Deagrarisation as a Determinant of Living Standards in Rural Areas of European Union Countries

The paper deals with the issue of living standards being compared in spatial layouts across the EU countries in 2010–2019. The situation in rural areas is related to urban ones allowing for analysis of spatial inequality in terms of standard of living across the member states. The aim of the paper includes not only comparisons of standard of living, but also identification of reasons of differences between the countries and verification of a role of agriculture in shaping the rural standard of living. The EU economies are clustered based on Ward method and ranked according to aggregate measures of the relative living standards created with a usage of Hellwig method. In addition to a compound measure of the living standards, they have been assessed in two dimensions: income (ID), and housing and environment (HED). Furthermore, the study searches for an explanation of differences in the relative living standards and tests whether deagrarisation may be identified as their determinant. It uses ANOVA and estimates linear OLS regression models. It is concluded that the standard of living appears to be more favourable for rural inhabitants than for city dwellers, especially with respect to the features of surroundings, while it is decreased by the income characteristics. Furthermore, the results indicate that the relative living standards of rural inhabitants are lower in more agrarian economies. The observed deagrarisation appears as a factor positively influencing the standard of living in its income dimension.


Introduction
The very general aim included in the socio-economic development is the pursuit of a better life, which becomes a source of important changes in terms of sectoral and spatial activity. A possibility to earn a higher income induces people's educational and professional choices, as well as the intertemporal, sectoral, and spatial distribution of their labour 1 3 and financial resources usage. In search of a better place to live, people make decisions about temporary or permanent migration-international or internal: between regions or urban-rural areas. All the choices strongly influence their standard of living, which may be understood, in line with the UN definition from 1954 and the other classical attitudes, as general living conditions that allow one to meet human needs (cf. Kalinowski, 2015, p. 13;Malinowski, 2020, p. 129;Majka, 2015, p. 38).
Some territories appear to be a better place to live than others, at least when considering material criteria of assessment. They may create the possibility to earn a high income on their territory or to maintain links with spatially external sources of income for their inhabitants, simultaneously offering an attractive living environment. Such areas not only stop the outflow of indigenous peoples, but also attract new incomers, and thus are characterised by a positive net balance of the population.
Generally, living standards have grown along with socio-economic development specified by economic growth and structural changes from agrarian to industrial and post-industrial economies. In the course of industrialisation, living standards within the European Union grew essentially; however, it was revealed to be in favour of city dwellers and factory workers in comparison to inhabitants of rural areas, who were mainly engaged in agricultural production. Moreover, early industrialised countries experienced higher growth in the standard of living than more agrarian societies.
Servicisation is a modern phenomenon that brings about new trends and patterns of engagement in economic activity, social ties, and spatial development. In spatial terms, it is accompanied by a growing role of rural areas as a place to live as a result of the urban sprawl inducing suburbanisation and counter-urbanisation (Dej & Zajda, 2016), as well as gentrification of rural areas (Zwęglińska-Gałecka, 2020. The dominating servicetype economic activity engages the growing share of human resources either in cities or in rural areas. Some service industries allow for remote work, while others require direct personal contacts influencing spatial labour concentration. Spatial movements of a new service class of well-educated professionals and managers induce a process of embourgeoisement or moyennisation of rural areas (Zwęglińska-Gałecka, 2020, p. 119). As a result, in rural areas, there are observed changes in economic terms specified by a structure of income sources with a decreasing share of rural income. Moreover, on-going outflow of human resources from agricultural employment additionally decreases a role of the sector in economic activity in rural areas. Deagrarisation is commonly pointed out as a factor influencing rural income in the EU countries (cf. e.g. Czyżewski & Kryszak, 2016). Deagrarisation of rural areas is also supported by the development of economic entities involved in a service activity for both gentrifiers and indigenous people. Rural areas simultaneously experience infrastructural changes related to the development of technical and social infrastructure and a new character of land use that limits its agricultural purposes (cf. Kapusta, 2014;Stanny et al., 2016). These changes inevitably influence the standards of living for rural and urban areas inhabitants across the EU countries, creating a chance to overcome existing differences, but simultaneously creating a new line of socio-economic divide.
Concerning the new trends in spatial development in the European Union, which may strongly influence the standard of living, this paper aims at identification of relationships between the role of agriculture in an economy and relative living standards for inhabitants of rural areas. First, the living standards in rural areas relative to urban ones are specified and compared across the EU countries. Furthermore, taking into account traditional structural determinants of the rural-urban divide, it is tested whether agricultural engagement is still an essential factor influencing the relative standard of living in rural areas.
A thesis about an existence of a relationship between the standard of living in rural areas and a share of agriculture in an economy is verified. A positive direction of the relationship may be induced by the prevailing role of agriculture in employment and as a source of income for rural inhabitants. Deagrarisation in this context can bring a loss of earnings and a decrease in the living standards of rural inhabitants as compared to city dwellers.
However, modern trends specified by changes in the character of rural areas, which more and more often become a place to live for people not involved in agricultural activity, can contribute to limiting or even reversal of the assumed relationship. Rural inhabitants, often highly educated and well-qualified, engage in non-agricultural economic activity, allowing them to improve their standard of living. The population coming to rural areas usually consists of people with high wealth resources and income, and thus able to secure high living standards. Thus, deagrarisation may appear as a factor improving the standard of living for inhabitants of rural areas.
Verification of the initial thesis allows specifying whether the traditional -agrarian character of a modern village is still preserved in the EU states or its functions evolved into the creation of a space for living comfort of the non-agricultural population.
The paper is organised as follows. Section 2 delivers a literature review concerning specificity of rural areas that experience a lot of developmental changes influencing standard of living. Section 3 is of methodological character and introduces the data and methods used in the research. The results are presented in Sect. 4, where the EU countries are clustered basing on features of the relative living standards in rural areas, the aggregate measure of the relative living standards is constructed which allows to rank the EU countries and finally the relationships between living standards and the role of agriculture in an economy is verified. The last section concludes.

Processes Influencing the Standard of Living in Rural Areas in the EU States
Standard of living is usually defined as general living conditions that allow one to meet human needs (cf. Kalinowski, 2015, p. 13;Malinowski, 2020, p. 129;Majka, 2015, p. 38). It is perceived either as the utility of life, or as economic provision, or the capability to live well (Barreiro-Gen, 2019). Berbeka (2006, p. 13) proposed a fairly broad attitude towards the standard of living, defining the concept as 'the state and availability of goods and services, as well as the conditions by which the individual (the community) can satisfy their material and spiritual needs and the scope of use' (cf. Malinowski, 2020, p. 129;Klepacki & Gotowska, 2013, p. 35). Łuczak and Wysocki (2013, p. 64) understand standard of living as general living conditions of society or a socio-economic group, that cover among others labour market conditions, income, healthcare, housing conditions, infrastructure, public finances, or environmental features. Similarly, Kalinowski (2015, pp. 18-19) stresses that standard of living must be perceived as a set of indicators resulting from a level of welfare and expressed by meeting material and immaterial needs and thus creating economic possibilities and fulfilling aspirations of individuals.
In line with those definitions, in the paper standard of living is defined as overall living conditions allowing to meet human needs by means of available goods, services and actual features of surroundings. Two dimensions of the standard of living are focused on, namely: the income dimension, which specifies the possibility of possessing goods and services created by income and its distribution, and the housing and environment dimension, which 1 3 considers some features of economic, natural, and social surroundings influencing living conditions.
Research on the standard of living being a compound phenomenon is always a challenging task. It is even more difficult when one is trying to compare the living standards in different spatial layouts across countries at various stages of economic development. Thus, it is important to search for some processes that determine living standards.
Rural areas, compared to urban ones, traditionally differ concerning not only low population density but also: dispersed social and economic activity (especially agricultural), poorer competitiveness resulting from less varied social and economic structures, lower availability of goods and services (Stanny et al., 2021, p. 3). Traditionally, the countryside fulfils three basic functions (Kapusta, 2014, p. 24): residential and tourist, production and service, and ecological. However, essential changes are observed in rural areas in the course of their development with their decreasing territory, growing population, infrastructural development, and deepening diversification of economic activity and main income source (Kapusta, 2014, pp. 25-26). The specificity of the development of rural areas strongly influence the relative standard of living.
In fact, the development of rural areas is a concept often limited to agricultural development or even agricultural growth, as rural areas are commonly regarded as almost synonymous with agriculture (Stanny et al., 2016, p. 33). However, there are three distinguished approaches to understanding the development of rural areas in the European Union: first, in which rural development is identified with the modernisation of agriculture and agrifood complex; second, in which a rural development is specified by levelling differences between rural regions concerning their changing sectoral characteristics; and the last, in which a rural development is specified as the development of rural regions in general by means of using all resources available on their territories (human, physical, natural, landscape, and others) and of integrating all components and sectors at the local level (Dema et al. 2019, pp. 201-203). In a vein of the last approach, it is postulated to consider overall urban-rural development as an integral and mutually coupled process, not only limited to improvement in living standards of rural areas or to narrowing the gap between urban and rural areas (Zhu et al., 2017, pp. 3-4). In fact the policy for rural development also evaluated and was sequentially based on an exogenous model, endogenous model and then their synthesis in a form of neo-endogenous model (Naldi et al., 2021, p. 32). A broadly discussed in the literature concept of sustainable development (cf. e.g. Kalinowska et al., 2022), understood as a process in which economic, social, and environmental goals are balanced and introduced concerning the needs of both present and future generations (cf. e.g. WCED, Our Common…, 1987, http) fits in with the postulate. It is oriented towards development of such model that allows to meet socio-economic needs, while significantly reducing the threat to natural environment (Zikic, 2018, p. 143) and focuses on standard of living, economic growth, environmental integrity, and effective governance of both social and ecological systems (Fisher et al., 2021(Fisher et al., , p. 1125. In the context of sustainable development that aims to provide people with opportunities for a peaceful life (UN, The Sustain-able…, 2020; cf. Kalinowska et al., 2022, p. 2), the study directly compares the situation between rural areas and cities across the EU states and identify economies where rural areas are a better place to live.
Agricultural development traditionally shapes the incomes of rural inhabitants and, therefore, influences the standard of living. It has been observed that regions with the prevailing share of employment and income derived from agriculture in the course of development started to experience problems with unemployment, poverty, and inequality (Woźniak, 2012, p. 376). Rural areas dominated by agricultural functions experience depopulation (Rosner, 2014, p. 78). The standard of living of the socio-economic group of farmers decreased when compared to a group of employees (Turczak, 2018, p. 347). It is often explained by the phenomena of the agricultural treadmill described by Willard Cochrane (1958). Industrialisation and new technologies allowed growth in agricultural production, resulting in a decrease in prices and agricultural incomes. It induced farmers to search for new solutions to increase production, pushing them into a vicious circle. Increasing efficiency did not result in improved profitability. The agricultural surplus was distributed to non-agricultural sectors, and growing production (supported by intensive chemigation and mechanisation) negatively influenced the natural environment (Czyżewski & Kryszak, 2016, p. 28). Many researchers argue that price liberalisation tended to be harmful to agriculture and resulted in a gap between urban and rural living standards, which are still substantial in many countries (cf. Fang & Sakellariou, 2013, p. 285). However, it has been observed that in the 'new' EU states real agricultural incomes have grown in line with the growth in real production. The increasing efficiency of labour was indicated as a cause of such a phenomenon. It was stimulated by an improvement in the technical level of work equipment and by a reduction in the agricultural labour force (Czyżewski & Kryszak, 2016, p. 18;Mantey, 2013). Thus, a deagrarisation may be essential for reducing differences between rural and urban standards of living. Furthermore, a decline in urban-rural inequality may be caused by increasing intrasectoral inequality, as the offsetting trends in inequality hypothesis suggests (Fang & Sakellariou, 2013, p. 285).
As deagrarisation, understood as the long-term process of decreasing the role of agriculture in the economy and society (Musiał, 2007) and a withdrawal from the dominance of the agricultural function in the economic structure of rural areas (Stanny et al., 2016, pp. 34-35), is progressing, multifunctionality is gaining importance as an essential factor determining the development of rural areas (Woźniak, 2012, p. 374, 376). The multifunctional development of rural areas is associated with the diversity of the rural economy and abandoning agriculture as the sole or dominant function (Bąk et al., 2021, p. 15). In line with the postulate of the 'new rural paradigm' (OECD, The new…, 2006), it is based on incorporating into the rural space additional non-agricultural functions, allowing for a deeper use of dispersed territorial capital (Zrobek, 2020, pp. 567-574). As Zrobek (2020, pp. 567-574) argues, the development of innovative multifunctionality is a prerequisite to improving living conditions in rural areas.
The multifunctional development of rural areas and changes in the sectoral engagement of rural inhabitants have also been associated with migration. Nowadays, phenomena of suburbanisation or counter-urbanisation are observed supporting the development of rural areas (Bański et al., 2020, pp. 175-176).
Suburbanisation is a process of functional bonding of suburban areas with a central city. In a macroeconomic dimension, it is induced by changes in capital circulation, while in a microeconomic dimension, it has its stems in individual decisions of people searching for specific environmental values of a new place of residence (Mantey, 2013, pp. 272-273). Thus, migration from cities to rural areas is specified by functional changes of rural areas into 'city bedrooms', which become a source of supply of mobile labour resources for cities. Changes in rural areas are additionally induced by their growing recreational function and evolution toward 'consumer landscapes' (Heffmer & Klemens, 2012, p. 82).
Counter-urbanisation is defined generally as the deconcentration of a population changing settlement system of a country or region or, more strictly, as migration from areas with a high level of concentration towards those with lower concentration. Counter-urbanisation means that migrants settle in areas not being directly influenced by the cities and their agglomeration, and they lose their permanent connection with the city (Dej & Zajda, 2016, pp. 51-54).
Suburbanisation and counter-urbanisation, in parallel to seasonal interest in rural areas and seasonal housing, are perceived as two patterns of another phenomenon-rural gentrification (Zwęglińska-Gałecka, 2021, pp. 539, 546). Gentrification itself is defined as the process of saturating the social structure of the countryside with people occupying higher positions in social stratification (Hoggart, 1997) or the process of the formation of the middle class in the countryside, and thus it is also called moyennisation (Mendras, 1994) or rural embourgeoisement (Guimond & Simard, 2010) (cf. Zwęglińska-Gałecka, 20215, p. 119). A type of migration inducing rural gentrification is also known as residential migration, lifestyle migration or amenity migration (Zwęglińska-Gałecka, 2020, p. 117). It is influenced by the motives of adjustment of the place of residence to growing needs in the sphere of living conditions and quality of life (Zwęglińska-Gałecka, 2020, pp. 117-118). Its consequences may be positivethe creation of new economic entities, new jobs, an increase in income, and negative-the exclusion of certain groups of the inhabitants from rural social life or the loss of the traditional character of rural areas (Zwęglińska-Gałecka, 2021, p. 538). Rural gentrification occurs throughout the European Union with the transition to the post-industrial phase dominated by a service economy, in which the village develops its new differentiated functions out of agricultural production (Zwęglińska-Gałecka, 2021, p. 547). It has been observed since the second half of the twentieth century, first in the highly developed countries in a classic 'aggressive' form and then in Central and Eastern Europe in its 'soft' form (Zwęglińska-Gałecka, 2021, p. 536, 539;Zwęglińska-Gałecka, 2020, p. 115). Furthermore, as the motives specified by the idea of 'rural idyll' or 'proximity to the wildness' gain importance it induced a new form of 'greentrification' (Zwęglińska-Gałecka, 2020, pp. 117-118).
The processes of deagrarisation and multifunctional development, as well as gentrification specified by migration and spatial usage changes are at various advancement stages across the EU countries. Moreover, rural space is additionally internally diversified, which is reflected in numerous systematisations initiated by the urban-rural continuum model by Sorkin and Zimmerman (cf. Stanny et al., 2021, p. 3). Intra-rural inequality and increasing rural heterogeneity arise as new challenges for many countries (OECD, The new…, 2006). In the European Union, rural areas are significantly diversified, covering those that are gradually depopulated, deteriorating, remote, and peripheral, as well as those outside large municipal centers and subjected to growing urban pressure (Bórawski (ed.), 2012, p. 126). Rural areas located in proximity to larger cities rapidly develop and increase their attractiveness for investments, population settlement, and infrastructure development. On the other hand, peripheral areas still suffer all the developmental problems (Ciura, 2010, p. 177). Nevertheless, the traditional line of division into urban-rural areas is still essential in most of the EU economies.
Trends observed in the course of modern development shape new patterns of spatial diversity in terms of standard of living. Simultaneously, the existing differences between rural and urban areas induce new patterns in the behaviour of people searching for a better place to live. It seems reasonable to investigate the rural-urban heterogeneity across the EU states to explain many of the observed phenomena.

Data
Data used in the research were acquired from the Eurostat database. They describe the situation in 26 European Union countries. Malta has been excluded from the research group due to essential lacks in data. The research concerns a 10-year period that covers annual data for 2010-2019. For each indicator, a mean value for the period was calculated. Such an attitude allows to smooth the cyclical fluctuations and focus on longstanding differences between the economies.
The relative standard of living in rural areas was assessed as a compound phenomenon and described by numerous indicators. To assess living standards in possibly the deepest way, variables expressing different aspects of the income level (its median and mean value), the income distribution and its consequences (inequality, poverty, and material deprivation) were taken into account. Moreover, considering that the standard of living is also specified by non-income features, a set of indices of housing standards reflecting a quality of the living place and its closest surroundings were analysed. Based on the availability of data, the preliminary set of indices covered: • X1 (mean_inc): Mean equivalised net income in rural areas to total mean equivalised net income (Eurostat, ILC_DI17; Eurostat, ILC_DI03)-the variable compares the average income of the inhabitants of rural areas with the value for total population, independently of the income distribution, e.g. with equal weights of each unit, and gives a generalised picture of the income level; • X2 (med_inc): Median equivalised net income in rural areas to total median equivalised net income (Eurostat, ILC_DI17; Eurostat, ILC_DI03)-the variable compares the income level achieved by a half of each population and expresses economic living standards that are typical for the community and are not distorted by the external values of incomes; • X3 (ineq_m/m): Mean to median equivalised net income in rural areas to total mean to median equivalised net income ratio (Eurostat, ILC_DI17; Eurostat, ILC_DI03)the variable compares the income distribution between rural areas and the whole population and reflects income inequality occurring in each community; • X4 (pov): At-risk-of-poverty rate in rural areas to at-risk-of-poverty rate in cities (Eurostat, ILC_LI43)-the variable compares shares of population achieving incomes below 60% of the national median income, and thus focuses on the bottom tail of the income distribution; • X5 (mat_dep): Severe material deprivation rate in rural areas to severe material deprivation rate in cities (Eurostat, ILC_MDDD23)-the variable compares shares of population suffering from economic strain connected with meeting basic needs; • X6 (hous_dep): Severe housing deprivation rate in rural areas to severe housing deprivation rate in cities (Eurostat, ILC_MDHO06D)-the variable compares shares of population experiencing difficulties specified by poor housing amenities; • X7 (rooms): Average number of rooms per person in rural areas to average number of rooms per person in total (Eurostat, ILC_LVHO04D) -the variable compares the housing conditions specified by a personal space available for living that allows for undisturbed rest of an individual and thus expresses a personal housing comfort; • X8 (noise): Noise from neighbours or from the street in rural areas to noise from neighbours or from the street in cities (Eurostat, ILC_MDDW04) -the variable compares the comfort of living in rural and urban surroundings in terms of noise; • X9 (pollut): Pollution, grime, or other environmental problems in rural areas to pollution, grime, or other environmental problems in cities (Eurostat, ILC_ MDDW05)-the variable compares comfort of living in rural and urban surroundings in terms of natural environment degradation; • X10 (crime): Crime, violence, or vandalism in rural areas to crime, violence, or vandalism in cities (Eurostat, ILC_MDDW06)-the variable compares comfort of living in rural and urban surroundings in terms of disturbances in social environment.
All the variables were expressed in a relative way, comparing the feature in rural areas to the total value or to the value for cities. Therefore, the variables describe the relative situation in rural areas and are comparable across the national economies independently on their general level of development. For stimulants, i.e.: X1, X2, X7, the value exceeding 1 indicates a more favourable situation in rural areas than in cities. The higher the value, the better the relative rural situation. For destimulants, i.e.: X3, X4, X5, X6, X8, X9, X10, the value exceeding 1 means more intensified problems in rural areas than in cities. The higher the value, the less favourable the relative situation in rural areas. In cases of either stimulants or destimulants, the value around 1 expresses equal results comparing both spatial layouts, while higher differences to 1 indicate growing spatial inequality and rural-urban divide. Such an attitude allows to focus on spatial differences between rural and urban areas that are not disturbed by a general level of economic development of the member state. The situation in which in some state, as compared to the others, high living standards in rural areas are observed but simultaneously they are enormously poorer than those in urban areas of the state is thus interpreted as unfavourable for the rural citizens. Therefore, the aspect of rural-urban disparities is focused on instead of absolute living standards across the EU countries.
Features of the relative standard of living were grouped in two dimensions: the income dimension, which expresses a role of a level and distribution of income earned by the population, which specifies possibility to meet the consumption needs, and the housing and environment dimension, which takes into account the living conditions in terms of housing and its environmental and social surroundings. The first dimension covered variables X1-X5, while the other: X6-X10.
Although the initial set of indices that describe the relative living standard in rural areas covered 10 variables, they were evaluated with respect to their information capacity (Table 1). Variables with a variation lower than 5% were excluded from the analysis as they have no potential to differentiate the situation across the EU economies. Therefore, the variables X3 and X7 were eliminated. Moreover, coefficients of correlation between the variables were calculated forming the basis to exclude those with a value exceeding 0.9 as duplicating information. There was identified a pair of X1 and X2 variables as the only ones so strongly correlated and X1 was eliminated. Finally, variables X1, X3, and X7 were excluded from the set of variables describing the relative living standards in rural areas.
Moreover, general statistics describing the role of agriculture in an economy were used in the research. For this purpose, data about activity in Section A of NACE Rev. 2 classification, i.e. 'Agriculture, forestry and fishing', were derived from the Eurostat database. A share of the activity classified in section A in total activity-covering all NACE sections was calculated. Three variables were used:  -lab: A share of employment in agriculture in total employment in the national economy (Eurostat,NAMA_10_A10_E); -gva: A share of gross value added in agriculture in the total gross value added in the national economy (Eurostat, NAMA_10_A10); -prod: Relative labour productivity in agriculture (compared to total labour productivity)-the variable is a ratio of the above-mentioned shares of agriculture in gross value added and employment (Eurostat, NAMA).
The data about the role of agriculture in an economy allow verifying the existence of any relationship between deagrarisation and the living standards in rural areas.
The robustness of such relationship was additionally tested concerning other variables, i.e.: -gdp: Real GDP per capita specified as gross domestic product at market prices in chain linked volumes (2010), euro per capita (Eurostat, SDG_08_10); -act: Activity rate in rural areas to total activity rate (Eurostat, LFST_R_AREDNU); -unemp: Unemployment rate in rural areas to total unemployment rate (Eurostat, LFST_R_URGAU); -low_work_int: People living in households with very low work intensity in rural areas to people living in households with very low work intensity in cities (Eurostat, ILC_ LVHL23); -edu: Population with tertiary education by educational attainment level in rural areas to population with tertiary education in cities (Eurostat, EDAT_LFS_9913); -health: Self-perceived health as 'very good' in rural areas to self-perceived health as 'very good' in cities (Eurostat,HLTH_SILC_18).
The variables can express influence on the relative living standards in rural areas of such factors as a general level of socio-economic development and affluence of societies, the situation at labour markets and human resources economic engagement as well as a level of human capital. These may be essential determinants of the standard of living, modifying the role of agriculture for rural areas development.

Methods
To fulfil the research tasks, different methods of analysis are used in the paper. First, to identify groups of the EU countries with similar characteristics in terms of relative living standards in rural areas, the Ward method (1963) has been used. This is a classical method of agglomeration that adopts a minimum total within-cluster variance criterion to specify groups of objects. It is perceived as one of the most efficient methods of agglomeration, although creates not numerous clusters (Electronic…, http). The variables have been standardised to avoid discrepancies resulting from different scales. In the clustering process, the Euclidean distance has been adopted as a measure of differences between the EU economies. As a result of the Ward method usage, a classifying tree has been created.
Second, to assess the relative living standard in rural areas in an aggregate way, linear ordering with a usage of the Hellwig (1968) measure of development has been conducted. In the process (Fig. 1), the abstract model of the relative living standards has been created with maximum values for standardised stimulants (i.e. X2) and minimum values for standardised destimulants (i.e. X4, X5, X6, X8, X9, X10). The Euclidean distance of each EU economy and the model (d i ) has been calculated and has formed a basis to specify the final aggregate measure of the relative standard of living in rural areas (LS) with the formula LS i = 1 − d i /d, where d is a mean value of the Euclidean distances plus its double standard deviation. The higher LS, the more favourable for rural areas is the standard of living. There has been also calculated two separate measures of the relative living standards concerning its income dimension (ID) and the housing and environment dimension (HED). ID has been based on variables X2, X4, and X5, while HED has been created taking into account variables X6, X8, X9, X10. As a result, three rankings of the EU countries have been achieved. Furthermore, the values of the aggregate measures of the relative living standards have been used in later analyses.
Third, to verify the relationship between the role of agriculture in an economy and the relative standard of living in rural areas, ANOVA, correlation and regression analyses have been conducted. In any case statistical significance has been tested at the 0.05 level. ANOVA was used to test whether the clusters of the EU states, distinguished basing on the set of individual features of relative living standards, differ with respect to their agricultural involvement. Some initial insights were also derived from correlation analysis.
It gave the basis to specify regression models (Fig. 2) assuming that the differences in the relative living standards in rural areas can be explained by an advancement of deagrarisation. To test the main hypothesis on the negative impact of agricultural pattern of development, a simple linear regression model 1 was constructed in the following form: where RLS represents the relative living standard in rural areas and is measured either by the aggregate measure LS or by the aggregate measures for each of the living standards dimensions: ID or HED, while agr represents the role of agriculture in an economy and is measured either by a share of agriculture in employment (lab) or by a share of agriculture in gross value added (gva). Positive verification of the hypothesis is specified by significant and negative parameters for agr variables.
After considering the impact of the control variables, a multiple linear regression model 2 was specified in the following form: where, in addition to model 1, cont n stands for the set of n control variables that represent the impact on the relative living standards of such characteristics as a general level of development (gdp), conditions of labour market (act, unemp, low_work_int) and human capital (edu, health). The OLS method was used to estimate the parameters of the (1) RLS = β 0 + β 1 agr + ε (2) RLS = β 0 + β 1 agr + β n cont n + ε Fig. 1 Scheme of the LS measure construction with the Hellwig method. Source own study models. In model 2 a backward stepwise procedure was used to specify the set of explaining variables.

Clustering the EU Countries Concerning Spatial Features of Standard of Living in Rural Areas
Focusing on relative living standards compared in spatial layouts, it is worth to notice that in the whole EU about a quarter of the population lives in rural areas (Fig. 3). However, the situation is strongly diversified across the member states reflecting individual patterns of the spatial distribution of the population. Slovenia, Slovakia, Ireland, and Lithuania belong to countries with the highest share of rural population-exceeding 40%. These countries may thus be pointed as those where people chose to live in the least densely populated areas, which results in specificity of their living standards. On the other hand, in the Netherlands merely every tenth citizen lives in rural areas. A low percentage of rural inhabitants is also observed in Spain, Belgium, Italy, and Cyprus, where the level falls below 20%. However, in any case the share of rural inhabitants cannot be treated as marginal, and the assessment of rural living standards related to urban ones appears as an important research task.
As the EU countries are strongly differentiated with respect to characteristics of their spatial systems, it is useful to identify some groups of economies revealing similar patterns and experiencing similar problems in the development of rural areas. Results of such clustering may create a base to specify common challenges and to share successful solutions aimed at limiting socio-economic tensions connected with rural-urban inequality and supporting development of rural areas. The task has been carried out taking into account different features of the relative standard of living in rural areas, with the usage of Ward method (Fig. 4).
Based on the classification tree it is possible to distinguish three groups of the EU countries, which consist of: -Slovakia, Cyprus, Romania, Hungary, and Bulgaria (group A), -Poland, Croatia, Lithuania, Spain, Latvia, Estonia, Italy, Greece, Portugal, Ireland, France, and Denmark (group B), -Finland, Slovenia, Czechia, Netherlands, Luxembourg, Austria, Germany, Sweden, and Belgium (group C). If the clustering process had been stopped later than at a distance of 10, groups B and C would create one cluster, expressing specificity of rural-urban disparities in group A. Some detailed characteristics of relative living standards in rural areas in the three groups are presented in Table 2.
Group A reveals the poorest results in the relative standard of living in rural areas considering all individual variables. Rural inhabitants in these countries not only earn lower incomes, more often suffer from poverty and experience material and housing deprivation than in cities, but also compared to the other groups of countries the disparities are the strongest. Furthermore, in the cluster some environmental problems in rural areas appear more frequently than in the other countries, although they are more favourable in rural areas than in cities. Generally, group A distinguishes by an exceptionally unfavourable for rural areas relative standard of living in the income dimension. However, even with such severe income inequality and strong environmental tensions as compared to other  countries, the living standards in rural areas cannot be unambiguously stated as inferior (or superior) to those in cities. Group B is characterised by the most favourable results concerning features of surroundings in rural areas-with the least severe nuisances specified by noise, pollution, or crime as compared to cities. Additionally, the housing and material deprivation in rural areas are lower than in urban ones, although they are at a medium level compared between the groups of countries. On the other hand, median income and poverty rate appear less favourable for rural inhabitants than city dwellers, but still the variables are at a medium level comparing the clusters. Summing up, group B offers the best environmental conditions for living in rural areas. They are only slightly levelled by unfavourable income results, and thus the general living standards are relatively high.
Group C achieves the best results in terms of relative living standards in rural areas in the income dimension. The median income is only slightly lower than in cities, while the poverty rate and the material deprivation rate are definitely lower in rural areas. Moreover, all the variables in the housing and environmental dimension of living standards are more favourable for rural inhabitants than for city dwellers, with the lowest housing deprivation when comparing clusters, and nuisances in surroundings at the middle level. Generally, a citizen of a country from the group achieves definitely better living standards inhabiting rural areas than urban ones.
The results indicate that rural standard of living is seriously limited by material problems specified by low income level, poverty, and deprivation. On the contrary, the environmental features of surroundings are more favourable for rural than urban inhabitants. Taking all features of living standards into account, it is possible to state that one living in rural areas usually enjoys better standards than being a city dweller.
The research revealed that policy aimed at levelling living standards in rural areas must be focused on problems with income inequality in line rural-urban areas. Such kinds of socio-economic disparities must be central, especially in countries from group A, while they create definitely less serious tension in countries from group B and are even reversed in group C. Moreover, in group A environmental issues should be constantly monitored as they seem to be the least favourable comparing all clusters. Nevertheless, politicians from group C should consider similar environmental monitoring as, although income of rural inhabitants is the highest, the advantages of rural surroundings are less significant. Still, the most serious challenges are faced by the countries that form cluster A where the problems coincide.

Aggregate Measure of the Relative Living Standard in Rural Areas
Although the clusters reveal some distinguishing features of the standard of living, each country encounters individual problems that create disparities between rural and urban areas. It is important to identify the economies that are characterised by the most and the least favourable relative living standards in rural areas. To simultaneously cover all the features of the standard of living, the aggregate measure LS was constructed. It allows to rank the EU economies concerning the relative living standards-from those where rural inhabitants enjoy high living standards as compared to city dwellers to those with the most severe problems in rural areas (Fig. 5).
The most favourable standard of living for rural inhabitants characterises France, Austria, Italy, and Denmark. These are highly developed economies with high national income, advanced economic structures, and mature societies. Such general features allow them to shape more balanced spatial layouts considering living standards. These post-industrial economies experience socio-economic changes connected with gentrification of rural areas, which are no longer perceived as unattractive places to live. On the contrary, the least favourable for rural inhabitants results are gained by Romania, and Cyprus, followed by Hungary, Slovakia, Bulgaria. They all created cluster A with the poorest results concerning the relative standard of living. Most of the countries at the bottom of the ranking are post-soviet states, with low average income and developmental delays in terms of structural transformation. The industrial pattern of their development, often historically determined, makes urban centers places with a wider range of economic perspectives for higher income and leaves rural areas as underinvested backward agrarian territories with problem of hidden unemployment and depopulation. Generally, the more developed economies as measured by GDP per capita, the higher a relative standard of living in rural areas More detailed analysis of the relative standard of living may be conducted concerning its two dimensions: income (ID), and housing and environment (HED), also described by the aggregate measures (Fig. 6).
The best results in the income dimensions (ID) achieve the most developed economies with the highest GDP per capita, such as Austria, the Netherlands, Luxembourg, and Belgium. In these countries residents of rural areas are usually richer than city dwellers. On the contrary, the strongest socio-economic tensions resulted from spatial income inequality in living standards unfavourable for rural inhabitants may be observed in the poorest economies, which accessed the EU in 2004 and later, such as Romania, Bulgaria, or Hungary. The correlation between ID and real GDP per capita is positive and very strong, reaching level 0.729 (calc. based on: Eurostat, ILC; Eurostat, SDG). It indicates that economic results at national level are not equally distributed in territorial layouts and the rural-urban divide is strong in income dimension. Moreover, the results in the income dimension of living standards appear strongly differentiated across the EU states ranging from − 0.194 to 0.915.
Considering the housing and environment dimension of living standards (HED), it is possible to point at favourable results in Italy and Greece. Relatively high ranks achieve other South European countries, such as Portugal and Spain or Croatia. On the other hand, Cyprus, followed by Slovakia, Romania, and Hungary are at the last positions. There is no direct association between the general level of development and the results in the HED of living standards (the correlation coefficient equals 0.199 (calc. based on Eurostat, ILC; Eurostat, SDG)). It suggests that although living in rural areas still means living in better surroundings, environmental problems may arise in any country regardless of its level of Fig. 6 Rankings of the EU countries by the relative standard of living in rural areas in its income dimension-ID (a) and housing and environment dimension-HED (b) in a period 2010-2019. Source own calculations based on Eurostat data (Eurostat, ILC) economic development. Nevertheless, it requires stressing that housing and environmental conditions are generally more favourable in rural than urban areas, and, additionally, they are less diversified across the EU countries than the income dimension of living standards.
The individual results of the economies are consistent with the characteristics of the EU clusters (Table 3). Group A is characterised by the lowest results either in any of the dimensions of the living standards or in the aggregate measure LS. Group B achieves the best results considering the housing and environment dimension, while group C-the highest achievements in the income dimension as well as in the aggregate manner of LS.
Generally, the relative living standard in rural areas grows in a course of a general level of socio-economic development. Moreover, it seems to be strongly determined by the pattern of development in its income dimension, whereas the housing and environmental living conditions appear to be more stable in a process of development. It is worth pointing out that the important dimension of socio-economic development, apart from the growing GDP per capita level, is specified by the diminishing role of agriculture.

The Role of Agriculture in the EU States in 2010-2019
As rural area development is traditionally associated with agriculture, it is reasonable to identify its role in the EU countries and verify whether more deagrarised economies are those with better relative living standards in rural areas or the opposite relationship appears. It is also essential to specify the potential influence of agriculture on the standard of living in its two dimensions to conclude about possible channels of the impact.
The role of agriculture in national economies is reflected by its shares in total employment (lab) and in total gross value added (gva). Moreover, its development may be assessed by relative labour productivity (prod) specified as a ratio of a share of agriculture in gross value added and a share of agriculture in employment. Such features of the analysed 26 EU states in 2010-2019 are presented in Table 4.
Data reveal that the role of agriculture in the EU states in 2010-2019 was inconsiderable, as it absorbed on average merely 6.4% of total employment and created 2.6% of total gross value added. It is definitely less than the share of population living in rural areas, indicating the fact that an agricultural character is no longer a dominant feature of these spatial layouts and reflecting the growing multifunctionality of rural areas. Most of the rural inhabitants are no longer involved in farming. Their economic activity may be conducted either in the nearest cities or be a local one but non-agricultural. These are the signs of deindustrialization of rural societies and may also result from modern migration movements of people-incomers for which rural areas create only a space to live and not to carry out economic activities. Gentrification may change the typical features of rural areas and the assessment of living standards of rural inhabitants, also bringing a new line of socio-economic divide. Furthermore, labour productivity in agriculture was relatively low and it accounted for about half of the total labour productivity. Such a situation may result in differences in living standards of populations engaged in agricultural and non-agricultural activity and negatively influence the achieved economic results when the share of agricultural engagement is high.
However, it is possible that the official data about the role of agriculture are disturbed by unregistered engagement and unpaid work in agricultural activity. It may cause an especially strong bias in societies with a high share of rural population and simultaneously in economies experiencing problems in labour markets, such as high unemployment and economic inactivity. Such a situation obviously influences living standards of the rural population creating the opportunity to obtain additional 'hidden' income, but simultaneously reflects poor perspectives of the people who are facing a long-standing risk of exclusion. It results in an underestimation of agricultural influence on living standards. On the other hand, the problem of overemployment in agriculture can also influence the results based on official statistics, as the situation lowers labour productivity and creates a barrier for improving living standards of rural population. Therefore, the results may overestimate a negative influence of agricultural engagement on living standards. Simultaneously, the data show that the role of agriculture was highly diversified across the EU economies. The lowest share of agriculture in employment and gross value added was observed in Luxembourg, where the shares were less than 1%. The highest shares the agricultural activity gained in Romania, where it absorbed 27% of employment and created slightly above 5% of gross value added. Such relationships resulted in the lowest relative labour productivity in agriculture in Romania, where it did not exceed 20% of total labour productivity. On the other hand, the highest labour productivity in agriculture was observed in Hungary, where it exceeded the total labour productivity and accounted for almost 104%. Hungary was the only country where agricultural productivity was not inferior to total results.
Comparisons between countries at different levels of socio-economic development measured by the real GDP per capita allow noticing a clear tendency to deagrarisation. The economies with higher GDP per capita are usually those with lower employment in agriculture (correlation − 0.564 (calc. based on Eurostat, NAMA; Eurostat, SDG)). Deagrarisation is even more significant concerning results of economic activity-as a share of the agricultural gross value added clearly decreases in the course of development (correlation − 0.806 (calc. based on Eurostat, NAMA; Eurostat, SDG)), which is consistent with Engels law. The basic character of agricultural products, which meet the everyday needs of the population, results in a decreasing share of food spending in households' budgets and thus in economic production. Simultaneously, no relationship is found between relative labour productivity in agriculture and GDP per capita (correlation − 0.012 (calc. based on Eurostat, NAMA; Eurostat, SDG)), suggesting that there is no technological bias in agricultural activity in favour of the most developed economies.

Relationships Between the Relative Living Standard in Rural Areas and the Role of Agriculture in an Economy
Verifying the research thesis that the agricultural character of rural areas influences the standard of living, in this part of the paper, the results concerning living standards are compared with data describing the role of agriculture. An initial premise supporting the thesis about the relatively lower living standards of rural residents in economies with the high importance of agriculture are the results of the variance analysis (Table 5). The variance between the groups of countries specified by relative living standards in rural areas appears to be essential with respect to the share of agriculture in total employment and the share of agriculture in gross value added, while not significant with respect to the relative labour productivity in agriculture. Furthermore, in both significant cases, the average share of agriculture is higher in the countries characterised by lower relative living standards of rural inhabitants. Similar conclusions may be drawn from the analysis of variance between two groups of EU countries (when groups B and C form one group).
Some initial conclusions considering individual data may also be drawn from an analysis of correlation (Table 6) Calculations reveal a negative correlation between the aggregate measure of living standard in rural areas and the share of agriculture in employment and gross value added. It gives the basis to conclude that the economies with the traditional-agrarian character of rural areas are also characterised by lower relative living standards for rural inhabitants. In any case, relationships with agriculture productivity are not found, suggesting that the rural standard of living may be influenced by a scale of agricultural activity rather than by its efficiency.
It can be observed that an especially strong correlation concerns the income dimension, while the correlations with the housing and environment dimension are not significant. Therefore, an agrarian profile of an economic activity interferes mainly with the level and Table 6 Correlation coefficients between the measures of living standard in rural areas and the role of agriculture in an economy LS the aggregate measure of the relative living standard in rural areas, ID the income dimension of the aggregate measure of the relative living standard in rural areas, HED the housing and environment dimension of the aggregate measure of the relative living standard in rural areas distribution of income among the population. It is supported by the correlations with the individual variables, which appear significant for all features of the income dimension of living standards, while in the housing and environment dimension only the variable concerning housing deprivation significantly correlates with shares of agriculture. Moreover, all the correlation coefficients are of expected sign. It additionally supports a thesis about a higher rural standard of living in deagrarised economies. Assuming the direction of the impact going from agriculture to the standard of living, it is possible to specify the relationships in more details using the OLS regression (Table 7).
The simple regression results prove that the scale of the agricultural sector in an economy significantly influences the relative standard of living in its income dimension. More agrarian economies, either in terms of employment or in gross value added, experience lower level of the aggregated relative incomes, poverty, and material deprivation in rural areas. The models for the aggregate relative living standard LS do not meet the statistical requirements (e.g. for residuals normality), suggesting a strong influence of other factors on the more compound measure of living standards. Meanwhile, deagrarisation appears to be an essential determinant of increased rural income. The income channel is thus the most important way of influence of agriculture on the standard of living.
Nevertheless, some additional variables have been used to check the robustness of the results and to identify other potential determinants of relative living standards in rural areas. A general level of socio-economic development measured by GDP per capita (gdp), some features of labour markets (relative activity rate in rural areas: act; relative unemployment rates in rural areas: unemp; relative low work intensity in rural areas: low_work_int) and some characteristics of human capital (relative rate of tertiary education in rural areas: edu; relative self-perceived very good health in rural areas: health) were considered as the potential explanatory variables. The results of multiple linear regression that fulfil the basic statistical assumptions are presented in Table 8. To specify a final set of explanatory variables, a backward stepwise regression was used.
The models explain about 87% of variability of the income dimension of living standards in rural areas across the EU countries and thus are of high explanatory power. Concerning either a share of agriculture in employment or in gross value added, they still prove negative influence of agricultural pattern of development on the income dimension of the relative living standards in rural areas. The signs of coefficients have not changed; however, the influence of agriculture appears weaker when considering other determinants of the living standards. Moreover, all additional explanatory variables reveal expected signs.
Considering agricultural employment as a determinant of the income dimension of the relative living standards, its influence is supplemented by a very weak but positive influence of GDP per capita and negative influence of relative unemployment. The most developed and affluent economies are those characterised by relatively high income of rural societies and less intense problems of poverty and material deprivation in rural areas. Furthermore, the higher unemployment among rural inhabitants as compared to cities also negatively influences living standards, limiting chances to earn income and deepening poverty and deprivation.
Considering a share of agriculture in gross value added as a determinant of the relative living standards, other explanatory variables complement its influence, i.e. low work intensity in rural areas as compared to cities and a relative share of highly educated people. A higher share of people excluded from employment limits the possibilities of a better standard of living in its income dimension. Simultaneously, widespread higher education in rural societies increases results in the income dimensions of living standards. Human capital resources as well as their economic engagement are thus the most important factors that together with a more deagrarised production lever rural standard of living.
Summing up, the regression results show that a higher share of agriculture in an economy lowers the relative living standards in rural areas. In other words, deagrarisation has a positive impact on the living standards in rural areas as compared to urban areas. Moreover, with regard to other factors that determine living standards, the negative influence of the agricultural pattern of development can be deemed to be stable.

Conclusions
The presented research revealed some important patterns when comparing the standard of living in rural and urban areas across the EU countries. At the actual stage of development of most of the member states, the standard of living appears to be more favourable for rural inhabitants than for city dwellers. However, as a compound phenomenon, it often reveals opposite characteristics regarding its detailed features. The rural standard of living is seriously limited by material problems specified by low income level, poverty, and deprivation. On the contrary, the environmental features of surroundings are more favourable for rural than urban inhabitants. Therefore, it may be concluded that policy aimed at increasing living standards in rural areas must focus on problems with income inequality across the spatial layouts.
Comparisons across the EU countries show that rural areas are becoming better places to live in the course of a general socio-economic development. The more affluent an economy is, the higher the relative rural living standard, especially in its income dimension. It indicates at the changing characteristics of rural inhabitants regarding their material situation specified by earning possibilities. The outflow from agricultural engagement and the inflow of gentrifiers never associated with agriculture can form modern bases for such an improvement in the income dimension of rural living standards. Both tendencies reflect a more general process of deagrarisation.
However, the results in the income dimension of living standards appear strongly differentiated across the EU countries, pointing at differences in the patterns of rural economic development. It sheds light on unique challenges that policy makers in each member state face conducting policy aimed at increasing territorial cohesion and sustainable development of rural areas. It suggests that initiatives decided upon and carried out by those involved on a local scale could be the most effective ones.
The deagrarian pattern of development is clearly visible when comparing the EU economies. The economies with higher GDP per capita are usually those with lower employment in agriculture. The deagrarisation in the course of development is even more significant concerning the results of economic activity. Simultaneously, the research allows to claim that there is no technological bias in agricultural activity in favour of the most developed economies. It suggests that socio-economic policy should support structural transformation of the whole economy rather than focus on sectoral instruments increasing agricultural productivity. It is necessary to create possibilities to work and improve employability outside an agriculture.
Finally, the results of the research confirmed that deagrarisation constitutes an important factor determining the relative standard of living in rural areas. More precisely, deagrarisation induces growth in the relative living standards in rural areas and the relationship is deemed to be stable. It thus forms an argument supporting the transformation of rural areas assuming their multifunctional development, transforming the main functions of the countryside and creating there a space of high comfort of life.
However, the impact of agriculture on the living standards is observed concerning the income channel, while there is no basis to claim that the living standards specified by housing and environmental conditions are influenced by changes in agricultural shares in an economy. Nevertheless, the features of surrounding must be kept under permanent monitoring to avoid potential negative tendencies during the transformation of rural areas.
Generally, deagrarisation and multifunctional development of rural areas create a new reality and strongly influence the rural standard of living. Especially the phenomenon of gentrification requires special attention and needs further analyses as it brings a new threat of division between the gentrifiers and the peasants. The favourable surroundings of rural areas stimulate the inflow of new incomers who usually earn higher income, and the situation smooths the official economic divide between rural and urban areas. Nevertheless, a considerable part of rural community may still be in a need of social support. Thus, there is a growing need for both in-depth research about intrarural divisions between the gentrifiers and the peasants, and cohesion policy measures focused on problems of integration of rural societies. Scientific diagnosis would help to direct the social support to groups at a risk of exclusion. Poor economic standards of living of many rural inhabitants may also induce migration movements either to cities or abroad and cause problems of depopulation of rural space, demographic ageing and the loss of the workforce, which also should be investigated further in scientific research and taken into account when planning policy measures aimed at rural development. Thus, the European Union policy should consider challenges of social and economic integration induced by the transforming character of rural areas and in advance collaborate tools to prevent negative phenomena. In this context, it is important to realise that the rural standard of living may be increased by supporting development of human capital or limiting discrepancies at labour markets.
Some ideas that are concise with the presented findings have already been incorporated into the EU Common Agricultural Policy (CAP) and other initiatives directed at improving living standards in rural areas. On the one hand, within its first pillar, the CAP directly supports farmers by income payments, which can limit disparities between living standards of people engaged in agricultural and non-agricultural activity. However, such actions may also perpetuate the agricultural structure of an economy that appeared to negatively influence the relative rural living standards. Income support must thus be offered simultaneously with incentives for rural transformation specified by deagrarisation. Such elements are included in the second pillar of CAP-the EU's rural development policy aimed at a wide range of economic, environmental and social challenges of the twenty-first century. It is to promote other than productive functions of rural areas, encourage economic diversification, protect environment, and, generally, to improve a quality of life. It covers also the community-led local development strategies supporting the integrated development of rural areas, which seem to be well-adjusted to the unique conditions of each spatial layout. Simultaneously, the CAP is led in parallel to numerous horizontal programs focused on the development of human capital and other factors of sustainable development. The wide range of policy measures directed at rural development gives the opportunity to overcome the problems faced by rural communities of the EU member states. However, strong scientific bases for such policy are necessary, and hence there is an urgent need for further research of the compound phenomena connected with development of rural areas.