Health, Wellbeing, and Democratic Citizenship: A Review and Research Agenda

Social scientists have examined the causes and consequences of people’s engagement with politics for many decades, yet we have only just begun to understand the roles that health and wellbeing play in people’s involvement as members of the body politic. Findings from a nascent body of research suggest that health predicts people’s decision to turn out to vote and whether they feel they can have a say in political decisions more broadly, but we still lack a systematic understanding of the variable, as well as specific, ways in which health and feelings of wellbeing shape people’s interactions with political life. We also know little about how—and if—these patterns vary across groups in society, regions, countries, or over time. In this contribution, we present a framework for analysing the ways in which specific health conditions may shape the connection between citizens’ wellbeing and their interactions with politics and how research should endeavour to trace the consequences of these links for people’s lives as citizens and their full participation in the democratic political process.

In surveys collected in 29 European countries between 2018 and 2020, 1 in 3 (34.7%) indicated that their health was 'fair', 'bad', or even 'very bad', while 65.3% rated their health as 'good' or 'very good' (www.europeansocialsurvey.org, Round 9). In the same surveys, over 1 in 4 (26.7%) indicated that they were hampered in their daily activities by illness, disability, infirmity, or mental problems. Hence, depending on how we define physical or mental wellbeing and the strains associated with them, between one quarter and one third of Europeans report that they are coping with less than perfect health. 1 Not only do these numbers constitute significant proportions of European societies, but as the negative correlations of indicators of individual health with measures of subjective wellbeing in Figure 1 show, they shape how people feel about their lives: those who say they are in ill health or are hampered in their daily activities because of illness are also significantly more likely to say that they are unhappy or dissatisfied with their lives.
The fact that significant proportions of democratic electorates report less than perfect health at any given point in time is interesting and important in its own right, but how does it impact our societies? The politics and economics of health are well documented, and health policy, of course, is a long-standing and prominent concern across all contemporary democracies. As a result, we know that inequalities in health conditions are systematic and considerable both within and across countries [1]. Long recognised as an important and costly public issue, health concerns have received attention from researchers and policymakers interested in how government action can redress current inequalities and the potential to reduce future ones [2]. 2 Political choices and policy designs matter for health outcomes.
In a different corner of the social sciences, political scientists for many decades have sought to understand inequalities in democratic political participation. How and why citizens engage with politics is a cornerstone of the study of politics; decades' worth of scholarship have established that people participate in different ways and at different rates because of unequal access to resources like time, money, and civic skills or differences in the motivation to engage with democratic politics. At the level of countries, socio-demographic heterogeneity and economic inequality are key drivers of inequalities in political demand and democratic representation, alongside differences in the formal political institutions that facilitate or hamper people's access to and engagement with political processes and institutions (e.g., [3,4]).
For much of their existence, scholarship in these areas of health policy on one hand and political participation on the other has proceeded on parallel tracks. Yet, there are good reasons to connect them by examining the nexus between 1) inequalities in health and wellbeing and 2) inequalities in political participation and representation. First, while focusing on the political determinants of health and health inequalities is important, we know comparatively little about the flipside of the coin, namely whether and how differences in health and wellbeing matter for politics and civic life (cf. [5,6]). Modern democracies devote significant resources to the wellbeing of their members to ensure full and equal participation in social and economic life, and yet, we have limited knowledge about how health and wellbeing in fact shape the ways in which people exercise their rights and responsibilities as citizens. Put simply, do policies that are designed to produce healthier populations also produce healthier democracies?
Second, health and wellbeing as determinants of political action are important for the study of politics for normative as well as empirical reasons. Commonly viewed as foundational to a well-functioning democracy, citizenship behaviours and attitudes are shaped-whether sustained and enhanced or interrupted and hindered-by the wellbeing of its citizens. It is an especially compelling argument from the perspective of democratic representation: if some groups in society are systematically underrepresented among those who make political claims and for reasons that are remediable, the democratic edifice is lacking.
Below, we begin by reviewing scholarship in health and health inequality on one hand and research on political behaviour and representation on the other to delineate the intersections between people's physical and psychological wellbeing and their cognitive and physical engagement with politics. Because wellbeing and political engagement take many intersecting and variable forms, the links between individual health and wellbeing and political attitudes and behaviours are complex and not always apparent. Moreover, these vary across countries, across different population groups, and over time.
More specifically, to understand how health shapes the quality of democratic life, we need to differentiate among specific types of health conditions and how these, in turn, affect particular kinds of citizenship behaviours. [7,8] Moreover, we will want to know how these relationships vary as a function of individual characteristics, such as age, gender, income groups, and levels of education, that endow citizens with resources or motivations to engage in politics. For example, well-documented skews in political engagement show that older, better-off, and more educated citizens participate more fully in democratic life. As a result, older citizens' voices are heard more consistently and their preferences are represented more reliably. At the same time, we know that different groups of people have variable odds of experiencing distinct health-related problems for different kinds of reasons (e.g., lifestyle, working conditions, or access to preventive and diagnostic care).
Do these patterns go hand in hand or are they, as the example of older citizens may imply, at odds with one another? Does it matter what the nature of the health condition is? For example, do muscular and back pain or heart disease and diabetes have the same effect on political engagement, and are they consequential for the preferences of different groups of people or how these preferences are expressed? Put simply, do health problems cut across the currently documented inequalities in political participation or do they serve to exacerbate them, and do the sources of good and ill health matter differently? These questions are important, but their answers are likely to be found only in careful analyses of rich data, with the help of different methodological approaches. Finally, we want to understand whether welfare systems and access to health care or treatments enhance access to the political arena or skew political demand [9]. Health policy is at the core of modern welfare states, often a central feature of political debates, and its design and operation affect people in ways that shape their resources and how they think about themselves as individuals and citizens. Given existing inequalities in the quality of care, access to care, and treatment, do citizens who need, seek, and subsequently receive treatment for particular conditions participate more fully or differently in politics than those who do not or cannot? Do these groups hold systematically different views about their country's health care policies and practices, and to what degree do we detect variance between the groups in their behaviours?

Understanding Health and Political Behaviour in Democracies
What citizens want from government and what they do to get it is one of the core concerns of the study of politics. Political scientists have spent decades studying this question and we can summarise their insights with the help of some stylised facts. First, people's socio-economic status (SES)-often proxied by income, education, and social class-correlates positively with becoming involved with politics: higher status individuals participate more. Second, the mechanisms that underlie the basic correlation between SES and political action revolve around high SES individuals possessing the necessary resourcesboth material and otherwise-to engage with politics in various ways [10]. Indeed, the classic studies of participation suggest that financial resources and civic skills in particular are key to understanding why some people participate more than others [11]. In addition to these resources, the so-called 'Civic Volunteerism Model' developed by Verba, Schlozman, and Brady argues that participation also requires social connectedness and psychological engagement [11]. Taken altogether, individual resources, mobilising networks, and motivation are at the heart of what makes people participate.
While research on inequalities in citizen politics and access to the necessary resources has therefore been a core concern in the study of politics, it has surprisingly little to say about the connection between health and political cognitions and behaviours. In stark contrast to long-standing literatures linking health and participation in education, work, and family life, standard texts on political participation and voter turnout like Milbrath and Goel [12] or Wolfinger and Rosenstone [13] barely mention health, and Verba, Schlozman, and Brady's [12] by now classic analysis of political participation completely ignores health as a factor that influences involvement in civic life (see also Burns, Schlozman, and Verba 2001 [14]). Similarly, even the most prominent analysis of political activism among older adults in America, which focuses on the importance of resources, does not include any mention of health as a key ingredient [15]. And Schlozman, Brady, and Verba's [16] most recent book Unequal and Unrepresented: Political Inequality and the People's Voice in the New Gilded Age-a summary of three decades of work into political participation in America by some of the most distinguished political scientists of our time-references the health of citizens exactly once, and only in passing, noting that the most politically active citizens also tend to be healthier. As Burden et al. [17 p167] summarise this literature: when it comes to the question of whether health influences political participation, 'the link between the two is often overlooked, sometimes assumed, and poorly documented.' Political scientists' neglect of health as a determinant of civic engagement has been as notable as it has been regrettable. A nascent body of research has begun to address this gap by examining how various health conditions may affect how people think about and actively participate in democratic politics (e.g., [18][19][20][21][22][23]). The most frequent starting point for these studies is a 'health gap hypothesis', which assumes that there is a positive relationship between a person's health and the quantity of their political involvement [5,6,24]. Echoing the civic volunteerism model foundational to research on political participation, the idea of such a health gap or health bias between more and less healthy individuals is rooted in the idea that engaging with politics has a psychological as well as a physical component and therefore requires resources that are linked to health. As Gollust and Rahn [18] note, 'The experience of illness … reduces the non-financial resources -including physical, cognitive, and social/emotional resources -that could promote civic participation.' As a consequence, health should affect political participation in various ways. For example, it impacts people's ability to engage in physical and mental tasks-consider, for example, the cognitive task of deciding who to vote for and then the physical task of casting a ballot-thus shaping their ability and skills to participate fully in democratic politics [18,25]. In addition, if concerns about illness overwhelm a person's attention to other domains in life, including politics, ill health can reduce the motivation to become or remain involved. Or, given that physical and mental health contribute to the creation and maintenance of social networks and participation in social institutions, such as the workplace or religion, it can affect participation via the likelihood of mobilisation and recruitment by social others into political activity. Consistent with the health gap hypothesis, one general conclusion arising from existing studies is that poor health impedes the full participation of citizens in the political process. Moreover, the effect seems to be independent of other important socio-demographic or psychological factors associated with full and active citizenship behaviours [5].
While such baseline findings about the deleterious effects of ill health may be intuitive, they have not been documented universally. Instead, the studies are generally limited in terms of scope and significance, focussing on different groups in society, with most research in this area still being early in the scientific lifecycle and exhibiting several notable limitations. Some of these limitations are empirical and others are conceptual and theoretical. For example, it is not exactly clear as to why particular health conditions alter the ways people engage in specific kinds of political acts and cognitions, or whether they do so consistently across different populations. In addition, important methodological issues and data-related challenges require clarification and examination. Taken together, despite a promising and emergent body of evidence, we are still some distance from fully exploring, let alone understanding, the various ways in which health-related factors shape how and how much citizens engage with democratic politics across countries and across populations and why. At the same time, because the provision of health care is increasingly the responsibility of government, it is important for both normative democratic theorists and empirical social scientists to fully understand whether and how the voices of the healthy or unhealthy are translated into public policy.

Taking Stock: Existing Data and Research
On the whole, then, our knowledge about the link between health status and the way individuals think about and engage with politics-namely, that ill health seems to be associated with lower levels of engagement-is currently based on studies that examine the connection between health and participation in very specific ways. For example, to date, most studies measure wellbeing with the help of self-reported health; they also most commonly examine electoral participation (voter turnout) to measure political engagement. Both make sense for theoretical and practical reasons, as voting is the most common political activity most people engage in during their lifetimes, while self-reported health is a reliable and readily available indicator of wellbeing. 3 At the same time, we do not know if other acts of participation are affected by health in the same way as voting is. Moreover, the finding that ill health demobilises citizens is not uniformly corroborated by the data. In fact, some studies have found that, rather than diminish people's motivation to be engaged, ill health can serve to politicise them (cf. [19,26]). This politicisation arises from self-interest, with the increased activism hypothesis (or what some have called the reversed health gap) predicting that ill health motivates political involvement because the stakes of public policies and the provision of healthcare are higher for those highly dependent on health services [5]. Thus, people with specific health conditions may be more, not less, active in work that involves political parties, contact with policymakers, and political demonstrations to make their voice heard to express grievances or needs [19]. Second, as a matter of social identity, illness and disease patterns may produce a shared experience and social identity, with mobilising networks and political entrepreneurs organising affected individuals to make demands on policymakers [27][28][29][30]. Good examples of such mobilising identities are networks of cancer survivors or associations devoted to raising funding and awareness for particular kinds of diseases.
While research into and thus evidence in favour of this politicisation hypothesis has been much more limited than research on the health gap, it offers a clue that there may be alternative channels for political engagement among some subsets of people with specific conditions. Turned on its head, the politicisation hypothesis may also be useful for capturing a theoretically curious, but empirically well-documented, phenomenon: that younger (predominantly healthy) voters are more likely to abstain from participating in political life. In line with the politicisation hypothesis, younger voters in good health may well have fewer incentives to take political action, believing that the system can be relied upon and serves their interests. Hence, the politicisation hypothesis, while plausible, requires more rigorous and complete testing than has previously been the case. Beyond the two stylised conclusions-that ill health is debilitating for political engagement and that some forms of ill health hold the potential to politicise citizens-we know of few results in this emerging literature that would qualify as uncontested. For example, looking at the correlations of health and various forms of political participation and the attitudes about politics in the sample of European citizens we mentioned at the outset, we see that they can indeed be found in the data. Figure 3, which shows correlations between health and political engagement, reveals that Europeans who report higher levels of ill health and being hampered by illness in their daily lives are, by and large, significantly less engaged with politics: they participate less and exhibit lower levels of happiness with the political system. 4 However, these conclusions are not uniformly supported by the data. While reports of health ('ill health') are strongly and negatively correlated with trust in the political system, satisfaction with the government, or a sense of being able to participate in politics (so-called political efficacy), feeling hampered in one's daily activities is essentially uncorrelated with reports of attending demonstrations or working for political parties. Importantly, those who report being hampered are just as interested in politics as other citizens, while correlations of being hampered and turnout in the last elections are modest at best.
We speculate that part of the reason for why findings in the emerging literature on individual health and political engagement are mixed is conceptual. In particular, there is a need to specify the cognitive and physical requirements and antecedents of specific behaviours and opinions more precisely in order to subsequently establish which dimensions of health matter and in what way. For example, are the critical dividing lines between physical and cognitive requirements, chronic and acute ailments, the social meaning of diseases and disease identities, or differential treatment via the health care system? Answering these questions requires a framework for specifying what we mean by health and political behaviour and to conceive exactly how we believe the two to be connected.

Healthy Citizens: Connecting Health and Wellbeing in the Study of Politics
We start by defining health as a personal property or asset 5 and political engagement as an individual-level activity that involves concrete and observable acts (e.g, voting, demonstrating, joining a group) as well as political orientations, including motivations (e.g., political interest), policy preferences (e.g., left-right orientations), feelings of political efficacy and trust, satisfaction with incumbents, and the like [5]. If political engagement, then, has physical and psychological manifestations and requires the necessary resources, motivations, and mobilising networks, our conceptualisations of individual health and wellbeing need to be defined relative to the political acts we are studying: some aspects of health status involve the ability to perform physical tasks and some involve emotion, others cognition. Importantly, perhaps, there also 4 European Social Survey data have been used in some of the most prominent European studies of health and wellbeing and political engagement (www.europeansocialsurvey.org). Here, we look at political actions and attitudes asked repeatedly over a number of years. Actions include voting, attending demonstrations, and working for a political party; attitudes include people's trust in political institutions and satisfaction with the government, political interest, and political efficacy. 5 We acknowledge that people can also be affected by the health challenges of social others, such as family members or friends. Behaviours are blurry conceptual lines between, say, chronic diseases and disability. Moreover, some forms of action, such as casting a ballot, require different kinds of physical resources than others, such as making an online donation to a political organisation, while some attitudes are more heavily cognitive in nature with others being more related to affective states. In combination, then, health and political acts need to be seen in relation to one another. Similar to Mattila et al.'s [5] discussion of the notion of ' accessibility' of political behaviour, we ask, 'What does this type of engagement require?' Once we can answer that question, we can define a particular kind of health status as potentially being facilitating or debilitating with regard to specific political actions or the formation of particular attitudes. That is, we think of political acts as being hindered or enabled, depending on the specific health concern involved and depending on the person's experience of how their wellbeing is thus affected.
Our generic individual-level model hence links health status and engagement, where the effects of health on behaviour are likely to be conditional on a) the specific attitude or action to be explained, b) the specific health issue(s) at hand, and c) intervening factors, such as demographic factors, socio-economic resources, attitudes (like identity), and mobilising agents and networks. In the aggregate, inequalities arise because individual-level physical and psychological resources necessary for political engagement are distributed unevenly across social groups in a population.
In this way, we expect the influence of health on engagement to be conditional, and in several ways. For one, ill health manifests in different ways and in ways that are differentially relevant to specific actions. Moreover, there are likely to be significant differences between somatic and psychological wellbeing-differences that do not necessarily point in the same direction of more or less involvement. In fact, absent the need to perform a physical task, physical wellbeing may serve to produce countervailing effects on civic attitudes. For example, as discussed above, good health may induce people to believe that all is well and thus produce more positive attitudes about government and politics, while ill health may have the opposite effect and mobilise to action.
These direct effects are interesting but only the beginning of the story. As mentioned above, it is likely that they are conditional on individual traits, characteristics, and attitudes. Thus, political ideology may moderate these effects, as may politically and psychologically relevant identities like gender or socio-economic status. To complicate matters even further, the logic of physical resources and political engagement may not translate neatly to the effects that cognitive resources and abilities may have on people's involvement with politics. While education has long been a potent predictor of physical and psychological involvement with politics, it is yet to be determined if the availability of cognitive resources functions in the same way that physical ones do.
As a last point in this complex chain of relations whose logic we seek to unravel, we expect these relationships to be embedded in particular political, institutional, and cultural contexts and thus to vary as a function of differences across countries. Hence, aggregated to the level of countries, we should see cross-national differences in the strength of the individual-level associations as well as aggregate correlations that are contingent on national factors (such as formal political institutions, political culture, or characteristics of the social policy and health care policy systems). Figure 4 illustrates our main argument: that health and wellbeing directly, as well as in conjunction with other identifiable factors, affect political action on a sustained basis.

A Brief Look at Some Data
In this section we conduct a brief empirical probability probe to see if some of these patterns can be detected in the European Social Survey data mentioned above. The intention is merely to establish whether the overall picture of these correlations holds true in order for future research to then engage with these relationships more rigorously and more in-depth.
In Figure 6, we correlate reports of political action and attitudes with reports of subjective health and wellbeing hampered in daily activities. Importantly we subset these by gender, education, and income, given that they are well known antecedents of political engagement and because they are likely to proxy for differences in the availability of resources to compensate for difficulties arising from bad health outcomes. Figure 5 summarises these factors -health, institutional and socio-economic factors -and their impact on a person's political engagement.
The top row shows correlations of health with political engagement for women and men, the second row show these correlations for respondents with low and high levels of education, and the third row for high-and low-income earners. The patterns are as interesting as they are straightforward. Generally speaking, reports of ill health are negatively correlated with measures of physical and psychological political engagement, particularly when it comes to attitudinal measures of political engagement, including political trust, satisfaction, and a sense of being able to participate in politics. In contrast, correlations with actual reported political activities are significantly lower. Thus, the data show that reported health status is indeed a significant factor in people's engagement with politics, but even more so is a person's subjective sense of wellbeing. Just as importantly for our purposes of demonstrating the complexity of the relations between health and political engagements, there are consistent differences in the strength of these correlations across the different demographic groups. Specifically, we see that ill health has a significantly more debilitating effect on attitudes and behaviours among women and individuals with lower levels of education and income. At the same time, we also see that our measure of reported (subjective) health has significantly stronger correlations with the various outcomes than reports of actually being hampered in daily life because of illness.
On the flipside, we also see that ill health appears to have the potential to be a positive driver of political motivation, especially among the highly educated and individuals with higher incomes. Among these two groups, we see a positive correlation between ill health and political interest, as well as a positive but more modest correlation with reports of having worked for a political party. Taken together, these correlations suggest that the health gap hypothesis has broad applicability, but that health status has different correlations with attitudes and actions as well as across groups of individuals with more or fewer political resources. In particular, it is striking that being hampered in their daily lives appears to mobilise and motivate people of higher socio-economic status to be politically involved to a greater extent than those of lower socio-economic status.

Looking Ahead: Some Suggestions
While these correlations only scratch the very surface of what may lie underneath the relationships between health and political engagement, they point to some of the key challenges to be addressed as part of future research in this area. On a theoretical level, they highlight differences between the dependent variables under examination-action versus attitudes-and why we may see them. More importantly, perhaps, they remind us of the importance to conceptualize and specify the physical and cognitive demands they put on people with different kinds of health experiences. Moreover, on the independent variable side, they raise the question of why subjective health has stronger effects on engagement variables than reported health; this suggests the need to specify how to capture the different kinds of limits that physical and cognitive abilities impose on citizens.
Among the various ways of categorizing these, we would imagine there to be important disparities, depending on whether differences in health are due to chronic or acute ailments, different kinds of chronic ailments, disease identities and stigmas, or how we categorize and incorporate the study of disabilities into the study of wellbeing and political engagement. On the cognitive resources side of the ledger, too, we will need to understand variation across different kinds of mental health conditions, for instance, and the extent to which they may or may not impede cognitive engagement with politics.
Aside from important conceptual and definitional issues, there are important empirical considerations to bear in mind. For one, the more specific and finely grained the disease patterns or impairments we are looking to study, the bigger and more precise our datasets need to be in order to achieve sufficient empirical leverage to tease apart how the different manifestations of ill health matter. Aside from the need for rich and plentiful data, there also is the need to move beyond purely correlational evidence in order to pin down the consequences of ill health for political behaviours. Causality is difficult to establish with existing data sources. Outside the laboratory, longitudinal panel data that trace people's wellbeing and political engagement over the life course may allow us to track changes in individual outcomes. Alternatively, certain statistical techniques can be helpful, provided the data are rich and samples sizeable, to draw inferences at the level of individuals. In particular, so-called matching methods, where we match individuals on all relevant characteristics except for the statistical treatment of interest-here, an individual's health status [31,32]-may be especially useful.
Aside from allowing richer, more comprehensive, and more extensive cross-nationally valid measurement of health status as well as political behaviour, comparing individuals across countries will allow us to establish with greater precision whether the impact of health varies across different health care systems. Do differences in health care systems shape and moderate the connection between health inequality and political inequality? Given the considerable cross-national variation in terms of how health care is funded or the nature and extent of provision [33], it will be important to establish whether people across specific social groups (e.g., age, income, and education) are systematically advantaged or disadvantaged when it comes to the formation and expression of political preferences, depending on the provision of healthcare in a society. Together, these building blocks will allow us to establish whether there is a connection between health and democracy, what form it takes, and whether that connection is spurious or causal. Of course, what policymakers are likely to do about these findings and whether they can serve to articulate proposals to address inequalities of participation resulting from differences in health outcomes has to remain an open question for another day.