations are concerned, the finding underscores the importance of fostering social connections and maintaining a robust support network, at least once a week, as essential components of subjective well-being. The significance of maintaining regular social relations has been relatively overlooked in the existing urban–rural literature. A recent contribution by Hoogerbrugge and Burger (2021) represents a notable exception, as they incorporate the frequency of social relations into the empirical analysis. However, there remains ample room for further exploration and investigation into the underlying mechanisms that underpin the influence of social interactions on life satisfaction within urban–rural contexts.

Overall, the results align with the borrow size concept of Alonso (1973) and the related recent literature, which emphasises the role of accessibility and network externalities of the higher rank cities on life satisfaction (Meijers et al. 2016; Lenzi and Perucca 2022; Koeppen et al. 2022). Indeed, living in rural areas where opportunities are low and public and private amenities are scarce does not contribute to people’s satisfaction. Therefore, the potential rural resurgence and urban decline expected after the pandemic it is likely to occur only in specific rural areas, those located in already economically strong areas.

Consequently, the urban–rural debate should acknowledge this finding when discussing the future of urban and rural areas. For instance, a policy for peripheral rural areas would imply accessibility to physical and digital infrastructures enabling households that do not like agglomeration to have access to a standard of living and a set of possibilities comparable to those provided by cities. In the absence of such amenities, regional and local policies should strengthen the links between urban and rural areas. The various positive effects of this strategy include the reduction of discomfort related to the disamenities of cities and the increase of density in less populated areas. Therefore, to exploit the mutual benefits, it is essential to consider the possible connections amongst different places when designing sustainable development policies.

The Italian National Strategy for Inner Areas (SNAI 2014), which represents a place-based state intervention addressing population decline in low-density territories, has categorized Italian municipalities into six distinct classifications based on their provision of essential services encompassing healthcare, education, and mobility. In future developments of this study, an avenue for investigation may also involve exploring urban–rural life satisfaction within the framework of this classification, thereby accounting for the influence of service provision in the residential context.

The present work does not come without limitations. First, survey data are stratified by region and not by lower aggregation level. Related to this, finer data would have allowed for a more granular analysis. Second, when using surveys that are already available at national statistical institutes, the potential set of variables and indicators that can be used in the research analysis is limited to the original purposes of the surveys. For this reason, some controls, including those representing certain functionings, are not present in the analysis. Finally, future developments could extend the research to the European context.