Introduction

For many years we have witnessed a net migration from rural areas to urban areas all over the world. The main underlying forces are structural changes from rural to urban industries and jobs. However, the city is also considered to be attractive in many ways that could improve human well-being. The city has offered more attractive job opportunities; more education possibilities for young people; a much wider spectrum of cultural activities, like museums, theatres and sporting events; many kinds of cafes and restaurants; specialty stores; and so on. Furthermore, objective measures show greater average life quality in the city. One can summarise this perspective on urban growth with the title of Edward Glaeser’s seminal book The Triumph of the City (Glaeser, 2011). However, in recent years, research from all over the global North has shown either no difference in subjective well-being between rural and urban areas, or even higher average levels of subjective well-being in rural areas (Burger et al., 2020; Okulicz-Kozaryn, 2017; Sørensen, 2013; Berry & Okulicz-Kozaryn, 2011). In the literature, this phenomenon is termed the urban happiness paradox.

However, the results are, to a large degree, heterogeneous across countries, and the difference between urban and rural subjective well-being can change at different speeds and different directions in different developed countries. New generations and new urban–rural migrations patterns, for instance, can rather quickly change the balance in the overall results. Furthermore, the results are not ‘global’ in the sense that they apply to all kinds of people. These and other insights about the complex patterns spur the scientific community to do more case-oriented research to pinpoint the detailed differences in single countries and try to explain the pattern of subjective well-being in different geographical areas. In case studies it is easier to do away with the rural–urban binary, as written about in the introductory Chapter 1. When focusing on selected case countries, in general it is also much easier to find or select data targeted at the specific research questions.

Difference in subjective well-being between urban and rural areas

Are people happier and more satisfied with their lives in rural areas, and if so, does this derive from the composition of the population, or is there a genuine causal association so that certain area-specific qualities on average make people happier? Which sociodemographic or other groups are thriving better in rural areas, and which place-based qualities in the rural areas have an impact on subjective well-being for people in general or for specific groups of people? These questions are of great importance for theory and empirical research in the field to answer.

Two very different and opposing theoretical strands stand out in recent scientific debate about rural areas vs the city in respect of people’s well-being, and two of the most cited authors are placed in each of these strands. On one side stands economist Edward Glaeser. In his seminal book Triumph of the City (Glaeser, 2011) he glorifies the city. The city has been one of the prime motors of the development from poor agrarian societies to educated, rich, democratic societies. In the city, different kinds of people meet, new ideas are put into practice, people get education, entrepreneurship and business flourish etc. Besides, the city beats the countryside in being more climate friendly. However, what about the inequality, poverty and slums in many big cities? Of course, Glaeser recognises that slums can be problematic, and that some cities have got problems which authorities should take care of, but overall, the city is a good thing, not just as a temporary step in the development of human life on earth, but also at present and in the future. Besides, most people in city slums are better off here than in the poorer countryside. In his opinion, the city slum is not inhabited with unhappy people because slums are a bad thing that make people unhappy. Unhappy people migrate to the city slums because of unhappiness, and because they seek, and often find, a better life in the city.

On the other side of this debate stands political scientist Adam Okulicz-Kozyran, who has written extensively on this subject. While some people might be better off in the city, Okulicz-Kozyran thinks that in general the city is a bad place for people to live: ‘One explanation that people do not kill each other consistently when crowded in cities is that our instincts are subdued due to culture, norms and so on as Freud has observed long time ago, and of course, there is police and other restraining factors’ (Okulicz-Kozaryn, 2015, p. 105). The countryside stands for Gemeinschaft as opposed to the Gesellschaft, and natural will opposed to economic rationality. In Okulicz-Kozaryn’s opinion, Edward Glaeser falls into a classic ecological fallacy when he states that in countries with the highest degree of urban development, people are on average happier than in other countries (Berry & Okulicz-Kozaryn, 2011; Okulicz-Kozaryn, 2017). However, this is an ecological fallacy only if Glaeser from this deduces that, in modern welfare states, people in cities are on average happier than people in the countryside, and this is indeed not his purpose. His main purpose is to convince the reader that cities have positive effects on the country’s population as a whole, not just the city’s inhabitants. In this respect, the existence of big cities is a macro effect, at least in highly developed countries.

While some of the writings of Okulicz-Kozyran are rather high-flown (e.g. Okulicz-Kozaryn, 2015), his analyses show sober, scientific quality. Together with co-author Brian J. L. Berry (Berry & Okulicz-Kozaryn, 2009), he analysed World Values Survey data merged with European Value Study data from around the turn of the century, including sixty countries around the world. There were two main results. First, after controlling for relevant individual-level characteristics like age, education and income, there was on average no difference in happiness between urban and rural areas in low-income countries. Second, when analysing high-income countries, the average level of happiness was statistically significantly greater in rural areas than in large cities. More specifically, this second conclusion applies mainly to countries with an Anglo-Saxon heritage and not to countries with a Latin heritage. In a follow-up article, Berry and Okulicz-Kozaryn (2011) analyse the urban–rural happiness difference in the US only, with data from the General Social Survey. In this article, the authors find a clear pattern showing an urban–rural happiness gradient with the urban–rural measured on a four-point ordinal scale, from rural areas and small towns to large central cities. Over the years from 1972 to 2008, average happiness nearly constantly follows the same trend from lowest in large central cities to highest in rural areas and small towns. This clear pattern is intact after controlling for background variables like age, income and marital status. However, after inclusion of variables for race and ancestral roots (Northwest Europe, Mediterranean, Africa and reference category ‘other’), the authors find a very weak and marginally statistically significant effect only for the most rural category vs big cities. That is, race and ancestral roots explain nearly all differences in happiness between urban and rural residence. Only in big cities with more than 250,000 inhabitants do the authors find average happiness a tiny bit lower than elsewhere. Despite these results, the authors end the article by stating that the continued migration from big cities to countryside among other things reflects a ‘fundamental feature of American life, the continuing pursuit of the happiness associated with lower density living and the persistence of cultural difference associated with it’ (Berry & Okulicz-Kozaryn, 2011, p. 881). In addition, the authors could have discussed the possible interesting connection between two different conclusions in their two articles. In the article published in 2009, they find that the urban–rural happiness gradient only applies to countries with an Anglo-Saxon heritage and not to countries with a Latin heritage. In the 2011 article, they find that a dummy variable for northwestern ancestral roots explains a lot of the urban–rural happiness gradient.

Theoretically, Okulicz-Kozaryn, with changing co-authors, draws heavily on Louis Wirth’s work on the effects of living in an urban environment (Wirth, 1938), and Wirth, again, draws mostly on the works of two grand old men from the first generation of sociologists, Emile Durkheim and Georg Simmel. The main hypothesis is that the city, through population size, density and heterogeneity, develops ‘anomie, alienation, and social disorganization’ (Berry & Okulicz-Kozaryn, 2011, p. 872). On one hand, the individual in the city gets more freedom through emancipation from tradition and customs, but also ‘loses, on the other hand, the spontaneous self-expression, the morale, and the sense of participation that comes with living in an integrated society’ (Wirth, 1938, p. 13). As Wirth further writes about the city dweller, ‘only rarely is he truly a neighbor’ (Wirth, 1938, p. 17). These things cause, Berry and Okulicz-Kozaryn (2011, p. 873) write, the city to destroy social capital and generalised social trust (drawing here too on Robert Putnam) with the cost, among other things, of lesser life satisfaction. Furthermore, in the city the pecuniary nexus displaces personal relations as the basis of association (Wirth, 1938, p. 17). Life becomes a rat race, always striving for more. In a later article, Okulicz-Kozaryn and co-author Joan Maya Mazelis tried to test these hypotheses in their statistical analyses by controlling the effect from the urban–rural continuum with city problems like crime and poverty. They found a robust effect after controlling and conclude, with some caution, that the city per se is lowering subjective well-being, and that the lower subjective well-being in cities is not caused solely by problems that often accompany city life (Okulicz-Kazaryn & Mazelis, 2018). Unfortunately, for some reason the authors do not control for northwestern ancestral roots, a factor that in a previous article by Okulicz-Kozaryn, mentioned above, was one of the most important factors in explaining the urban–rural happiness gradient in the US.

With some hesitation, Okulicz-Kozaryn, in one of his latest publications (Okulicz-Kozaryn & Valente, 2019) admits that there is some evidence from the US that the gap between urban and rural dwellers in subjective well-being over the last decade has diminished, and that the youngest generation even seems to be happier with living in the city than in the countryside. Nothing points to an age effect, i.e. a passing effect, and he cannot find an explanation for this among his usual arsenal of explanatory factors. Furthermore, he agrees with Glaeser that cities are more climate friendly than sprawling people out in the countryside. His conclusion is, however, a bit surprisingly, that we ought to reduce the human population on earth so that the climate can cope with human sprawl (Okulicz-Kozaryn, 2015). Likewise, Glaeser recognises that not all cities should be triumphed, for instance the outdated industrial Detroit, and he recognises urban poverty: ‘The occasional success story doesn’t mean that urban poverty isn’t awful. It is’ (Glaeser, 2011, p. 75). On the other hand, Glaeser as a real economist thinks that people sometimes willingly, and for good reasons, choose to be less happy than what is possible: ‘humans are quite understandably willing to sacrifice happiness or life satisfaction if the price is right’ (Glaeser et al., 2016, p. 169), and the price can be, for instance, high income or low housing prices. This statement runs counter to much of the theory on subjective well-being. As written about in Chapter 1 in this book, much theory argues that the concept of subjective well-being has advantages over the concept of preferences to learn something about utility. Clearly, agents could be willing to offer some pleasantness and perhaps feelings of happiness for some other goals, but the same would hardly account for life satisfaction. However, this is also a question of definition of the concept of life satisfaction.

The heated scientific debate about big city life vs rural life does sometimes seem pervaded by misunderstandings and ideological oppositions, with one camp discussing homo economicus and the other camp discussing the natural human being and a lost paradise. However, as one of the forerunners in happiness research, Ruut Veenhoven (1984) wrote, ‘[t]‌here is some truth in the socio-biological assertion that evolution did not design us for city life. Yet it did not predispose us to rural life either. Current sedentary life in farms and villages is equally remote from the original hunter/gatherer life as urban life in streets and stockbuildings.’

An example of possible misunderstandings is Okulicz-Kozaryn’s claim that in developing countries there is, on average, no difference in happiness between urban and rural areas. He shows, correctly, that after statistical control for background variables like education and income, happiness is the same. He also writes that Veenhoven, back in 1994, was wrong in writing that average happiness was greater in urban areas in developing countries. The problem here is that Okulicz-Kozaryn perhaps overdoes the statistical controls. The question is whether one should consider variables measuring, for instance, education and income as control variables or mediating variables. They can be both, but in developing countries, these variables are probably mostly mediating variables. The city causes education and income to rise. This is one of Edward Glaeser’s main points in his book Triumph of the City and is elegantly analysed empirically in an article written by Easterlin et al. (2011). Part of the subjective well-being gap in developing countries could be because of selectivity of rural–urban migration, where the higher educated from the rural areas migrate to the city, but as the authors write, ‘while selectivity of rural–urban migration could conceivably contribute to the observed urban–rural differences in life satisfaction, it is highly unlikely that it could be quantitatively important’ (Easterlin et al., 2011, p. 2194). Furthermore, the authors write that the levelling of the happiness gap between urban and rural areas in developed countries ‘is due largely to a convergence in urban and rural occupational structures, income levels, and education’ (Easterlin et al., 2011, p. 2195). Two historical trends are mostly responsible for this convergence. First, there is a weakening of close bonds between place of work and place of residence caused by the development of general-purpose technology. Second, there has been an increase in the elderly opting for rural residence, caused by changing demographics and rising income.

In the 2020 World Happiness Report, Burger et al. (2020) use cross-sectional data from the Gallup World Poll across 150 countries (2014–2018) to analyse differences in subjective well-being between urban and rural areas. Again, we see the same picture, with greater well-being in cities in the developing countries and a closed well-being gap in the developed countries or even a little higher level of well-being in rural areas. In their analyses, the authors use three different measures of subjective well-being, life evaluation (the Cantril ladder) and both positive (enjoyment and laughter) and negative (worry, sadness and anger) affect. The general pattern, described above, is most pronounced when using life evaluation as the dependent variable. When using measures for positive and negative affect, the number of countries with a non-statistically significant difference between rural and urban areas is much greater. After having investigated the general pattern on the full data set, the authors selected two extreme cases in respect of the size and direction of the happiness gap between rural and urban areas, sub-Saharan Africa and the Western world. In these two cases the authors investigated what factors lie behind the urban–rural gap. In sub-Saharan Africa, the higher subjective well-being in urban areas is mostly due to economic situation, economic optimism and education. Whether one should consider these factors as causes or mediating factors is debatable, as mentioned above. In countries belonging to the Western world, these same factors still push subjective well-being up in urban areas in comparison with rural areas. However, other factors show significant effects that point in the direction of greater rural area subjective well-being, namely a higher degree of community attachment and housing affordability and a lower percentage of single households (Burger et al., 2020). Measures for social capital and feeling of safety also point in the direction of rural area advantage, but are not statistically significant. Some of these factors can be deduced from the theory of Wirth, Simmel, Tönnies and Putnam, referred to above and used as the theoretical basis by Okulicz-Kozaryn in his analyses.

On data from the European Values Study from 2008, Jens Fyhn Lykke Sørensen (2014) finds greater average happiness for rural dwellers than for urban dwellers inside the EU region. However, as also applied for the US, the difference is small, although statistically significant after relevant controls. Furthermore, and what applies for the US case too, the findings are not homogeneous across countries/states. Sørensen seeks to explain the higher average subjective well-being in rural areas by including some explanatory factors in his regression analyses. For most of his explanatory hypotheses, these are deducible from the classical theories discussed above: a higher level of feeling of insecurity in the city, higher sense of community feeling in rural areas and a greater amount of social capital and social trust in rural areas. He finds variables in the data that are more or less well suited for measuring these concepts. Among these three explanations of the higher level of subjective well-being in rural areas, only the amount of social capital seems to matter. The measure for social capital, participation in voluntary work, explains part of the difference between rural and urban areas, but the vast part of the difference is intact and still highly statistically significant.

Besides hypotheses deduced from the more classical sociological theories, Sørensen also draws on social reference group theory to explain the lower subjective well-being in city areas. In subjective well-being research, reference group theory states that people use comparison with a reference group when evaluating their subjective well-being, for instance people living in their neighbourhood. This theory is often used as an explanation of the Easterlin paradox, i.e. that wealthy people in general are happier than are poor people, but as societies get more and more wealthy, we do not see any parallel increase in happiness (Easterlin, 1995). Sørensen argues that wealth inequality is more pronounced in cities than in rural areas, and if people evaluate their well-being according to the reference group theory, you might expect lower average well-being in the cities. The hypothesis is interesting, but Sørensen’s operationalisation and analysis strategy is not convincing, and he does not find any statistical explanation either.

Another theory often used to explain the Easterlin paradox is aspiration theory (Bjørnskov et al., 2007). In relation to research in subjective well-being, this theory argues that humans are always striving after more and never really satisfied. We aspire to some goal, but as we reach this goal, we adapt to the new situation and get new aspirations. Because of this, we are caught in a never-ending rat race and with no gain in subjective well-being no matter what we obtain. With reference to Okulicz-Kozyrin and Wirth, we can expect this constant striving for more to be more prevalent in the city, where the pecuniary nexus is hypothesised to displace personal relations as the basis of association.

In the first introductory section of this framing essay, we wrote that cities clearly have a lot to offer in relation to the countryside, when it comes to amenities like possibilities for education, broad spectrum of job opportunities, restaurants, museums, sporting and cultural events etc. However, there is another side to this. Often, the countryside has something else to offer that many people enjoy too. Nature seems very often much more present in the countryside. Of course, there is a lot of romanticising about the rural idyll (Shucksmith, 2018). The reality of modern agriculture is far removed from the romantic image, but at least some rural areas possess some qualities that no city can offer. For instance, van den Berg et al. (2010, p. 1208) found that ‘green space in a 3-km radius around the home significantly decreased the relationships of stressful life events with number of health complaints and perceived general health’. Furthermore, the results from their analyses suggested that the buffering effect mostly applied to ‘more large-scale nature areas, such as forests, dune areas or agricultural fields’ (van den Berg et al., 2010, p. 1208). Brereton et al. (2008) used Irish survey data merged with GIS data to investigate the effect on life satisfaction from a series of spatial variables measuring the distance from respondents to different kinds of amenities. They found, among other things, a strong and highly statistically significant positive effect from proximity to coast. In their analyses, the authors control for living in Dublin, the only big city in Ireland. This dummy variable has a negative effect on life satisfaction. However, the authors do not investigate any possible interaction between the Dublin dummy and proximity to coast, and we therefore do not know if the positive effect from proximity to coast is restricted to the countryside, or if it applies to people living in the wider area of a big city too. This question relates to a broader question about nature and green spaces in cities. Can urban dwellers gain the same positive effects from natural surroundings as rural dwellers if the city planning is right? Obviously, even a city with lots of good green spaces cannot match a good countryside on all criteria. But then again, not all people are alike. In a review article, Wolch et al. (2014) show that parks and urban green space in many instances have a positive effect on physical activity, health and psychological well-being. So perhaps the fulfilment of needs just depends on who you are. As human beings, we might need some kind of green surroundings, but while some just need city parks, others perhaps need a forest, and perhaps even others might just need a cat and some potted plants in their small city apartment. Likewise, some people might thrive in big cities with high density, people everywhere, restaurants, sporting events etc., while other people thrive better in low-density open landscapes.

This last point brings us to the question of personality and values. Research indicates that people with somewhat alike personality structure, via several different social mechanisms, cluster together in geographical areas (Rentfrow et al., 2008). Verma and Thakar (2019) find that personality, conceptualised with the big five (or the Five Factor Model), correlates with personal values, for instance that extraversion correlates with hedonism. This, in a way, relates personality structure with the concept of subjective well-being, where hedonism often is opposed to eudomania, see Figure 1.1 in Chapter 1. The connection between values and subjective well-being is shown too by Morrison and Weckroth (2018) on European Social Survey data from Finland. The authors show how people in metropolitan areas (Helsinki) on average possess more hedonic values than do people living in the countryside, and that on average people with hedonic values have lower subjective well-being.

Research design and the question of causal mechanisms

Causality is ever-present in the discussion about hypotheses and results, although not always explicit. All of the research discussed above uses statistical controls in some way, and the prime reason to use such controls is trying to come a bit closer to conclusions about causality pattern and effect, i.e. the effect of living in a rural area in comparison with an urban area. As mentioned above, it is not always easy to decide if one should consider variables like, for instance, education and income as controls for spuriousness or as mediators of an effect. If one is interested in long-term effects, and especially in developing countries, variables like level of education and income are probably more mediating factors than controls for spuriousness. The city is one of the driving factors in the development of education, rising income and living standards etc. In this way, you can say that control for education and income is overdoing the controls. On the other hand, in developed countries, the city effects have been more or less spread out to the whole country. As Burger et al. (2020, p. 82) write, the rural dwellers ‘are able to “borrow” the positive effects of cities’, and at the same time they are not ‘subject to the negative externalities of urbanization’ (2020, p. 82). So, in that case, it is perhaps more correct to consider these variables as controls for spuriousness. However, even when Glaeser and co-authors analyse data from American cities, they do not control for income or employment, ‘as these represent outcomes that may be caused by an area’s economic success’ (Glaeser et al., 2016, p. 141). At least one should not control for this kind of variable thoughtlessly. The conclusions about effects cannot be deduced from statistical analyses on cross-sectional data. However, this is not a recommendation to abstain from including variables other than the dependent and the main independent variable, and to adjust the ‘effect’ from an urban–rural variable on subjective well-being; one should just be cautious with the conclusions.

Many individual-level variables have been found to correlate with subjective well-being, some of which have been mentioned in the discussions above. Furthermore, many of these variables are correlated with, or interact with, degree of urbanisation too. Some of the most important of these variables are age, marital status, having a vocational education, income, health and social trust. Taking such variables into consideration is crucial when investigating the correlation between urban/rural and subjective well-being.

Nearly all previous research in subjective well-being differences between rural and urban areas have used only cross-sectional data or repeated cross-sectional data. Classical experimental design is out of the question, but a few studies have taken the advantage of also using panel data to be able to learn something more about the causal mechanisms involved. With individual-level panel data, the researcher follows individuals across time with two or more measuring occasions. With this sort of data, the researcher can analyse what happens to people’s subjective well-being when they migrate, when either moving from the city to the countryside or vice versa, and the researcher can compare this development with other non-migrating individuals. In this way, with the control for base-level subjective well-being, the researcher is able to come somewhat closer to the so-called counterfactual situation; for instance, what would have happened to Mrs Johnson’s subjective well-being if she had not migrated from the city to the countryside? Naturally, one cannot from this kind of analysis say that Mrs Robinson, who is staying in the city, and is the same age as Mrs Johnson, is married just like Johnson, etc., would experience the same change in subjective well-being if she too had moved to the countryside. No matter the character of the research design, one should be careful not to draw too ambitious conclusions. People are different from each other and trying to decide which is better for humans to live in, the city or the countryside, is a bit like comparing which taste better, apples or bananas.

In theory, it is of course possible to imagine some globally existing causal mechanisms in favour of either cities or rural areas. This is close to Okulicz-Kozaryn’s view, discussed in the first part of this section. However, even Okulicz-Kozaryn recognised that some people seem to thrive better in city environments. Therefore, if we do not seek rescue in concepts like false consciousness, the causal mechanisms in place will depend on the specific geographical case and which agents we are investigating. On the other hand, it is possible that we can learn something about average effects for some specific groups of people living in, for instance, the Western world shortly after the turn of the century. In fact, this is what we, as social scientists, usually acknowledge, but in this case, it is safe to say that the problem is especially wicked.

Presentation of the chapters

All chapters in this section take on a quantitative approach, and all chapters, except Veenhoven’s, in one way or another investigate differences in subjective well-being between urban and rural areas. Lolle investigates the case of Denmark, a geographically small, universal welfare state. He uses person-level, register panel data merged with survey data from thirty-eight municipalities to analyse differences in subjective well-being between urban and rural areas by way of multilevel regression analyses. The survey data include several different domain satisfaction measures as well as different subjective well-being dimensions.

In the following chapter, Viganò et al. explore whether the difference in subjective well-being between urban and rural areas in Italy has changed between 2008 and 2018, and if so, in what way and how much? In 2008 Italy experienced the same development as is seen in general in the Western world, namely with rural areas having a higher average level of subjective well-being. The question now is whether this pattern is intact ten years later. In their analyses, Viganò et al. investigate more specifically whether effect factors on subjective well-being are more or less the same in urban and rural areas, and if this has changed during the investigated period. Pasqualini too, in her chapter with France as a case, investigates changes in subjective well-being in urban and rural areas. However, she looks specifically at changes during the COVID-19 pandemic. Her reference measure is just before the start of the pandemic, and she has data from eight follow-up rounds through the pandemic.

Colley et al. perform primary and secondary analyses to critically examine the value of outdoor recreation for the well-being of rural residents. The primary focus for the authors is exploring the inequalities in the use of outdoor recreation in rural areas. However, they also discuss future outdoor recreation in relation to the COVID-19 pandemic, and ask if the pandemic might act as a moment of change and reshaping habits. Lund uses the same survey and register data as does Lolle. Also in this chapter, the analyses explore differences in subjective well-being between Danish localities. However, a main purpose for Lund is to discuss definitions and mapping of neighbourhoods. Neighbourhood studies have been done since early Chicago School sociology, but only a few have been interested in trying to map them validly in quantitative studies. Lund has developed a new method for mapping such neighbourhoods more flexibly than with the borders of official administrative units or by just using squares on a map. He shows his method in use on research questions about subjective well-being and neighbourhoods, and with a focus also on urban vs rural.

Veenhoven et al. stretch the focus of the book away from quality of life in rural areas to the incorporation of rural elements in urban environments. The authors introduce the reader to biophilia theory and to the furthering of urban green, and the main research question is whether urban green has a positive effect on happiness. The authors also present a new method for doing synthesis analysis, collecting results from existing research on the correlation between happiness and urban green from the World Database of Happiness. They use results from seventeen empirical studies between 2004 and 2018 from eleven countries around the world.