The Voter Experience Around the World: A Human Reflexivity Approach

ABSTRACT

The experiences that voters have of elections are pivotal in the democratic experience of citizens. However, there has been relatively few multidirectional theorisations of the nature of this experience and the implications. This article reviews existing canonical approaches to understanding the voter experience which are informed by rational choice theory, behaviouralism and constructivism. It offers an alternative human reflexivity approach which anchors the voter experience in structure-agency relationships using realist social theory. The voter experience is defined as the simultaneous process of gathering and responding to knowledge, perceptions and emotions about the electoral process through observing and (non)participating in electoral activities. The citizen is reflexively situated in this experience and is involved in a process of interpreting, re-interpreting, and responding to stimuli, structures and other actors. Using cross-national data, the article identifies the overall global characteristics of the voter experience around the world. Older and more educated voters tend to have a more positive voter experience. Poor voter experiences are also found to lead to citizens ‘checking out’ of future elections or disengaging from the voting process. The article concludes by setting out the research agenda that arises from the new framework which the special issue takes forward.

KEYWORDS:

• Democracy
• elections
• electoral integrity
• voting
• voter turnout
• voter trust

1. Introduction

Election day is a pivotal moment in the democratic experience of citizens where they have the opportunity to cast their vote (James, 2012; Norris, 2014; Orr, 2015). While the experience of casting a vote can be a ritual involving a close feeling of community and self-expression, it can also be one of frustration with long queues, perceived and/or real racism, or fear and intimidation as party activists threaten would-be voters. The consequences of these experiences at the polls can have wide-ranging impacts, from undermining a voter’s confidence in electoral democracy, to their willingness to return to the polls. Indeed, when elections fail, the wider perception in elections can be one of the distrust, alienation and suspicion, rather than a moment of empowerment, equality and deliberation.

The voter experience has been under-theorised, however. The common approach since the 1960s has been to adopt a rational choice approach in which the voter experience is reduced to a series of cost–benefit calculations (Downs, 1957a). This allows less nuance for the rich diversity of experiences around the world – from voting in peaceful and democratic elections in Cape Verde; to registering to vote in contexts of extreme political uncertainty in the post Capitol Hill riot mid-term US elections; to participating in rallies held during the context of the covid-19 pandemic in India. It also assumes that voters are constrained in their decision-making to follow what is rational for them to do, offers no explanation for why citizens might be pre-disposed towards particular positions – or how they come to be socially and politically constituted as actors on election day. Meanwhile, the greater availability of big data has encouraged the individual voter experiences to be reduced group characteristics and lessons to be drawn from past behaviour through a behavioralist lens. However, humans have reflexivity and an ability to interpret, re-interpret and reconstruct their voter experience reflexively.

Existing frameworks have also tended to focus on the issue of how institutional design can affect turnout, via the voter experience (Rosenstone & Wolfinger, 1978). But they have said much less about how institutions can shape a broader set of outcomes of the election – and how the voter experience – and voter actions – can play an important contingent factor in shaping these outcomes. Existing frameworks also struggle to conceptualise the interaction between institutions, agents (such as parties and candidates) and the wider voter experience.

This opening article to the special issue therefore seeks to re-conceptualise what we mean by the voter experience using realist social theory. This is an approach to social science which has been presented as an alternative to positivism or constructivism by scholars such as Margaret Archer (1995, 1996) and Andrew Sayer (2010). It emphasises the importance of structure and agency relationships in understanding social change. Agents (in our case, voters)¹ find themselves in strategic-relational contexts which are not of their choosing – and which can shape their understandings, options and choices – but they still have some reflexivity to respond to these contexts. We therefore define the voter experience as the simultaneous process of gathering and responding to knowledge, perceptions, and emotions about the electoral process through observing and (non)participating in electoral activities. The experience is one where the citizen is reflexively situated - they are capable of re-interrupting and responding to stimuli, structures and other actors reflexively. In so doing, citizens can be the causes of changes in society and its structures.

This introductory article to the special issue begins by reviewing canonical theoretical approaches to understanding the relationship between voters and their experience of elections. It then introduces the human reflexivity approach as a new and alternative paradigm for understanding the voter experience, alongside the concept of the voter experience cycle and uses a vignette to illustrate how it compares to existing canonical approaches. The analytically distinct stages of the voter experience are set out alongside an analysis of cross-national data to identify how individual and structural conditions impact perception of elections and the actions voters will take at a global level. Finally, we present the special issue ahead. Articles in the special issue help to advance this newly set out research agenda to consider whether there are regularities in how institutional design affect voter experience, or whether they are more context specific. The conclusion article draws together the broader lessons for theory and practice of democracy and elections.

2. Existing Approaches to Understanding the Voter Experience

Existing approaches to understanding the voter experience are closely bound to the mainstream approaches to political science. The main three paradigms are outlined in turn.

2.1. Rational Choice Theory

Much of the existing research on the voter experience, particularly on policy interventions aimed at improving voter turnout through reductions in the ‘costs’ of voting, draw on rational choice theory (RCT). The building block of the theory is methodological individualism: the axiom that analysis should start with the individual rather than society. Rational choice theory posits that voters are rational, self-interested actors, based on work set out by Downs’ (1957b) Economic Theory of Democracy, which was further refined by scholars including Riker and Ordeshook (1968) and Blais and Daoust (2020). The voter experience, according to this prism, is reduced to a series of utility maximising calculations. Voters are making decisions in response to a series of external events, institutional structures and actions of others in order to decide their own behaviour. However, individuals have preferences which are conceived as endogenous. They are assumed to have prior preferences for certain outcomes. They can rank alternative options according to their preferences, which are assumed to be transitory. The theory has been used to explain how voter experience is connected to behaviour, in a wide variety of areas. According to Downs, the decision to vote (or not) will rest on a calculus made by the voter, a multiplication of inputs of whether the vote will ‘matter’ (p) and the expected benefit of one candidate winning over another (B), minus the costs (cognitive, time or material costs) (C) (Downs, 1957b). Political competition influences the pB term, thus the dynamics of parties, candidates, issues and campaigns were seen as important to voters’ attitudes of whether the voting in the election was important to them. Empirical research has focussed the costs (‘C’ term of the equation). Cognitive costs included understanding the procedures and deadlines for registering, knowing when and where to vote, understanding the ballot format and voting procedures, and making a decision from the available options. Time costs include travel time to the polls, line-ups and wait times, and administrative issues like filling out registration paperwork. Finally, voting can have material costs, whether that be time off work to attend a polling station, the costs of acquiring the right forms of identification, transportation expenses, or childcare costs. Innovations to boost voter turnout, such as mail-in voting or easier access to voter identification, were therefore designed to reduce the costs of voting. Later theories would include duty to vote (D) as part of this calculus to capture variables such as civic duty and satisfaction derived from the voting process (Blais & Galais, 2016; Riker & Ordeshook, 1968).

There are three core problems with the approach. Firstly, at the broadest level, methodological individualism has multiple weaknesses. Society is not constructed from free-floating individuals, but societally based actors. Understandings, preferences and choices are constructed in the historical and cultural context in which we find ourselves. A voter entering a polling station in Cape Town will have entirely different prior cultural, political and life experiences to one entering a polling station in Caen or Caracas. These settings exist prior to individuals but are not accounted for by methodological individualism.

Secondly, alongside questions of whether the concept of rationality is plausible, the approach is deterministic. The plausibly of the rationality assumption has long been questioned. The chance of any one voter influencing the results of an election are slim, yet we turn out to vote regardless. The presence of a D term (duty) was argued to capture some of this nuance (Blais & Galais, 2016), and recent work covering motivation also addresses these concerns (Blais & Daoust, 2020), but scholars are aware that voters may not make precisely these calculations in their heads in the choice of whether to vote or not. But as Colin Hay (2002) notes, if voters always behave in ways which are rational, then they have no free-will and agency.

Thirdly, the rational choice approach reduces all aspects of the voter experience to cost–benefit calculations. However, there is much more to an experience that this which it is important to understand and capture. What actually is their experience? What do they see? How do voters feel? What are their fears? What do they not see? What takes place during the electoral process which they are not aware of? How do they reflect on their experience and reshape their perceptions of and participation in the electoral process?

2.2. Behaviouralism

Behaviouralism rose to prominence in the post-war years, in part, thanks to the rise of election studies capturing socio-demographic and attitudinal variables to help political scientists understand voter behaviour. Again considering turnout, but also the choice of who to vote for, the Columbia School (Berelson et al., 1986) suggested that a voter’s experience of the campaign likely had less of an influence than previously expected, since voters pre-existing choices were reinforced through campaign materials and more largely experienced through personal connections and discussions. However, the salience of issues presented could influence their decisions. Scholars from the Michigan School (Campbell et al., 1960) expanded this work with large-scale surveys, and theorised that voters’ choices drew largely from immutable characteristics, including socio-demographics and socialisation, which in turn translated to partisan affiliation.

Studies within the behaviouralist tradition undoubted advanced the scholarship of electoral institutions to an enormous degree, including the voter experience, however there are two underlying criticisms that realism provides an alternative position to (see: James, 2020, pp. 20–23). Firstly, there was often a positivist aspiration to identify iron laws: repeated patterns of voter experience and behaviour that are universal and timeless. However, in different cultural settings the effects might differ. Alternative methods for identifying causal methods have therefore been proposed (Mahoney, 2021; Sayer, 2010). Moreover, knowledge of past underlying patterns of behaviour may cause actors to act strategically in the future. These actions may therefore undermine the findings about the underlying factors that shape the voter experience. Secondly, analysis interprets individual behaviour through an analysis of group characteristics. Hence, there is little scope for agency for voters to reflect on their experience and act creatively (Bevir & Blakely, 2018).

2.3. Constructivism

A third approach is constructivism. Beginning with an anti-foundationalist epistemological and ontological position, constructivism focusses on the social meanings that actors attach to institutions, rituals and practices. The focus is therefore less on developing a causal explanation of the effects of political institutions. The role of the social scientist is therefore to understand and describe rather than to predict and generalise. Ethnography, interpretivism and other qualitative research methods are therefore often the prescribed research methods (Bevir & Blakely, 2018). Constructivist approaches have been widely adopted in many fields of international relations and political science, but have been less used in the study of electoral institutions. Exceptions include important work on electoral ethnography. Banerjee (2017) used ethnographic methods to ‘focus on ordinary Indians’ experience of elections, and on what they mean to them’ (p.4). This important work uses concepts of the social imagination to capture how subjects understand and imagine their social existence, drawing from work by Charles Taylor (2007). Klumbyte (2006) explored how citizens understood their experience at the ballot box in post-Soviet Lithuania, in the context of political memories that had been formed.

A focus on political culture and context stands to improve our understanding of the voting experience and provide much richer and more nuance which cannot be captured through quantitative research. The approach, however, tends to limit opportunities for causal explanation about how broader structural factors shape the behaviour of individuals – especially those outside of the actors’ awareness of them. Moreover, the approach limits the scope for structural conditions such as electoral institutions to be featured in the analysis. Electoral laws such as voter identification requirements are not purely an ideational construct. They exist independent of voter’s understanding of a situation and can have affects on behaviour. However, how they are understood by voter’s is an important frame of analysis. Finally, suggesting that all aspects of election are social constructions tends towards very relativistic analysis in which objective measures such as the fairness of an electoral context does not exist – except in the mind of the voter. But there is a difference in the fairness of an election where it has been rigged, where there has been violence and intimidation.

3. A Human Reflexivity Approach to the Voter Experience

An alternative human reflexivity approach is set out here instead, which is premised on social realism as set out by Margaret Archer.² She developed a theory of structure and agency in a Realist Social Theory that she termed the morphogenetic approach (Archer, 1995). Her work was expanded on in successive monographs which considered how humans could reflect and respond to their social context (e.g., Archer, 2007). Andrew Sayer provided some key texts in enabling the overall philosophy of social science to be used in empirical research (Sayer, 1984). Realism is often perceived as not being applicable to real-world problems and to specific empirical research questions (Kurki, 2010). Scholarship has, however, increasingly sought to apply the approach to more conventional political science topics such as evidence-based policy (Pawson, 2006), political marketing (Savigny, 2007), political leadership (James, 2021a, 2021b), and the design of electoral institutions (James & Garnett, 2020). The approach is thought to be compatible with the use of mixed research methods such as qualitative research, historical comparative research, but also quantitative methods (Porpora, 2001).

The relationship between structures and agents is a key feature of realism. Margaret Archer argued that much of social science had adopted a starting point of either ‘methodological individualist’ or ‘methodological collectivist’ analysis. ‘Methodological individualists’ would include rational choice theorists, for example, who begin with an atomistic approach in which individuals are treated as free-floating individuals separated from society. At the other extreme, ‘methodological collectivist’ sociologists might undertake aggregate-level analysis in which the individual is subsumed into a wider (class, ethnic, gendered, etc.) group without any reflexivity and individual free will. By contrast, Archer conceptualised structure and agency through an ‘analytical dualism’ and focussed analysis on how they interact over time, through a process that she termed a morphogenetic cycle. The approach holds that social structures exist prior to agents at moment T1. They therefore constrain and influence strategic choices, understandings and meanings at moment T2. But structures do not ultimately over-determine the choices that agents make. They are not dictated to follow what is ‘rational’ because they have some reflexivity in how to understand their social experience – and choice between what actions (if any) to take. Human agents, through reflexivity, in turn, contribute towards the creation of new conditions at moment T3. In short, agents are born into social systems which are not of our choosing – but over time, gain some understanding of them and develop some capacity to reshape them.

The idea of human reflexivity is at the heart of social realism and is put forward as an alternative framework for understanding the voter experience in this article. Voters (but also other actors such as party leaders and electoral officials) are capable of reflecting internally on their context and situation and deciding a course of action in response to this. Their situation, however, is not of their choosing (given that structural conditions will already exist). Hence the structural context in which they find themselves, despite their ability and desire to undertake reflection, has the capability of shaping their actions.

A consequence of reflexivity is that there are no iron sociological laws for the voter experience. If strategic choices are possible by a variety of actors, then it is not possible to predict outcomes with complete certainty in the way that it would be if we were studying atoms or chemicals in test tubes. Causation does not work on a do not work on a ‘when A, then B’ manner. Voters, autocrats, political parties, activists and media outlets may undertake strategic behaviour and change their approach in response to a previous election and previously apparent trends. That said, drawing from social realism, we can note that social structures have emergent properties which can have power effects on social actors. Emergent properties are the powers that social phenomenon have when they activated in combination with other phenomenon and actors. In some cases, the emergent properties of structures will be very strong. For example, legally disenfranchising Black citizens during the South African apartheid system, had profoundly strong emergent properties. Although the effects are mediated by social actors (e.g., electoral officials enforcing the laws), the voter experience was profoundly shaped. In other cases, the properties of emergent structures may be weaker, however, allowing individual reflexivity a greater role.

The human reflexivity approach therefore seeks research on how structures and agents collide to shape the voter experience – with a critical concern with how this varies around the world.

4. Defining the Voter Experience

We offer an alternative social realist approach to understanding the voter experience – given it’s place within the morphogenetic cycle. The voter experience is the gathering and responding to knowledge, perceptions and emotions about the electoral process through observing and (non)participating in electoral activities. Electoral activities include receiving information about candidates and parties throughout the electoral cycle, registering to vote, casting their vote. It also includes receiving information about the electoral process itself such as how to vote and evaluative information such as news and claims about the fairness of the election. The experience is one where the citizen is reflexively situated - so that they are capable of re-interrupting and responding to stimuli, structures and other actors reflexively. The voter experience also informs their actions and therefore is connected to the evolution of future social structures.

The voter experience is therefore more than cost–benefit analysis. It involves a richer set of sensual experiences. The experience is also one in which humans are reflexive actors – they are not just an externally imposed phenomenon. Experiences do not happen to people, they have an inner, internal voice which enables them to navigate it (Archer, 2007). However, individuals do not have complete freedom to socially construct and interpret their environment as constructivists might have it - because their prior structural contextual conditions matters.

However, voters are not the only actors to have creative capacities. Other political agents such as parties, candidates, electoral officials and the media also have reflexive capacity to respond reflect on past experiences – and this can also shape the voter experience. Hence the voter experience also embodies the interactions with other agents. In this article, Archer’s morphogenetic cycle is adapted here to understand the development of the voter experience through a voter experience cycle (Figure 1).

Figure 1. The voter experience cycle: a human reflexivity approach.

4.1. The Voter Experience Cycle

In the first stage (T1) of the voter experience cycle, there is structural conditioning. There is an existing societal context which has the capability of shaping individuals through emergence. Citizens may have had some prior influence over this through earlier iterations of the cycle (e.g., through voting in earlier elections). However, at the start of the voter experience cycle these structural influences are conceived as existing analytically prior to the forthcoming voter experience. In the context developed at T1, a variety of actors undertake strategic actions to influence the electoral environment at T2. At this stage, we therefore see the first moment of social interaction between structure and agency. Key actors include incumbents, candidates and parties, electoral officials plus a wider range of stakeholders.

Their strategic activities become apparent to the voter in the voter experience, at moment T3. This experience involves much more than a rational calculation about whether to vote or not. It may involve feelings of fear, intimidation, perceived or actual racism, anger, rejoice or safety. Their actions will involve (not) casting a vote for a specific party or candidate – but also sharing their views and experience with others. All of these activities have already been shaped by societal structures and strategic agency at T1 and T2. They will therefore be shaped by logics of both calculus and culture. However, we emphasise that voters will not all experience this election environment in the same way. The strategic and structural characterises of the context will affect individuals differently. Depending on a host of past individual and group experiences, each voter may interpret the competitive environment, fairness of the competition and other facets of electoral integrity differently.

There is a second moment of social interaction at moment T3 wherein the voter experiences the various stimuli of electoral activities. They exercise reflexive agency by interpreting the information gathered and making decisions to (not) act. These experiences are aggregated at moment T4.

At moment T5, the individual and collective agency of voters can interact with the strategic context and the agency of stakeholders to shape future contexts (T1). There is the possibility that their actions regarding electoral issues will be discussed by decision makers and that law-makers will enact legislative changes to future elections. They might create reverse pluralities through their vote shares, or there might be low voter turnout, campaign rallies against violations of electoral integrity or the deliberate sabotaging of ballot papers. In many cases, however, the main consequence will be the election of a new government who will have the opportunity to reform – or keep – existing structural conditions for the voter experience.

The voter experience at T3 is not limited to the act of casting a vote on election day. It also includes pre-election day activities such as voter registration and campaign rallies, alongside post-election media consumption. It is important to distinguish the voter experience cycle from the electoral cycle, however. The electoral cycle is a well-known tool amongst the electoral community used to differentiate between phases of the election (ACE, 2017). The voter experience cycle presented here, however, represents place of the voter experience in conceptual time – not chronological or physical time. This framework considers the place of the voter experience with respect to structure and agency. T3 thus involves all moments of the citizen experience: pre/post-election including post-election protests – even though chronologically they follow the physical counting of votes because conceptually, post-election protests occur before their consequences at moments T4 and T5.

4.2. A Vignette: Human Reflexivity and Voter ID

A vignette on voter identification is introduced to illustrate how the reflexivity approach departs from traditional approaches. Voter identification laws have traditionally received little academic or policy attention outside of the United States, where their requirements were deeply embedded into historical narratives of voter suppression of minority voters. Historical research has documented that both Republican and Democrat party elites increased the administrative barriers for voting to maximise their chances of winning as part of Jim Crow politics (Key, 1949). Redressing these problems were a central component of the civil rights movement. Civil society groups such as Human SERVE championed reforms to electoral administration to make it easier to vote at the federal level in the 1980s, with academics Frances Fox Piven and Richard Cloward simultaneously researching and advocating easier voting practices which culminated in the National Voter Registration Act 1992 (James, 2012, pp. 96–104). Voter identification requirements increased substantially between 2006 and 2022 (Kuk et al., 2022). Throughout this period, academic and policy debate grew considering the effects of voter ID. Proponents of voter ID have claimed that it is necessary to prevent voter fraud and fraud was widespread (Lott Jr, 2016). At the same time, however, voter ID has been accused by some civil rights groups and politicians as an unnecessary burden and even a deliberate form of voter suppression.

Viewing this through a rational choice lens, voter ID laws alter the costs of voting (time or resources to acquire the proper forms of identification). There is a simple cost calculation effect. The voter experience is simply the calculus undertaken in the mind of the voter about whether to vote or not – given the varied administrative and resource burden. According to the most basic forms of this approach voters must do what is ‘rational’ and cannot deviate from this with any agency. Not all studies take this view – but where writing adds nuances and departs from this simple rational choice approach, there is an implicit acknowledgement of the problem with the rational choice approach.

The reflexivity approach is much more layered (Figure 2). The effects of voter ID still include the additional time involved finding, requesting and providing identification to a poll worker when it is asked for (T1). However, the history of voter identification laws in the US are more than a historical anecdote. The civil rights movement and Jim Crow politics provide the narratives through which media outlets may report the new story, campaigners might seek to ‘Get the Vote Out’, and other actors might seek to suppress the vote (T2). When a Black or Hispanic citizen is confronted with the requirement to show voter identification at the polls in the US or reads/watches news stories relating to voter identification requirements, the historical narratives can shape their understandings of fairness of these requirements (T3). They may accept or reject a narrative of ‘voter suppression’ – but in doing so they are exercising reflective agency. Citizens who interpret the new requirements as ‘voter suppression’ – a deliberate attempt to discourage them from voting – may then respond to the requirements with a new determination to vote.

Figure 2. Voter ID vignette.

The aggregate effects of individual experiences are captured at T4, which might include a decline in turnout – but counter-mobilisation efforts or increased confidence in voting may mean higher turnout. The voter experience may then lead to attempts at reform at T4 – with states seeking to remove or increase the identification requirements. This cycle has had consequences beyond the US. The narrative that voter identification could be being used by incumbent governments has spread from the US to other countries such as the UK – where pressure groups, opposition MPs and columnists accused the Conservative government of introducing voter ID for partisan advantage (Toynbee, 2022).

A reflexivity approach importantly considers academic knowledge about voter ID and experience as not independent from the actual object of study. The research questions to be addressed are also shaped by the historical narratives, the actions of agents and the voter experience. Academic studies can be reported in national news media (Gelman, 2018), which in turn, can be used by agents to reflect their strategies and shape the voter experience, but not as a rational accumulative development of scientific knowledge, as understood by behaviouralists. Instead, academic researchers become actors in the plays they are writing. This may further contribute to variations in the scientific findings: actors are responding reflexively to the studies themselves.

5. The Voter Experience Around the World

As the voter ID vignette demonstrates, our alternative approach opens up a new research agenda on the voter experience, as we ask how the stages of the voter experience cycle influence each other. In this introduction, we now demonstrate, in broad comparative strokes, how the tactics of parties, electoral officials and observers influence the voter’s individual and group experience, and thus also a variety of outcomes, including their choice of whether to vote or not, their trust in elections, and their perceptions of the role of elections in democratic governance. We ask three main questions: What shapes the voter experience (T1-3)? What is the nature of the voter experience at T3? And (how) do voters reflect on their experience and reshape their participation in and perceptions of the electoral process (T4)?

This article now uses cross-national data to map out broad global nature of the past voter experience and the relationships between stages of the cycle. Articles in the remainder of the special issue follow up this research agenda by focusing on specific country contexts and/or parts of the voter experience.

5.1. T1: Societal Structures

Which societal structures affect the voter experience at moment T1? There was a strong literature which points to how structural contexts play an important role in shaping cross-national and within country political outcomes and experiences. Theories of development, political economy and sociology centre around how socio-economic configurations shape the political. Modernisation theories, as initiated by Lipset (1959), point to structural factors such as technological innovation, productivity, occupational specialisation, urbanisation and cultural values being connected. As societies undergo processes of modernisation, involving a transition from agrarian to industrial capitalist societies, a move from dictatorship to democracy accompanied these changes. Industrialisation, for example, has been argued to generate a working class, who unionise to develop movements for mass franchises. Levels of economic growth, literacy, technology are therefore important structural features with which electoral actors can interact to create contextual conditions. Political agency prevents history being uni-dimensional, but the emergent power of these structural factors is important to explore.

The overall regime structure is therefore likely to have a profound effect on the overall dynamics of how political systems function. Autocratic regimes are characterised by rulers having previously had excessive power and instruments available to them to maintain rule. By comparison, political systems where civil liberties are enshrined in law and respected in practice will structure the actions of electoral stakeholders. Where societies are more open to civil society and there have previously been opportunities for activism, voters would be more empowered to vote. The broadest characteristics of political regime and history of a country or region is therefore important in shaping the voter experience.

The broader constitutional system also provides an important structural context (Norris, 2015, pp. 113–132). Constitutional systems, whether they are majoritarian or contain multiple veto points, are federal or unitary, are argued to structure incentives, normalise and legitimise modes of political behaviour (Lijphart, 1999). There are also more specific structures relating to election. Electoral rules are important structural conditions for elections. Electoral laws prescribe who can (and cannot) vote, how boundaries are drawn and how votes should be converted to seats (Rae, 1967). The organisational characteristics of electoral management bodies (such as statutory independence, capacity and workforce features) are known to have consequences for election quality (James et al., 2019). Each of these are also in interaction with broader structural conditions where they were created through previous agency.

5.2. T2: Strategic Agency

In this context developed at T1, a variety of actors undertake strategic actions to influence the electoral environment at T2. Incumbents, for example, may work through the existing state structures to influence election laws to their benefit before the election. For example, there might have been attempts to prevent certain individuals or groups from running for office or casting a vote through voter suppression tactics – such as removing names from the electoral roll or making it more difficult to vote (James, 2012). In Table 1 we see this played out in the broadest strokes through cross-national regression analysis with only two independent variables. The Perceptions of Electoral Integrity Index captures the quality of elections across a variety of stages of the electoral cycle, and for our purposes, across a variety of actors. It is measured first in individual questions on a 5-point (Likert) scale and aggregated into 0–100 indices (Garnett et al., 2023) (see Appendix). Table 1 shows that the above structural variables – namely the dynamics of the regime (V-Dem Polyarchy Index, on a 0–1 scale) and the level of development (GDP PC PPP (from Coppedge et al. (2023))) – are related to strategic actions of key electoral players. In Model 1, we suggest this is seen by incumbents working with the existing regime structure to influence elections for their benefit – the more restrictive the regime to competition and civil liberties, the more electoral laws will be built up to favour the incumbent regime.

Table 1. Structural conditions influence electoral actors.

	(1)	(2)	(3)	(4)	(5)
	Laws favoured incumbents	Voters were threatened with violence at the polls	Elections triggered violent Protests	EMB Index (higher number is more positive)	Campaign Media Index (higher number is more positive)
V-Dem Polyarchy Index	−3.28***	−2.19***	−1.34***	68.77***	37.61***
	0.20	0.21	0.20	3.43	3.21
GDP PC PPP ($1000s)	0.01**	−0.02***	−0.01***	0.17***	0.04
	0.00	0.00	0.00	0.04	0.04
_cons	4.77***	4.03***	2.85***	18.11***	26.35***
	0.11	0.12	0.11	1.88	1.79
N (Elections)	520	519	520	520	520
N (Countries)	159	159	159	159	159

Notes: Standard errors in second row *p < 0.10, **p < 0.05, ***p < 0.01. Cross-national multi-level regression (elections nested in countries).

Any candidates and parties (incumbent or opposition) may likewise have decided to deploy tactics from the menu of manipulation which could include organised vote rigging, voter intimidation, or not accepting the results of the election and turning to violence instead to resolve disputes (Schedler, 2002). In Models 2 and 3, we focus on the use of violent tactics by candidates and parties. Again, we see that the resources available to these parties and candidates, plus the openness of the regime to disputes, seem to relate to whether they will resort to these types of violent tactics to have their voices heard.

Electoral officials (or electoral management bodies, ‘EMBs’) are themselves actors who perform within the constraint of their organisations, but who have strategic agency. They may deploy resources successfully – or fail to do so (James, 2020). To illustrate, in Model 4, the additive EMB Index considers how an EMB fulfils their roles of delivering impartial elections, informing the citizenry and allowing for public scrutiny. We see evidence that both the resources at the country’s disposal and the level of democracy are related to the quality of their conduct. We suggest that if the country is in a low-resource environment then the EMB was also likewise working within low resource constraints, which makes it more difficult for them to deliver quality elections (Figure 3). It is similar if the ideals of democracy set out in the Polyarchy index are not achieved.³

Figure 3. Relationship between GDP and EMB index. Predictive Margins with 95% confidence intervals, from Table 1, Model 4

Other actors include observers who monitor the election, citizen observation groups, paramilitary groups or civil society organisation running voter registration drives. Important too are journalists, reporting on the news, and the moderators of media platforms, deciding how and whether to moderate news and social media. We suggest model 5 points to the potential influence of the environment in which they are acting on the impartiality of their reporting.

While these models cannot determine a direction of causality between these key factors of wealth, quality of democracy and quality of elections, it is nonetheless clear that these variables work together to create the environment in which various stakeholders choose to act, and voters experience electoral events.

5.3. T3: Voter Experience and Agency

Which political structures (T1) and agency of other actors (T2) affect the voter experience at T3? Political structures such as electoral laws and cultural norms can have biases which disadvantage different genders, for example, as they can be patriarchal in nature (Lovenduski, 2005). The adoption of electoral laws to redress this is therefore often argued to be important (Hughes et al., 2019). Resources such as education and income have been argued to be important for enabled citizens to politically mobilise (Verba et al., 1995) – but these are unevenly distributed amongst populations. Structures can also disadvantage ethnic or racial groups (Bonilla-Silva, 1997) – with voting procedures in particular often argued to be deliberately designed to discourage participation with active attempts to prevent some ethic groups from participating (Piven et al., 2009). Likewise, political structures have often been argued to disadvantage young people. Elected representatives and party leaders tend to be older members of the population and manifesto pledges geared towards the grey vote. The voting experience is often claimed to not be designed to fully engage younger citizens – who might prefer more digital methods of participating (Pickard, 2019; Xenos et al., 2014). Structures do not determine outcomes, because political agency matters, but agency and outcomes are still shaped by structural context.

The World Values Survey (Haerpfer et al., 2020) provides broad comparative insights into voters’ experiences of the electoral process, through a battery of questions asking for their perceptions of the quality of elections in their country: from the quality of TV news coverage, to whether voters were offered a genuine choice at the polls. These data are different from other measures of electoral integrity (for example PEI, V-Dem) in that they provide the voter’s reflections on these stages of the electoral cycle. These perceptions are based on first-hand experience at a polling location or watching the coverage of the election, or second-hand experience, be that what they hear through their friends and family or on the news. Some of the indicators do tap directly into the voter’s direct personal experience (ex. their consumption of the news media would inform whether ‘TV news favors the governing party’ and ‘journalists provide fair coverage’ and their experience of choosing who to vote for should inform whether ‘voters offered a genuine choice’), while others rely on perceptions of the wider experience in their country (for example, they likely are not watching the vote counts in-person, but are reflecting on the activity in a wider sense). And while ideally, we would be curious to know whether the respondent themselves had experience a type of electoral malpractice (for example whether they were personally bribed or threatened), this battery is our best approximation cross-nationally of how the respondent thinks the election was experienced by citizens like themselves. They are measured via four response options, here modelled as a 1–4 scale, with 4 denoting the action occurs most often (see Appendix).

The models in Table 2 present, in broad strokes, how individuals with a variety of different characteristics have experienced the same election quite differently. While we certainly cannot include all the major individual and group-level factors that will influence how an individual experiences an election (for example, ideology and minority status undoubtedly have an impact, but are difficult to study cross-nationally for the purposes of this introduction), we do provide some of the key individual socio-demographic factors that may be important, while nesting individuals within countries and including fixed effects to hold the structural factors mentioned earlier constant.

Table 2. Citizens perspectives on their experiences of elections.

	(1) +	(2) +	(3) +	(4) +	(5) -	(6) -	(7) -	(8) -	(9) -
	Votes are counted fairly	Journalists provide fair coverage of elections	Election officials are fair	Voters are offered a genuine choice in the elections	Opposition candidates are prevented from running	TV news favours the governing party	Voters are bribed	Rich people buy elections	Voters are threatened with violence at the polls
Age/10 (Decades)	0.03***	−0.05***	−0.03***	−0.05***	0.01***	0.04***	−0.03***	−0.03***	0.03***
	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
Female	−0.03***	−0.01**	−0.04***	−0.02***	0.02***	−0.03***	−0.03***	−0.01	0.01
	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01
Education Secondary	0.00	−0.00	0.06***	0.10***	0.06***	−0.01	0.05***	−0.03***	0.05***
	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01
Education – Post-Secondary	0.02**	−0.02	0.11***	0.14***	0.02	0.01	0.05***	−0.05***	0.02
	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01
Income Level	0.02***	−0.00*	−0.01***	−0.01***	0.00	0.02***	−0.02***	0.00*	0.01***
	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
_cons	2.61***	2.42***	2.82***	2.70***	2.54***	2.48***	2.79***	2.10***	2.70***
	0.02	0.02	0.02	0.02	0.02	0.02	0.02	0.02	0.02
N (Individual)	77,172	73,947	71,678	75,560	76,142	76,181	74,768	75,119	77,175
N (Country)	58	58	58	58	58	54	58	58	58
R-sq	0.00	0.01	0.01	0.01	0.00	0.00	0.00	0.00	0.00

Notes: Standard errors in second row *p < 0.10, **p < 0.05, ***p < 0.01. Multi-level regression models (individual nested in country) with country-fixed effects. Reference category – Education (Less than secondary). The symbols + and – after the model number denote whether the statement indicators a positive trend for electoral integrity or a negative trend for electoral integrity.

Consistent with previous research, we find that older respondents have had a more positive voter experience (Norris et al., 2019). Income level also has a rather high level of consistency in relationship with having a more positive experience of the election. The results for gender and education level, however, are less consistent across all indicators.

To illustrate these divergences of experiences, consider Model 7. Younger voters and those of lower income levels are more likely to believe that voters were bribed. We expect this was based on their own and their community-level experience of elections. Voters at lower income levels, for example, may have been more susceptible to bribery due to economic constraints (Carlin & Moseley, 2015). In sum, these models demonstrate variations in the past experiences of elections for citizens, despite participating in the same election.

5.4. T4: The Aggregate Effects of the Voter Experience

The aggregate effects of the individual voter experience and action become clear at moment T4. Much literature suggests turnout as a key outcome that can be impacted by the voter experience. One can easily consider how the experience of an election may influence a voter’s decision whether or not to turn out to vote. For example, in a classic rational choice theory, if a voter believes that their vote will not be counted fairly, there is little incentive to go cast a ballot (Birch, 2010). Recent research, however, has demonstrated that this doesn’t always pan out in all situations (Martinez i Coma & Trinh, 2017) – again demonstrating that context matters. That even in places where elections are clearly lacking in electoral integrity, turnout can be high, due to other factors, such as fear of reprisal for nonvoting or a desire to work within the existing system to gain some material benefits (Golosov, 2017).

The voter experience also shapes broader trust in the electoral process, and consequently also satisfaction in democracy as a whole (Garnett, 2019). Thus, the aftermath of elections takes place not only in the counting of the ballots and declaration of the results, but also in the minds of voters – whether they accept the count and results. The consequences of this, as seen, for example, in the January 6th riots on Capitol Hill in the United States, can extend beyond the minds of voters into concrete actions, be that violent protest or peaceful acceptance. The voter experience also affects a voter’s future interactions with elections and the political system in general. For example, at an individual level, a voter might not ever forget an incidence of harassment. A protest or low turnout might shape the cultural legitimacy of certain electoral practices which can have broader consequences.

Tables 3–5 show the impact of three facets of the voter experience - whether votes are counted fairly, whether voters are bribed, and whether voters were threatened with violence at the polls - on turnout and attitudes towards elections.⁴ Each of the models provides evidence that even when controlling for the country’s unique structural factors (country-level fixed effects), and the individual-level factors mentioned earlier that influence how a voter experiences an election, we find that the voter’s experience of the election in the case of these three variables – how they see the votes being counted, whether they perceive widespread bribery, and whether they fear conflict at the polls – in turn influence these outcomes. Poor experiences lead to lower voter turnout, lower confidence in elections and lower perceptions of the role of elections in democratic life. In other words, a poor experience can lead a voter to ‘check out’ of the electoral process. Conversely a positive experience of elections, can lead the voter to participate and trust the results.

Table 3. Votes counted fairly.

	(1)	(2)	(3)	(4)
	Voter turnout (Binary)	Confidence in Elections	Difference honest elections make in life	Importance of free elections for democracy
Age (/10, decades)	0.36***	0.02***	0.04***	0.08***
	0.01	0.00	0.00	0.01
Female	−0.13***	−0.04***	−0.02***	−0.12***
	0.02	0.01	0.01	0.02
Education Secondary	0.09***	−0.13***	0.08***	0.19***
	0.04	0.01	0.01	0.03
Education – Post-Secondary	0.55***	−0.13***	0.16***	0.47***
	0.04	0.01	0.01	0.03
Income Level	0.02***	0.02***	−0.00***	−0.01**
	0.01	0.00	0.00	0.00
Votes are counted fairly	0.40***	0.29***	0.09***	0.22***
	0.01	0.00	0.00	0.01
_cons	−0.77***	1.49***	2.99***	6.82***
	0.14	0.02	0.02	0.05
N (Individual)	73,489	74,966	76,640	75,847
N (Country)	57	57	58	58
R-sq		0.10	0.02	0.01

Notes: Standard errors in second row *p < 0.10, **p < 0.05, ***p < 0.01. Model 1 – multi-level logistic regression (individual nested in country), Models 2–4 – multi-level regression models (individual nested in country) with fixed effects. Reference category – Education (Less than secondary).

Table 4. Voters are bribed.

	(1)	(2)	(3)	(4)
	Voter turnout (binary)	Confidence in elections	Difference honest elections make in life	Importance of free elections for democracy
Age (/10, decades)	0.36***	0.02***	0.04***	0.09***
	0.01	0.00	0.00	0.01
Female	−0.14***	−0.05***	−0.02***	−0.13***
	0.02	0.01	0.01	0.02
Education Secondary	0.09**	−0.11***	0.08***	0.19***
	0.04	0.01	0.01	0.03
Education – Post-Secondary	0.55***	−0.10***	0.16***	0.48***
	0.04	0.01	0.01	0.03
Income Level	0.02***	0.02***	−0.00**	−0.01
	0.01	0.00	0.00	0.00
Voters are bribed	−0.03***	−0.16***	0.02***	−0.01
	0.01	0.00	0.00	0.01
_cons	0.37***	2.66***	3.18***	7.44***
	0.14	0.02	0.02	0.05
N (Individual)	71,931	73,420	75,081	74,350
N (Country)	57	57	58	58
R-sq		0.04	0.01	0.01

Notes: Standard errors in second row *p < 0.10, **p < 0.05, ***p < 0.01. Model 1 – multi-level logistic regression (individual nested in country), Models 2–4 – multi-level regression models (individual nested in country) with fixed effects. Reference category – Education (Less than secondary).

Table 5. Voters are threatened with violence at the polls.

	(1)	(2)	(3)	(4)
	Voter turnout (binary)	Confidence in elections	Difference honest elections make in life	Importance of free elections for democracy
Age (/10, decades)	0.37***	0.03***	0.04***	0.08***
	0.01	0.00	0.00	0.01
Female	−0.14***	−0.05***	−0.02***	−0.14***
	0.02	0.01	0.01	0.02
Education Secondary	0.09***	−0.13***	0.08***	0.18***
	0.04	0.01	0.01	0.03
Education – Post-Secondary	0.57***	−0.13***	0.16***	0.47***
	0.04	0.01	0.01	0.03
Income Level	0.02***	0.02***	−0.00**	−0.01**
	0.01	0.00	0.00	0.00
Voters are threatened with violence at the polls	−0.07***	−0.08***	−0.04***	−0.30***
	0.01	0.00	0.00	0.01
_cons	0.39***	2.42***	3.32***	8.04***
	0.14	0.02	0.02	0.05
N (Individual)	71,485	73,071	74,659	73,985
N (Country)	57	57	58	58
R-sq		0.02	0.01	0.02

Notes: Standard errors in second row *p < 0.10, **p < 0.05, ***p < 0.01. Model 1 – multi-level logistic regression (individual nested in country), Models 2–4 – multi-level regression models (individual nested in country) with fixed effects. Reference category – Education (less than secondary).

5.5. T5: Structural Reform

At moment T5, the agency of voters can interact with the strategic context and the agency of stakeholders to shape future contexts (T1). There is the possibility that their actions electoral issues will be discussed by decision makers and that law-makers will enact legislative changes to future elections. Space does not permit empirical analysis.

6. Conclusions

The experiences that voters have of elections are pivotal in the democratic life of citizens. However, there has been relatively few multidirectional theorisations of the nature of this experience and the implications. This article reviews existing canonical approaches to understanding the voter experience and offers an alternative human reflexivity approach. This provides a new definition of the voter experience and has provided a temporal conceptualisation of it - and the linkages between other aspects of the electoral process. The voter experience is the gathering of knowledge, perceptions and emotions about the electoral process through observing and (non)participating in electoral activities. The experience is also one in which the citizen is reflexively situated so that they are capable of re-interrupting and responding to stimuli, structures and other actors reflexively.

The new framework re-raises a research agenda to explore the linkages in the conceptual model. The key questions, which are taken forward in this special issue are:

• What is the nature of the voter experience? To what extent does this vary worldwide? Are there common trends and experiences?

• What shapes the voter experience?

• How do voters reflect on their experience and reshape their participation in and perceptions of the electoral process?

• What, therefore, are the consequences of specific forms of voter experiences?

• How can research on the voter experience be used to improve it?

This introduction has therefore commenced the research agenda to provide a recent mapping of the overall global characteristics of the voter experience around the world. Older and more educated voters have tended to have more positive voter experiences. Poor voter experiences are also found to lead to citizens ‘checking out’ of future elections, in terms of actions like turnout and broader feelings of trust in the system. This points to pressing problems which researchers and policy makers will need to address as part of the overall effort to protect democracy and prevent democratic backsliding in the years and decades ahead.

Subsequent articles focus in on different parts of the voter experience, how voters and other actors navigate this and the consequences for democracy. This includes the requirement to produce voter identification or nature and extent of electoral violence. Articles also consider the role of parties trying to bribe voters and the (in)effective role of the financial transparency system. They also include the legal design of elections such as whether elections are being held concurrently, and whether there are short or long voting hours available. The conclusion then provides a rejoinder of the lessons learned and the consequences for the study and practice of elections.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Social Sciences and Humanities Research Council of Canada [grant number 890-2020-0006].

Notes on contributors

Toby S. James

Toby S. James is Professor of Politics and Public Policy at the University of East Anglia UK and Distinguished Fellow at the School of Policy Studies at Queens University, Canada. He is also co-Director of the Electoral Integrity Project. His research covers the quality of democracy, electoral integrity, electoral management, voter registration, public policy and political leadership.

Holly Ann Garnett

Holly Ann Garnett is an Associate Professor of Political Science at the Royal Military College of Canada and cross-appointed at Queen’s University. She is co-director of the Electoral Integrity Project and studies issues relating to the quality of elections in Canada and around the globe, including electoral management bodies, registration and voting, cyber-security, and election technology, civic literacy, and campaign finance.

Notes

1 We use the term citizen and voters interchangeably throughout the article. We refer to the term ‘voter experience’ because our focus is on elections – rather than citizen’s experience of democracy as a whole. Citizens who decide not (or who unable to) to cast a vote, however, are still a focus of our study.

2 The terms social realism, critical realism and scientific realism are often used interchanging in the literature – although it is also argued that there important differences.

3 ‘The electoral principle of democracy seeks to embody the core value of making rulers responsive to citizens, achieved through electoral competition for the electorate’s approval under circumstances when suffrage is extensive; political and civil society organizations can operate freely; elections are clean and not marred by fraud or systematic irregularities; and elections affect the composition of the chief executive of the country. In between elections, there is freedom of expression and an independent media capable of presenting alternative views on matters of political relevance. In the V-Dem conceptual scheme, electoral democracy is understood as an essential element of any other conception of representative democracy — liberal, participatory, deliberative, egalitarian, or some other.’ (V-Dem Codebook)

4 These three perceptions were chosen for their direct relationship to the voter’s experience in casting a ballot, and thus likely one of the areas their direct experience will most influence their beliefs and actions.