Introduction to political behaviour

De Baripedia

The study of political behaviour is not only the study of behaviour as such, but also political opinions, political attitudes, beliefs and values more broadly, all of which are part of political behaviour.

The word "behaviour" is a bit wrong. We have a field of study that is much broader than behaviour as such. This field also extends to opinions, beliefs and values without necessarily transforming them into behaviour. It is not only behaviour and action, but also attitudes, opinions, beliefs and values that are behind it.

We will become aware of the discipline and give some insight into what we are studying in political behaviour.

Two main areas of study

There are two main areas of study in political behaviour that can be summarized by saying that there is conventional political behaviour on the one hand and unconventional political behaviour on the other.

Conventional political behaviour

What is called conventional political behaviour, is also called electoral behaviour. The purpose is to study the behaviour of electors when there are elections. First, to study political participation, that is, who participates and who abstains for what reason; and second, who votes and how.

It is possible to simplify this field of political behaviour by saying that there are three fundamental questions, namely who votes, how and why. We observe who votes, what people vote, and then we try to explain why people vote and why they voted for this or that party.

As we are in Switzerland, it is necessary to broaden the notion of electoral behaviour a little, because, as its name suggests, electoral behaviour refers to elections, so we study electoral behaviour, who votes, who votes, for which party and for which candidate, but in Switzerland, there is an important direct democracy which means that we do not only vote for elections, but also on concrete objects, public policies, proposals, political reforms among others, and therefore, we can also apply the study of electoral behaviour to behaviour in popular voting, namely the study of behaviour in direct democracy votes. So, all the questions we ask ourselves about electoral behaviour can also be asked when we study voting behaviour in direct democracy votes, i.e. popular votes in Switzerland.

Non-conventional political behaviour

Non-conventional political behaviour makes it possible to invoke two types of collective action: protest politics and the new social movements that belong to this field.

Collective action is the field that encompasses the whole. Collective action refers to the collective mobilization to defend common interests. A group of citizens is mobilizing to defend common interests. This field of political behaviour studies how these behaviours are formed.

Within collective action, we can define a little more specifically what is called the protest policy. The protest policy is a set of actions by groups that wish to make a claim and therefore make demands to the government, parliament or other types of decision-makers. It is, therefore, a group that wishes to make a claim and therefore alert the authorities.

The protest policy can take different forms. It can take the form of social movements, but also of revolt, civil war, terrorism; all means aimed at bringing these demands to the attention of the general public and, if possible, at influencing policies.

The third level within political action and protest politics is the so-called new social movements. We say "new social movements" in distinction with "classical social movements" such as trade unions. These new social movements are, for example, the ecologist movement, the pacifist movement, the gay movement, these are movements that are created to defend the interests of a specific segment of the electorate or to defend a cause such as the environment in the case of the ecologist movement.

This mobilization through social movements takes unconventional forms and that is why it has been distinguished from conventional political behaviour. This is, for example, the demonstration, the strike, the boycott, therefore forms of collective action that differ from the institutional channels of voting, the collection of signatures to launch referendums or initiatives.

Institutional channels will lead to conventional political behaviours such as signing petitions, signatures in order to launch initiatives or referendums. This is distinguished from non-conventional behaviours such as strikes, demonstrations or boycotts.

Examples of questions that are asked

To know what we are studying, when we cover political behaviour in Switzerland and abroad, these are the kinds of questions we ask ourselves:

  • To what extent does age influence participation in elections and polls? There is a whole stream of literature that focuses on political participation and the variable "age" is a key variable to explain participation. The effect of age is not only the effect of aging, but it is also the effect of the life course and it is also the generational effect, but also the fact of belonging to a specific generation. All this is combined, there is this age effect on political participation. Why do some people engage in social movements and others not? In other words, are there individual predispositions that make people more or less willing to engage in collective action and social movements?
  • What are the main individual determinants of electoral behaviour? This question is how can we explain how which segment of the electorate votes more for such a party, are there regularities that can be identified to better understand why certain types of people according to their age or social class or political value tend to be more for one party than another?
  • How can we explain the rise of populist right-wing parties in Europe? It is a very large field now in literature. We are trying to understand, to find regularities, kinds of rules that would make it possible to decipher the vote for these right-wing populist parties. This is, for example, to know how parties such as the SVP have been so successful and whether the explanations we have in Switzerland also apply to similar parties elsewhere in Europe, whether the same causes produce the same effects, whether there are regularities behind the rise of right-wing populist movements in Europe, which is an important issue addressed in political behaviour.
  • To what extent does associative engagement influence the integration of foreigners? This is a research that Marco Giugni and Matteo Gianni are currently conducting to see if the involvement of foreigners residing in Switzerland in associations has an influence on the type and degree of integration of foreigners. The question is whether we can ensure an integration model through associative integration.
  • What is the impact of citizenship models on the mobilization of immigrants in European countries? This is an international research, we would ask ourselves the question because there are different models of citizenship, some involve land rights, others blood rights, some are very liberal in integration, others are very restrictive, and we are trying to find out if this has consequences on the degree of mobilization of immigrants in these countries concerned.
  • To what extent do election campaigns and the media influence the formation of opinions before an election or a vote? It is a dynamic perspective in which we are interested in the way in which citizens form their opinions before a vote or election and therefore in the way in which this formation of opinion is influenced by the environment and by the referendum election campaign. The idea is to know if opinions were formed in advance and we knew in advance what they were going to vote and the campaign didn't have much effect or, on the contrary, do campaigns have a massive role in opinion formation.

We will leave aside non-conventional political behaviour and focus on conventional political behaviour.

Three main models for explaining the vote

The literature in this field includes three classical grandes écoles d'explication du vote. These three schools date from the beginning of the 20th century or the first half of the 20th century, so they are already more than fifty years old, which is why we speak of a classic school of explanation of vote. However, there has been an evolution towards other models of explanation of vote. But it is important to start with these grandes écoles, which still have a great model for explaining the vote.

Political behaviour is a relatively young discipline. It is a relatively young discipline because it is linked to the availability of data. Opinion polls were born in the 1920s, 1930s and 1940s. Until then, when we wanted to study political behaviour, we had to do so on the basis of aggregated data, i.e. the results of elections or votes per canton or municipality, for example. The distribution of results by commune was studied. But this is at the aggregate level, i. e. in general, for a community there were no survey data available for a very long time to study political behaviour at the individual level, i.e. each individual separately. This explains why this field of political behaviour is relatively recent and has developed since 1945 and 1950.

Socio-structural school

Paul Lazarsfeld.

The first major school of explanation of vote is also called the Columbia School because it was developed at Columbia University by several researchers, including a famous researcher, Paul Lazarsfeld. Lazarsfeld conducted the first serious non-commercial scientific opinion survey in the United States. It should be noted that it did not do it for the whole country, in this case, it focused on a county in the state of Ohio. It was a study limited in its geographical scope, namely only one county in an American state, but which was on the other hand very impressive in terms of its research design since Lazarsfeld conducted a panel survey in six waves. This means that he interviewed the same people six times in a few months or years. This is called a panel survey, also called a "longitudinal" survey. So, for the first time, there were data that no one had had before in studying voting behaviour and opinion formation at the individual level.

Lazarsfeld's study focused on the 1940 presidential elections. He tried to understand the reason for the vote, namely, why some voters voted Republican and why some voters voted Democratic. What interested him was the explanation after the vote, he had not been interested in predictions. Today, we are seeing more and more results of opinion polls, at least in the media to tell us what the outcome of the upcoming election will be, that is to say, that forecasts are made that are predictions. In this study, as in many other scientific studies, the purpose is not to make predictions, not to make predictions, but to try to understand afterwards why people voted this or that.

To go directly to the essence of its conclusions, the results of this study founded the sociostructural model, known as the Columbia model, which, as its name suggests, emphasizes the importance of socio-structural factors in explaining the vote. One of the key conclusions of this study is that "a person thinks, politically, as he or she is socially. Social characteristics determine political preferences. As the use of words indicates, this model of explanation of vote is very deterministic in nature, such as "tell me who you are and I will tell you how to vote". Under this model, individuals know well before the vote what they are going to vote. Moreover, almost by definition, this knowledge of what people will vote is stable over time because the integration of an individual into his or her social context is relatively stable and therefore so is his or her vote. There is a high stability of the vote due to the stability of the insertion.

In this model, the determinants of voting are socio-demographic or sociostructural characteristics such as socioeconomic status, such as education level, income or social class; religion and place of residence.

Once we know these three characteristics and as long as they complement each other, we know for whom more or less Americans were going to vote at the time. In this model, the vote is highly predetermined, there is a very high predisposition of the vote according to the characteristics of the group to which an individual belongs. Thus, there is a pre-structuring of the vote according to the social and socio-economic characteristics of the group to which an individual belongs.

There is a link between this model of explanation of vote and the literature on cleavages. The idea is that if a cleavage is prominent, if an individual identifies himself in this cleavage as in a religious cleavage that would oppose Catholics and Protestants, then simply knowing the characteristics of the individual on this religious dimension makes it possible to know more or less correctly who the individual will vote.

In Switzerland, historically, in the Catholic cantons, there was a very strong opposition between individuals who practiced religion, and those who did not, namely the laity. Everyone was Catholic, almost everyone was a believer, but not everyone practiced diligently and the distinction was not practiced. This division was politically reflected in the opposition between the Christian Democratic Party and the Radical Liberal Party. Practitioners easily voted PDC and lay people easily voted radical liberal. It wasn't as caricatured as that, but barely. Voting could fairly easily be anticipated based on knowledge of religion and religious practice. In the non-Catholic cantons, the dividing line was different, it was between Catholics and Protestants. Catholics voted PDC and Protestants voted radical or perhaps socialist and more recently SVP.

Psychosociological school

The second major model that also followed the Columbia model over time is what has been called the Michigan model because it was developed within the framework of the University of Michigan, which conducted the first American opinion surveys at the national level. Lazarsfeld focused on a county that is Ohio, Michigan conducted the first presidential scientific opinion surveys at the national level. This then led to the American electoral studies project, which is still being piloted from the University of Michigan.

For the Michigan School, the key explanatory factors for understanding electoral behaviour are not the socio-geographical characteristics as claimed by the Columbia School, but what they call "psychobiological" factors. Unlike Columbia School, which focused on the individual inserted into its group, Michigan School focuses on the individual as such with his or her psychosociological orientations. More precisely, the key variable at the heart of the Michigan model is partisan identification, which is identifying with a party, feeling close to a political party.

Identification is a psychological emotional attachment to a party. In the theorization proposed by the Michigan School, we identify with a party very early in life, in adolescence through political socialization within the family. There is a very high intergenerational transmission that causes an adolescent to be influenced by his parents' preferences and in the case of political socialization within the family, he acquires this identification with a party that then only strengthens itself with age according to this model.

As in the first model, the emphasis is on stability of preferences, there is a kind of lasting loyalty to a party that will then influence the choice of that party when the electorate expresses itself.

In this model, the main determinant of voting is its loyalty to a party that is a lasting emotional trait. The idea is that this partisan identification works as a kind of cognitive shortcut. The world is a complex place, ordinary citizens cannot fully master the whole complexity of what are the right solutions to all the problems that exist, so they will rely on their partisan loyalty. They will use their partisan loyalty to simplify their representation of the world and guide their electoral choices, read will vote according to the party they feel close to and think it will be able to solve problems. In other words, the average citizen relies on shortcuts of information because we are not necessarily able to inform ourselves exhaustively, to master all the parameters of a problem well, so we try to rely on shortcuts of information, we also talk about heuristics, which helps us to make a decision without necessarily entering into a complex and sophisticated information processing and decision-making mechanism. Identifying oneself with a party can be used to shorten information rather than to go and look for all the information oneself, to compare the parties' programmes, their ins and outs. This is the idea of an information shortcut that is also applied in other contexts.

In this model, this partisan identification is the key variable, there are others that are also integrated into the model, but play a much more secondary role. The Michigan model also refers to other types of political attitudes such as opinions on current political issues, but also sympathy for candidates. As much as party loyalty is a stable factor in the long term, so much so, it is well understood that attitudes about issues and sympathy for candidates are short-term factors that can also fluctuate during an election campaign itself. So, the basic idea of this model is that, in general, preferences are stable because identifying with a party is a stable factor, but it is recognized that sometimes, as an exception to the rule, there may be fluctuations in the electorate's own preferences because of preferences on issues or preferences on party candidates that may change over time in an election campaign. It is, therefore, the exception, the rule is stability because of stable party loyalty.

School of rational choice

Article détaillé : The rational actor.

The School of Rational Choice is linked to the University of Rochester because that is where Anthony Downs studied and taught. Downs is the author of reference for all the literature on rational choice, he is a little bit the founding father of the School of Rational Choice. His book An Economic Theory of Democracy published in 1957 remains a reference. This school has not only developed in the field of political behaviour but also in other areas of politics.

We're changing our perspective. The Columbia School and Michigan School models assumed that there is a strong link between a voter's profile and his or her vote, whether it is the socio-demographic profile with the Columbia School or the psychosociological profile with the Michigan School. The two schools agree on the idea that if we know this profile, we know about who the person is voting for. In this case, the profile explains the vote.

For the School of Rational Choice, the analytical cursor is shifted a little and the emphasis is on individual decision-making mechanisms. This model is less deterministic than the others, we cannot know in advance how an individual will vote, we must look at the mechanisms that lead to decision-making to understand this person has voted. The mechanisms on which the School of Rational Choice emphasizes are cost-benefit calculations, a so-called "utilitarian" approach to voting. It is believed that individuals decide on the basis of a cost-benefit calculation.

What are the benefits, what are the costs associated with a voting decision? What is the benefit of voting for this party or what is the loss of voting for this party? The voting determinants in this model are based on a utility calculation. This is the same logic of homoeconomicus applied to homopoliticus. It is assumed here that the homopoliticus behaves like a rational being who will try to make cost-benefit calculations and vote according to these cost-benefit calculations. It will, therefore, seek to maximize its usefulness.

There are three strong assumptions in this model: voters are aware of their preferences and make an effort to gather the necessary information to be able to make cost-benefit calculations. So they will seek information in order to make a rational choice.

  • voters are able to identify exactly the costs and benefits associated with a voting decision and then are able to vote rationally and therefore choose the party that effectively maximizes their utility.
  • voters are not influenced by their environment. Voters are at the heart of their own decision, they will seek information, compare and weigh costs and benefits and make their choice. They are not influenced by party propaganda, they are not influenced by the context in which they live, they are not influenced by their families, etc.

Deficiencies of traditional models

These three models have many weaknesses and defects. There has been a huge literature to criticize, amend and correct them. If we talk about political behaviour, we have to start from these three models because they are the basis from which we can start thinking a little more seriously and with slightly more recent models.

What are the shortcomings of these classic models? There are several of them and we will focus on the main ones.

Empirically, the studies that were done in the 1970s, 1980s and 1990s did not really confirm this strong weight of sociological and psychosociological factors. The theses of the Columbia School and the Michigan School that one could really explain the vote if one knew the social characteristics and partisan preferences of individuals, studies have not confirmed this. The explanatory power of these models is low. It is possible to explain something, but not much.

Why these models were not so efficient and, why they tended to lose performance over the years and decades, is because these explanatory factors at the heart of the models have declined over time. There has been a historical decline in the factors that explain the vote, such as social class or religion, or the fact of identifying with a party as postulated in the Michigan model.

Why was there this decline? This is because there have been changes in society that have led to this decline, such as changes in social structure. Society has changed significantly from a primary society in which the primary sector was highly developed to a society in which the secondary sector and especially the tertiary sector has been highly developed and this change in the social fabric has had major consequences from a political point of view. The tertiarization of the economy had major consequences on voting. The primary sector has shrunk, the secondary sector has shrunk and the historical links between the primary class or the working class, and a or other party generally on the left, these links have been greatly weakened. The same applies to geographical mobility, there have been great geographical mobility, which has led to much greater social mixes, much greater cultural mixes, which has also weakened traditional links between groups and parties. Overall, there has been a decline in class and religious loyalties and a decline in partisan identification.

The second factor that has contributed to the decline in the heavy explanatory factors is the development of education. It is what is sometimes called revocation, which is the fact that the level of education is greatly increased in Western societies, leading to an increase in the independence of minds, in the decision-making autonomy of voters and has made them less captive, less prisoner of their traditional allegiances. As education levels increase, people can afford to form opinions more independently and autonomously, they are less influenced by organizations, groups or parties. There is an electorate that is more independent, but we also have an electorate that is more volatile. Where previously, when the major models of explanation of vote explained voting with fairly stable behaviour from one election to another with citizens who are more independent, more autonomous, more critical, we have greater volatility, greater instability in electoral behaviour at the individual level. It is easier than before to change parties from one election to another.

The third essential factor is the rise of audiovisual media, first with television, but now also with electronic media, this rise of audiovisual and electronic media has also radically changed the situation with regard to election campaigns and voting campaigns. Once again, the result is individuals who are less captive, less influenced by organizations such as political parties and more influenced by the media and everything that is done in the media, by media coverage or advertising and who influence electoral behaviour much more than before. Overall, there is less party influence in political communication and more media and campaign influence with short-term effects. Where traditional models, particularly the Columbia and Michigan models, emphasized the stability and importance of long-term explanatory factors such as social inclusion or party identification, we now know that short-term factors are much more important than before. This does not mean that there is no longer any importance of long-term factors, but in the short term, clearly, there is a relative increase in importance.

Another shortcoming of these classic models of explanation of vote is that they were all more or less explicitly based on a simplistic conception of the electorate. Simplistic because it was homogeneous, i.e. these models took into account individual differences, but above all socio-demographic and possibly psychosociological differences, but did not take into account the fact that individuals differ from each other in their relationship to politics. Individuals and citizens differ from each other in their relationship to politics and in particular in their interest in politics and political competence. Not all citizens are equally interested in politics, some are very interested and even a lot of people who are involved in politics are making a career out of it, while others are not interested at all in politics. On the other hand, some citizens have a very good political knowledge, understand the issues and master, inform themselves, while others do not have the cognitive skills and motivations necessary to inform themselves and therefore do not have the knowledge necessary for informed participation and voting. So, interest in politics determines the degree of attention to politics and it also determines political participation, namely if you are interested, you are very likely to participate, if you are not interested you are very likely to abstain. Motivation and interest in politics also determine the attention given to politics and the political message. Political competence, on the other hand, determines the ability to integrate messages delivered in the public space. There can be a wonderful information campaign with positions on the right and positions on the left, rich debates and if at the individual level, people do not have the necessary skills to understand, internalize and assimilate these communications, it will not influence their opinion and contribute to the formation of their opinion. Somewhat more competent people will take this into account and weigh the pros and cons trying to get an idea based on the information provided in the public space.

The point here is that interest in politics and competence in politics, in other words, a motivational factor and a cognitive factor, namely an interest factor and a competence factor, both of which will condition and play an important role in the process of forming people's opinions. We are now trying to take into account the heterogeneity of the electorate, we are no longer betting on a homogeneous electorate, we are increasingly trying to take into account the diversity and heterogeneity of the electorate.

The last gap in classical models and especially for the school of rational choice, there is a huge focus on individuals. The school of rational choice is the paradigmatic case of a focus on the individual since for the school of rational choice, the individual makes his cost-benefit calculation independently of the context and independently of any form of external influence, it is he who is at the centre, to be informed, or even which party pays him the most, which one costs him the most and according to that make his choice as, for example, know which party is closer to him on a left-right scale, and we will vote for the party that is close to us according to our interests, but regardless of the context. The criticism made here is an excessive focus on voters and their characteristics and an insufficient consideration of the context in which individuals form their opinions.

This criticism applies mainly to the school of rational choice, but it also applies to Columbia School and Michigan School. The Columbia School provides that an individual votes according to the characteristics of the group to which he belongs, but even for this school, the group is not taken into account, it is only taken into account through the individual characteristics of the voter, namely whether he is a worker, whether he is Catholic or not. The inclusion of this voter is not taken into account or, for example, the role of trade unions in articulating workers. Even this model, which is a sociological model that places the individual at the heart of the group, even this Columbia model did not seriously take into account the role of the group. What is taken into account is the social characteristics of the individual and not the group in reality. However, individual opinions are not formed in a political vacuum, but in a specific institutional and political context. This specific political institutional context is likely to influence the way in which individuals form their opinions.

There are two elements of the context that should be mentioned here that are:

  • political offer: we speak of a political offer to designate partisan competition. It is therefore a characteristic of the parties that run for office, of the differences between parties, of the characteristics of the parties, of the characteristics of the candidates; it is called the political offer. Political demand would be the characteristics of individuals, the characteristics of voters are the characteristics of demand. The individual asks by voting in response to an offer made to him by political parties by presenting lists and candidates. The idea is that supply matters as much as demand. As traditional voting schools suggest, we must not focus on demand, we must also take supply into account because supply will have an influence on demand, on the meeting between demand and supply.
  • election campaign: there is a growing conviction, and studies show it, that the short-term factors that are conveyed in an election campaign influence voting, partisan choice and the choice of candidates. We are far from the traditional schools that assumed stable ties with individuals who vote for the same party from one election to the next. On the contrary, recent studies show that traditional allegiances, traditional loyalties to parties are declining and weakening, while short-term factors such as the role of election campaigns are becoming more and more important.

Electoral research: recent developments

In electoral research, we have tried to correct the shortcomings, to remedy the shortcomings of traditional models without necessarily abandoning them completely, we continue to take into account the role of social class, religion or other individual factors on voting, but we try to add other explanations.

Taking into account the context

First, we take into account the context that will become more systematic and rigorously taken into account. First, there is the institutional context, such as the electoral system, which influences the voting behaviour of voters, but also the upstream behaviour of parties when they present lists. Recent literature in electoral research attempts to take into account the role of the electoral system, such as the degree of proportionality of the system. We know that if the system is more or less proportional, there will be upstream consequences on how individuals develop their voting strategies.

The degree of polarization of the party system is to what extent we have parties that make proposals that differ from each other and the more polarized a system will be, the more different the proposals they will make because the more polarized the system is, the more the parties oppose each other on certain issues and therefore its likely to propose different solutions to the problems that arise. So, the more polarized a system is, the more varied and richer the offer will be for voters. On the other hand, the more consensual a system is, the less difference there will be between parties, the more difficult it will be for voters to differentiate between parties. Similarly, in addition to polarization, the degree of fragmentation of the party system is also taken into account, i.e. how many parties are presented. The election campaign and the media are also taken into account, trying to see how political communication and media coverage influence opinion formation.

Taking into account the heterogeneity of the electorate

The second innovation is that we are trying to take into account the heterogeneity of the electorate. We assume that the electorate is no longer homogeneous as with traditional schools, but that it is on the contrary a heterogeneous electorate that differentiates itself, that is diverse and varied with interested persons and others not, competent persons and others not, and we try to model and see how these differences of interest in politics or these differences of political competence affect the opinion-forming processes and the electoral choice processes.

It is a trend that is part of the political psychology that is booming in the United States and Europe. Political psychology places more emphasis on the psychological mechanisms behind opinion formation and choice.

Now, we assume that the stake vote is much more important than before. The issue vote is not the main ideologies, but what are the important issues of the day, what are the preferences of voters on these issues and which parties are perceived to be the most competent to solve these problems from the voters' point of view. This is now a powerful factor in partisan choice. With Switzerland, we can very well show that the SVP vote is partly a vote of stake because of the immigration issue that is imposed on Switzerland and because of the SVP's reputation, reputation for competence or at least reputation for dealing with this immigration issue. Many voters certainly gave their votes to the SVP because it was the party that had the best solutions for the most important issue of the day in the eyes of voters. These are short-term mechanisms that tend to become important in electoral behaviour.

Methodological innovations

The latest innovation is that the complexity of the models has been made possible by methodological innovations and in particular the use of hierarchical models or multi-level explanatory models. In statistics, models are developed that take into account both the characteristics of individuals and the context in which they vote. In other words, individual and contextual factors are simultaneously taken into account in the explanation of the vote and the interactions between individual and contextual factors are also taken into account. It is this complexity with the simultaneous consideration of contextual and individual, but more jointly, factors, i.e. the consideration of interactions between contextual and individual factors and this is the best way to explain an electoral choice. For example, we try to show that being Catholic does not have the same impact if we live in a Catholic canton or a religiously mixed canton. So there is an interaction where being Catholic will be a more or less powerful factor in the vote depending on the context, i.e. between Catholic cantons and religiously mixed cantons. The idea is therefore to take into account both individual and contextual factors.

Examples

Example 1: Explanation of the USC vote

This study analyses the composition of the UDC electorate and the evolution of this composition over time.

Source: Oesch et Rennwald 2010

The graph on the left shows the composition of the Socialist Party's electorate in 2007, results from an opinion survey after the 2007 federal elections. We mentioned the SELECT studies that have been conducted since 1995 after each federal election, where a Swiss opinion survey is conducted to assess individual behaviour.

In 2007, the Socialist Party had achieved about 20%, which also represents the average score of the Socialist Party. For different socio-professional categories, there is also a difference less or more than this average, which gives an idea of which segments of the population are most likely to vote for the Socialist Party and which segments of the population are least likely to vote for the Socialist Party.

Starting with the last line, we see that there is a socio-professional category that votes massively for the PS, which is what we call socio-cultural specialists; while the PS has averaged 20%, it has done 34%, or more than fourteen percentage points among socio-cultural specialists. The socio-cultural specialists who are sometimes called the new middle class are the employees who are active in the field of health, social, education, culture, but also in the media, thus a slightly superior middle class, but not too many, which has grown in number, what could be called, in a rather trivial way, the "bobo", namely the "bohemian bourgeois". These are people who, on the one hand, are relatively wealthy in terms of resources, but who, on the other hand, are close in terms of values to the redistributive values of the left. These are people who, if they would vote selfishly as postulated by the rational choice model, would actually go from side to side because their socio-economic situation would in principle lead them to support Liberal programs. However, these people easily vote on the left because they are in solidarity with society in general and they are also attracted by the other programmatic values of the left, such as international openness and solidarity.

We see that all other categories are below the average score of the socialists, including those called here "production workers", "service workers" and office workers". These are the people who were called workers at the time. Production workers are people who are active in the typically industrial field with repetitive tasks that have little autonomy in their work. These people vote less socialist than the average.

With regard to the SVP, for the 2007 elections, the SVP made 28% in 2007 with wide variations from one socio-professional category to another. For the SVP in 2007, one could say that the Columbia model still says things: the integration of individuals into society, their social class continues to say things about voting.

Who votes UDC? The little freelancers. Where the SVP made 28%, it made 44% among the small self-employed. Almost one in two small independents voted SVP. They are farmers, traders, craftsmen or self-employed people who do not run a large company. This is known in the jargon as the "old middle class". It is one of the two bastions of the SVP. The second stronghold of the SVP is the production and service workers. Where the SVP made 28%, it made about 40% among workers. This is something a little more surprising given that the SVP is a right-wing party from an economic point of view and therefore not a party that defends the interests of workers. What makes a higher than average proportion of workers vote for the SVP? Perhaps it is not because the SVP defends workers better than the Left, because there it would be possible to discuss, because trade unionists would say that the SVP does nothing to defend workers' interests, it is not the SVP that protects workers against foreign labour, or indirectly by closing borders so that there is no more competition. If the SVP is so successful with workers, it is not for the economic values or the economic programme that the SVP defends, but for the cultural programme that the SVP defends, i.e. for the desire to close borders in cultural terms, namely the defence of traditions by international closure not so much driven by economic considerations, but by cultural and identity but also historical considerations. In the two-dimensional space, we find here the result that if the SVP is progressing, it is largely due to its program on the openness - tradition axis and not for its program on the economic axis. From this point of view, it is not so different from the PLR, it is the famous win-lose divide. To put it simply, one could say that those who feel they are winners, in their perceptions, are sociocultural specialists, while the losers are those who fear international openness and who are afraid of this competition, which is not only economic, but also cultural and identity-based. These groups are the small self-employed on the one hand and services and production on the other. On the other hand, it is a bit of a mirror effect compared to the PS, the SVP also scores much worse among ethnic and socio-cultural specialists.

We are talking about class cleavage here, but we are saying that class cleavage still plays a role in electoral behaviour in Switzerland, but this role has changed. The class divide, traditionally, pitted workers against employers. This was the reformulation of Marx's theses on the conflict between labour and capital and for a very long time, in European states, there were marked differences in voting between workers on the one hand and employers on the left, with workers voting on the left and employers voting on the right. In Switzerland, as in other countries, there has been a reformulation of the class divide with a new class divide. There was a phenomenon of misalignment and realignment of voters with respect to parties. More concretely, voters in popular circles tended to slide from the left to the populist right. This has been observed in Switzerland, but also in other countries such as France or Austria, the Netherlands or the Scandinavian countries. In all these countries, we can show that there was this shift movement during the 1980s and 1990s by workers who voted for the left and who now tend to vote more towards the right population. Not all of them do, but a good number of them. On the left, however, there has been a strengthening of socio-cultural specialists as a bastion of the left. This is a powerful result that applies in Switzerland as in other countries.

Source: Oesch et Rennwald 2010

This graph shows the same thing. In 1995, between 15% and 20% of service, production and office workers voted for the SVP, and we can see how this has changed over the past ten years with the scores we have achieved with workers who vote from 35% to 40% for the SVP. Of course, the SVP has progressed everywhere over the past twenty years, but it has especially progressed among the popular electorate, hence this reformulation of the class divide.

The class divide still matters, but it has changed in nature. It has been reformulated due to changes, misalignments and realignments between social classes and parties. When we talk about misalignment, it is for example the fact that the workers have gradually distanced themselves from the Socialist Party or the left in general in order to vote SVP, which is therefore a realignment. With this process of moving partisan allegiances across social classes, the content of the class divide has changed. We are now talking about the "new class divide" that would oppose on the one hand the winners of globalization such as managers and the new middle class, and on the other hand the losers of globalization or at least those who see themselves as losers or who are afraid of being losers, namely the working classes and the former middle class of the small self-employed such as artisans, farmers or traders.

The table above illustrates the evolution of the middle classes for the SVP. Three segments of the working classes with the grey office employee segment, the dotted line service employee segment and the black production worker segment. In all three cases, there was a very sharp increase in the proportion of workers who voted SVP. Of course, the SVP has made progress in all segments of the population, but especially among the working classes.

This graph is a simple arrangement of voters on a two-dimensional space.

Source: Oesch et Rennwald 2010b: 276

On the horizontal axis, there is a dimension that could be called "for more government or for more market". This dimension is constructed with two questions asked in the opinion surveys with a question concerning social spending, namely whether the respondents are in favour of increasing or decreasing the Confederation's social spending, and the other question is whether the respondents are in favour of increasing the taxation of high incomes or on the contrary are in favour of reducing the taxation of high incomes. In both cases, depending on whether the answer is "yes" or "no", "favourable to spending", "unfavourable to spending", "favourable to raising taxes" or "unfavourable to raising taxes", this can be interpreted as left-wing or right-wing values on the economic axis of "more redistribution" with social spending and tax increases, and "less social redistribution" with reduced spending and taxes.

The vertical axis represents a first question that is "do you support" Switzerland's accession to the European Union or, on the contrary, "do you support a Switzerland that is going it alone? The second question is whether we are in favour of a Swiss who gives the same opportunities to Swiss or foreigners or, on the contrary, whether we are in favour of a Switzerland who favours Swiss over foreigners.

Then, the position of each voter on these two dimensions is calculated. In this graph, the average position of the sub-segments of the electorate of each party is calculated. For example, we can observe the average position of persons who belong to the class or class of socio-cultural specialists and who are PS voters compared to voters who are also socio-cultural specialists, but who on the other hand voted SVP. In the middle is the average position by socio-professional category. If we take the average position, they're all a little bit towards the center and there's not much difference. This is because among managers, for example, there are some who are very right, others less right, others up and others down, and when you take the average, it puts them quite centrally.

First of all, we can see that the SVP electorate is relatively homogeneous where all the sub-segments that make up the SVP electorate are quite close to each other, they are all relatively close to the "defence of traditions" pole, and a little to the right from an economic point of view, but in fact relatively centrist. In comparison, the SP electorate is much more fragmented. On the other hand, it is clear here that it is possible to be a production worker and have very different values. The production workers who voted SVP are very much in favour of closure and traditions and are rather right-wing economically. The workers who voted PS are neutral on the openness dimension, but they are more favourable on a redistributive policy.

The distance that there is between sociocultural socialists and workers socialist voters of production, on the vertical dimension, openness - closure, there is a world difference between socialist sociocultural voters and socialist workers voters. This is a bit of a problem for the SP at the moment because the SP, in Switzerland at least, is somewhat kept at a large distance between its popular electorate and the "new middle class" electorate of sociocultural specialists. Some, the workers, would like more authoritarian policies, less international openness, more security, fewer immigrants, fewer asylum seekers, while the socio-cultural ones are much more open, supportive and in favour of a generous policy towards a foreign policy. This is the great dilemma of the SP, because if the SP seeks to seduce its new electorate of the new middle class, it risks alienating its popular electorate. If the DP hardens its positions and becomes a little more firm on international issues, compared to others, immigrants and asylum seekers, then it will probably please its popular electorate and displease its sociocultural specialist electorate.

There is a real dilemma for the SP, which is a dilemma that the SVP does not have because its electorate is relatively homogeneous. It can be seen that the SVP electorate is mainly characterised by values of closure and the defence of Switzerland's traditions of sovereignty. The "business body" of the SVP is not so much its economic profile, it is independence, sovereignty, neutrality, the tightening of migration and asylum policies. All his electorate likes it, whether they are workers or not.

As we can see, on the horizontal dimension, the SP does not have too many concerns because the whole of its electorate is relatively homogeneous on this dimension, they are all aligned on -1 and -1.5 being all grouped almost vertically, which means that on redistribution issues, the PS electorate is homogeneous. Workers because they are in favour of a redistributive policy that serves their interests and socio-cultural specialists because they are ready to make an effort of solidarity with the less favoured classes.

To summarize this example, first of all, the class divide has changed in meaning, but it remains important as an explanatory factor of electoral behaviour. It is no longer the same class divide, but if we take into account these changes within the divide, these realignments, the class divide continues to explain electoral behaviour quite strongly. On the other hand, socio-cultural specialists, namely the so-called wage middle class or the new middle class of employees in the social, health, education, culture and media sectors, these socio-cultural specialists have become the new bastion of the left. In addition, the working classes and the former middle class, the small self-employed such as craftsmen, traders and farmers, have become the bastion of the SVP, which is the strongest voting class for the SVP. There is therefore a widening gap between the winners or supposed winners and the losers or supposed losers of globalization. From a normative point of view, this cleavage is expressed by the conflict on the vertical axis between open and tradition, which others call integration - demarcation, sometimes called libertarian - authoritarian.

Example 2: Explaining the success of the SVP

There is an alternative explanation to the SVP vote, which no longer refers specifically to the class position and the vote that goes with the class position, but is based more on what is called the stake vote. The stake vote is the idea that electoral behaviour is increasingly influenced by the major political issues of the day, regardless of the class to which one belongs. We take individuals from all social categories and try to see what their preferences are on the issues, what issues voters consider important, what their preferences and positions are on these issues and can we, on this basis, try to define for which party an electorate voted. Depending on voters' preferences on the issues, the question is whether it is possible to predict or understand why voters vote for a particular party.

By "stake", we mean the major political problems of the moment such as the migration crisis, ecology, nuclear energy or unemployment. These are concrete issues that arise every day and that citizens are aware of, leading them to vote for one party or another according to the proposals made by these parties. The idea here is to say that we must move away from these structural explanations of the "social class" type to move more towards short-term situational explanations being factors that are likely to influence and modify electoral behaviour in the short term. We can imagine that an election campaign succeeds in putting a new issue at the centre of the political agenda and that the party or parties that position themselves on this issue benefit from it in the election.

We are moving away from allegiances, lasting and stable loyalties between individuals and parties in order to better understand what is going on and what are the factors that would change partisan preferences and therefore the electoral choices of voters in the short term.

The stake vote

There are two main types of explanations related to the issues.

The first explanation is directly derived from a rational choice model with the voter voting rationally making a cost-benefit calculation. The idea is that voters will vote for the party that is closest to them on the issue. The parties that have the most similar preferences with voters are those for which voters will vote. If, for example, we are in favour of reducing immigration, we will vote for a party that supports immigration restrictions. It is possible to calculate distance models that ask respondents to position themselves on scales and parties to position themselves on the same scales, and then it is possible to calculate the probability that an elector will vote for which party based on their distance.

The general idea is that of the proximity model with a whole series of large types. With the proximity model, we try to explain electoral behaviour according to the distance or proximity existing between voters and parties on important issues.

The second idea that is very close is that the elector will vote for the party that has the most important issue, that is, the party that is deemed most active and competent on the issue in question. There is an issue vote related to a specific issue. It is not the position of voters and parties on different issues, it is what is the main issue currently at stake in the country, what is the party that has this issue and has developed over the years a reputation as a party that deals with the issue, that is able to manage and find solutions to this issue; if the issue becomes prominent among the population, this party will benefit from it in the election. This is the trend we call "issue ownership", on the party side, we try to develop reputations of skills on issues. For the Greens, it will be to develop their reputation for competence in environmental issues, for the Socialists, it will be their competence in social policy and redistribution, for the PLR, it will be to boost competence in economic policy, for the SVP, it will be to develop competence in immigration, security and European policy. This is quite stable because it is not easy for a party to change its reputation for competence, it takes months and years. To some extent, this issue is becoming important, so the party that owns it can win a lot of votes. For example, the victory of the SVP on 18 October 2015 is considered to be very much linked to the migratory and humanitarian crisis linked to the refugee movement, particularly in Syria, and the SVP, without doing much, if anything, about it has been in the news. The current situation has made this the issue that was raised in the media throughout the election campaign without the SVP having to campaign. Since the SVP is known as the party competent to have simple, even simplistic solutions to the issue, voters rely on it and then vote.

Source: Nicolet and Sciarini (2010: 451)

This diagram is taken from a post-election survey conducted after the 2007 federal elections. This is done quite systematically in these surveys, first of all a first question is asked, which is an open-ended question that consists in telling the respondent that there are different problems facing Switzerland at the moment and therefore we ask the respondent what is the most important problem today. It is an open-ended question and then the answers are grouped according to different categories. Once we have asked this first question, we ask a second follow-up question, which is "in your opinion, which party is the most competent to solve problem X? ». Separately, elsewhere in the questionnaire, this person was asked which party he/she voted for in the elections. These are three pieces of information that can then be put together.

This diagram is the entirety of the respondents who were interviewed in 1716 people. There were 1716 people who participated in the elections and chose a party. On the first line, there is the distribution for the first question mentioned which is an open question. For 35% of respondents, the most important issue was immigration, security and integration of refugees, for 16% it was the environment and for 31%, a broad category, the concern was related to the economy and the social state. If we add it up, we do not reach 100% for the simple reason that there are still other issues that are not taken into account here.

The second line aims to answer the question of which party is the most competent to solve this problem. These are always net percentages, i.e. the people who responded. Of the 35% who answered "immigration", 27% answered "immigration", a good part answered either SVP or PS. That is, 75% of those who cited "immigration" as a problem believe that the SVP is the most competent. Then, on the last line, we looked at what these people voted for. 17% of the 1716 people said they voted SVP, being the most competent party on immigration which is the main problem for them. This does not mean that 17% of voters voted SVP because immigration is the most important issue and the party is the most competent. We do not go so far as to establish a causal and correlative relationship. In any case, it is an indication of the importance of the "immigration" issue and the SVP's competence on this issue for the SVP vote. This does not mean that all these people voted SVP for that reason, but there is probably something going in that direction.

The rise of the SVP has been the spectacular factor in Swiss politics for twenty years now, and so there has been a lot of work done on this topic.

Electoral potential and exploitation of the potential

To explain the SVP vote, we began by saying that there is an explanation related to the socio-professional position, which is the class position that can already partly explain why we vote SVP rather than another party. We have seen that there is the question of the stakes, particularly with immigration, which plays a structural role in Swiss politics and which makes it easier for the SVP to vote. A third type of explanation refers to the party's strategies and the effects of these strategies in terms of mobilization.

The explanation we will put forward is that the SVP owes its success to its formidable capacity to mobilize its electorate. We are not going to study here how the SVP mobilizes. We will already show that we can see the effects of these mobilization strategies, they are tangible in the surveys. We do not analyse SVP strategies or political communication, we analyse the result of this communication and these messages. To do this, the standard and most obvious question when measuring electoral choices in a survey is "for which party did you vote". This is the key question, because without this information, there is almost nothing we can do. An attempt has been made to develop other measures and indicators that also provide information on party preferences without being limited to electoral choice.

The problem with electoral choice is that once a person has said they have voted SVP, we have no knowledge of that person's preferences over other parties. The person may have voted SVP, but he or she could also have voted PLR. Or a person voted for the Vers, but he could very well have voted for the Socialist Party, and once the person said he voted for the Greens, all the information is lost for the rest.

What we are doing is trying to implement measures that ask the question about all parties. This measure is called the voting probability measure. In the survey, we propose a scale from 0 to 10 and what are the chances that one day we will vote for a particular party. The same question is proposed for the main political parties in order to have a comparative view because we have information not only for the party we have chosen, but also for the others we have not chosen. This allows comparisons between parties to be made much more accurately than the raw question of "electoral choice".

Once all persons in the survey have been asked how likely they are to vote one day for the main parties in the canton, for example, it is possible to calculate on this basis the average probability of voting for a party. It is not complicated, it consists of summing and averaging the probabilities. The scores of each respondent are summed and divided by the number of respondents. This average probability of voting for a party with electoral potential. It is a measure of the electoral potential of each party. It is possible to do this for each party separately.

Afterwards, on this basis, it is possible to calculate the concretisation rate or the exploitation rate of the potential. A simple ratio is calculated, i.e. a division between the effective electoral strength of the party, i.e. what percentage of votes the party obtained, divided by its potential derived not from the votes cast, but from the survey, which is the average probability of voting for that party. This obtained rate is a measure of the parties' ability to convert potential into effective support.

Electoral potential of the parties (average probability of voting)

Let us start with the electoral potential measured in the surveys, in other words the average probability of voting for one or the other party.

Source: Données Selects (mes calculs (M. Sciarini), N=4064-4261)

This graph shows, for surveys conducted after the federal elections of 1995, 1999, 2003, 2007 and 2011, the average probability of voting for each party. In other words, it shows the electoral potential of each party. For all parties, the potential is much higher than their actual electoral strength.

If we take the case of the Greens, they have a potential of 44%, in other words, on average over the entire sample, the probability that someone will vote Green is 4.4/10. Translated into percentage terms, this is 44%. However, at the end of 2015, the Greens are currently at 7%, 8%. This is the extreme example of the difference between potential and effective electoral strength. It should be noted in passing that there is such a difference between the potential and the Green vote for two reasons. The first is that there is the whole electorate responding in this graph and including people who did not vote, and among the people who did not vote, there are many young people who like the Greens. The popularity of the Greens among young people increases the potential of the Greens, but does not translate into votes because young people do not vote or vote little. The second explanation is the competition between the Greens and the Socialist Party. The Greens and the PS share most of the same potential electorate, but in the end we vote much more often than the Greens.

The first point is that the potential vote is much higher than the actual vote. That said, there is a very strong relationship between the two, there is a correlation of 0.8 or even 0.9 between potential and voting at the individual level, so it is very closely linked. The second point is that this potential fluctuates slightly from one survey to another, but not massively, there are some fluctuations, but overall it is relatively stable. There has been a loss of potential for the socialists, but they have recovered a little in 2011. The two left-wing parties have, according to these measures, the highest potential.

The main point that can be derived from this graph is the following and concerns the UDC curve. As can be seen, this potential is stable and quite low. It has never exceeded 40%. The electoral potential of the SVP is relatively stable and relatively low, it is the lowest of all the parties considered here including new parties such as the BBD and the Liberal Greens. What can already be said here is that the success of the SVP is not due to an increase in its potential, which remained constant or even decreased in 2011 compared to 2007. The important thing to remember is that the UDC potential does not increase and is relatively low. This is still spectacular if we link it to the SVP's electoral curve, which has an impressive positive slope.

Materialization/operation rate

This graph illustrates the rate at which the potential is realized. In other words, it is the ratio between the electoral strength and the party's potential.

*mes calculs = M. Sciarini

What we see here is the extremely sharp increase in the rate of achievement of the SVP. In 1995, 1999, 2003, 2003 and even 2011, the SVP was able to increase its implementation rate almost systematically. In other words, the SVP has been able to almost systematically increase its ability to mobilize its potential electorate. This is the main reason for the success of the SVP. It is not an increase in its popularity among the electorate, the SVP remains about as popular as it was twenty years ago, that is, not very popular, but on the other hand, people who think they are voting SVP actually vote much more strongly than the other parties. The concretization rate of the other parties is barely over 40% and even less than 20% for the Towards, which is in sharp contrast to what we see for the SVP.

Thus, the rise of the SVP over the past twenty years is essentially due to its growing capacity to mobilize its electorate, which has remained relatively stable in terms of potential. It is the idea that the SVP is the type of party that provokes very strong reactions, either you are "friend" or "enemy". If you are a "friend", you vote for the party, if you are an "enemy" you will never vote for the party. The SVP's ability is to have convinced more and more friends to vote. He has no more friends than before, but his friends vote more frequently for him or more and more for him.

Comparative opening

The rise of populist parties is a phenomenon that applies in many European countries.

Sciarini ouverture comparative élections au Parlement européen de 2014.png

This table seeks to show the analogies between party families. As we can see, if we take figures from the European Parliament in 2014, there are several parties, including in countries other than Switzerland, that have exceeded 20%. The FN made 25%, the Austrian Freedom Party made 20%, the UKIP in the United Kingdom made 28%, the Cinque Stelle in Italy made 21%, the People's Party in Denmark made 27%, the Freedom Party in the Netherlands made 13% and even Sweden, which has long been spared by populist parties, has seen a party reach almost 10% which is the Swedish Democratic Party. This phenomenon described by the SVP also applies to other countries and in most European countries.

Sciarini 2015 ouverture comparative élections parlementaires nationale.png

The results of national parliamentary elections make us aware of the differences that can occur from one election to another. If we take the case of the FN, which is the most obvious, it made 25% in the European elections, but two years earlier, the FN had only made 9% in the legislative elections in France.

We have to be careful with these figures because the European elections are often intermediate elections, which are called second-rate elections, which means that they are elections where voters take advantage of the opportunity to sanction the government in power, to give a signal to the government. They do not have much importance, so we can afford to vote against them. This may explain why the parties of disgruntled people are so high. They are more generally in many countries than in parliamentary elections, but even there, there are parties such as the FPE in Austria with more than 20%, the Cinque Stelle with more than 25%, the Popular Party in Denmark with more than 21%. It is still a real phenomenon, but at the same time we take into account the differences that can exist between two elections depending on their meaning and scope.

Example 3: Gender, age and participation

The study of political participation is another field of electoral behaviour. Sequentially, participation comes before electoral choice, you must first understand why voters go to the polls and participate before trying to understand what they are voting. Logically, we first try to understand who participates and who does not participate and why before we understand who votes for which party and for what.

Comparative analysis

Sciarini 2015 taux de participation aux élections et votations fédérales 1.png

This first graph shows at the aggregate level the evolution of the participation rate in Switzerland in federal elections and federal referenda since 1919, i.e. since the end of the First World War. In both cases, for the red elections and the black votes, there was a sharp decrease in voter turnout. Elections were 80% for the elections at the end of the First World War and there was an incessant decline to a low point in 1995 with less than 45% turnout. There was a lower and more fluctuating turnout for votes, but there was a parallel trend between the 1940s and 1970s to reach a low point for votes in the late 1970s with an average of 40% turnout for votes.

For elections, the figure is the participation rate for the current year's elections. For the votes, it is the average participation rate on all the votes that have taken place over four years, since people vote four times a year in Switzerland in federal votes. If you want to have participation for four years, you have to calculate the average participation for four years and then you can compare these two curves.

Thus, there is a secular decline in participation with a low point reached in the 1990s for elections and then a slight increase in participation since 1995 with stability over the last three elections. On October 18, 2015, there was approximately 43.8% participation, 44% in 2011 and approximately the same in 2007. Now, in the short term of the three elections, we are a little below 48% turnout stability for the elections and we are about the same at the same level of 43% for the votes in the last three surveys.

The general image that emerges is a sharp decline in participation and the question that must be asked is where does this sharp decline in participation come from?

We will focus on two factors that can help us understand political participation, which are two factors that explain participation and abstention, two quite fundamental factors that are gender and age.

Sciarini 2015 taux de participation aux élections fédérales de 1995 à Genève.png

These are figures on the evolution of participation in federal elections in the canton of Geneva. The figures in this graph are only from the canton of Geneva. The advantage of these figures is that they are real figures. It is not participation as measured in opinion surveys, it is real participation. Since 1995, the Canton of Geneva has had the good idea of collecting and archiving data on the participation of all Geneva citizens in electronic format, which makes it possible to monitor the evolution of participation over time.

This is a revelation for the 1995 federal election. It is a participation curve that is almost a school curve, that's how we imagined it and that's how participation is really within the population.

This curve is a school curve. At the age of 18, when young voters obtain their right to vote, there is a higher turnout than for those aged 20 to 25, because we have received a new right and therefore we must use it and participate a little more than the next category. There is a first movement that is a "U" movement. The low point of participation is between 20 - 29 years old and after that, it rises almost linearly with age up to 65 - 69 years old and then it drops quite sharply in old age.

Sciarini 2015 taux de participation aux élections fédérales de 2015 à Genève.png

If we use the same figures for 2015, we find the same curve with the same inflection after a fall that is a little less abrupt, we go down less low than in the previous graph where we went down below the level of participation of young people, but this is mainly due to the fact that there are age groups up to 90 years and over, while in the next graph we have grouped all people aged 85 and over. This has the effect of increasing the participation rate on average.

The other thing that is quite interesting between these two graphs is that before, the high point of participation, the peak was reached between 65 and 75 years for men; in 2015, the high point of participation concerns 75 and 79 years for men and 70 and 74 years for women. So, apparently, there is a movement here that is causing us to vote more and more later, which would be logical enough with the increase in life expectancy because we are healthier and in better shape to be even more integrated and vote later than before.

If we look at the curve for women and men, in both cases in 1995 and 2015, young women participate more than young men. Then, there is an identical point for 20 - 24 year olds in both cases and after, gradually, the difference in participation between men and women tends to increase with age and is massive among older women and men. If we take the 85-89 age group, there is about 40% participation for women and more than 30% participation on average for men. This gap is found for those 85 and over, 40% participation among women and more than 55% participation among men. There is a strong age effect, but this age effect is coupled with a gender effect in interaction with age. There is no difference between men and women according to these graphs for young people, but as time passes, as age advances, there is an increasingly different proportion of women and men participating.

We will now try to explain why there is this difference in participation according to age on the one hand and gender on the other.

Gender factor

First, there are sociostructural factors that have historically explained the differential in participation between men and women.

The first sociostructural factor is the lower social and professional integration of women. The fact that women were much less integrated into the world of work, she worked less frequently than men, remained much more numerous to care for children and therefore remained as housewives, this made women less socially integrated, less professionally integrated and had a direct impact on political commitment, on political participation. This is in contrast to the classic model which was that the fact that women were less socially, professionally and professionally integrated than men leads to their lower political participation because social and professional integration is a powerful factor of social and political commitment in general.

The second sociostructural factor is the over-representation of women in marital states that promotes social isolation. If we look at the statistics, we see that there are many more older women than older men. Women have a longer life expectancy than men even today, even if it tends to converge and, above all, there are many more widowed women than widowed men. In 2015, in the category of widowers, there are 80% women and 20% men. In the general population, there are 51% women and 49% men, while in widowhood, there are 80% women and 20% men. This partly explains this dropout because widowhood is a powerful factor in social isolation. If you are a widower, you tend to be more isolated, you no longer have a spouse, you have children who have left home and this social isolation contributes to political abstentionism. Since there are more widowed women than widowed men, this contributes to a lower participation rate among older women on average, it being understood that there is a greater chance of being widowed in old age than before. It is possible to correct as if there were as many male widowers as female widows, the curves are narrowing dramatically.

Another explanation of sociocultural factors and more precisely the persistence of traditional models of the role of women. This is almost independently of the socio-structural factors, which is the fact that for a very long time there has been this maintenance of the traditional model of women's vision in society and of women's role in society in private and public spaces, which has had the effect of reducing the participation rate of women compared to men.

The last factor is specific to Switzerland and is an institutional factor. Women were given the right to vote very late in Switzerland. Women were granted the right to vote in Switzerland at the federal level in 1971, earlier in some cantons, but only in 1971 at the federal level. In some conservative cantons in central and eastern Switzerland, it was not until later than 1971 that women were granted the right to vote at the cantonal level. The last canton to grant the right to vote was the canton of Appenzell Ausserrhoden, which granted the right to vote to women in 1991 following a decision of the Federal Court. It was the Federal Court that imposed the right to vote on women in Appenzell Ausserrhoden because this discrimination was not in line with equal rights between men and women in the Federal Constitution. There is even a canton in 1991 that had still denied women the right to vote until then.

What are the consequences of this late granting of the right to vote to women and how has this affected the participation rate even today?

The reason is that some of the oldest women have reached the age of majority without the right to vote and have had to wait until late in their lives to obtain it. They have therefore missed a political socialization that is the one that any citizen who has the right to vote from the age of 18 can obtain. They did not learn about politics because they were not allowed to vote. It is considered that this delay in learning political socialization is still producing effects among older women today. One can even imagine that this institutional effect is combined with the sociostructural effect of the high frequency of widowed women. It is conceivable that widowed women will withdraw all the more willingly from politics and participation as they have been trained to participate through their marital ties. They voted because the husband did it too and because she talked about politics at home and maybe even, the wife voted as the husband told her, it was something common at the time, and once the husband died, these women find themselves alone, they were not socialized in their youth in politics and she drops out all the easier. So there is a conjunction of the effect of widowhood and the granting of late votes to women, which can still explain why we have this decline, this early exit of women from participation that we are still discovering here and even in 2015, even though it is tightening a little bit.

That was really the classic image and we have to address the revisionist thesis, which goes against the background of these explanatory factors presented and says that we have to be careful because these explanatory factors were good at explaining how women participated, but it is no longer able to explain how women participate today.

The idea is that there would have been a catch-up or convergence effect. There would have been a catch-up effect in the sense that women would have started to participate as much as men and therefore the participation curves would be getting closer over time. The explanations given for this possible catching-up effect are, first and foremost, the increased social and professional integration of women. Women are no longer confined to the household and the household, women's labour force participation rate has risen sharply throughout the developed world and therefore with this increased professional and social integration, women also participate politically more than men. Women are more socially and professionally integrated than before, so they are also more politically integrated than they were before and they end up catching up with men.

According to this revisionist thesis, the gender gap in terms of political participation has disappeared. In fact, it has disappeared in many developed countries such as the United States, in the Scandinavian countries even around Switzerland in countries such as France or Germany, there is no longer any difference in participation between men and women. Not in Switzerland. In Switzerland there is still a small participation differential for elections. According to polls, there is no longer a participation differential for voting. In federal elections, women participate as much as men. However, in elections, there is always a small difference between men and women. In Switzerland, the catch-up and convergence effect, even if it has occurred and even if it is not yet complete, the catch-up effect is perhaps also due to the fact that this institutional factor is losing weight over the years because fewer and fewer women have reached adulthood without the right to vote. Women who are affected by the lack of voting rights in their adult years are disappearing and therefore the institutional effect will gradually weaken and disappear. There are still differences, in Geneva there is almost no difference in participation between men and women overall. There is a differential, but it is really very small.

The effect of age

It's a little more complicated, because the age effect is a multiform, protean effect, because behind the "age" variable, there are different types of mechanisms.

There are three main age effects:

  • Life cycle effect and biological aging effect: the age of the arteries and simply something other than the age of aging, this is what the different ages represent in a person's life course, what it means to be 20, 40, 60 years old in terms of life course. In terms of life course, being 20 years old is not the same as being 40 years old. Being 20 years old is probably still in school, probably not yet married, probably not yet having a child, while 40 years old is still more likely to be married, to have children, to have a profession. It is well known that age affects participation through social integration and political experience. The 20-year-old individual who is unlikely to be married, who will not have children and who will still be in school or who will start his or her professional career will be poorly integrated socially, that is, poorly integrated into society. This low degree of integration will result in a low degree of political participation. On the other hand, getting married, having children, starting to participate in associations, going to meetings with parents, for example, has the effect of forging social ties that will then benefit political participation. The more we progress at the professional level, in our social life, the more we integrate politically and participate politically. The second mechanism is political experience, which could be called political competence; electing, voting, presupposes knowledge and skills that you do not necessarily have at 18 or 20 years of age, because you have to be interested in politics, many young people are not interested in it, interest comes with learning, the further you go and the more you start to understand what it is about, what it is for, why it is important and the more you learn and the more you have an information stock that increases with age and therefore the more you are able to vote. With this dual mechanism of social integration and increased political interest and competence, age is a powerful factor of participation. In old age, we are probably much more competent at 70 than at 20 in terms of political knowledge, but we can have health problems and social isolation that mean that we will withdraw into the private sphere and withdraw from the public and therefore refrain from participating, hence the fall in old age.
  • Cohort effect or generational effect: what is important here is the date of birth, i.e. when was born. It is not the age you are at that age that matters, it is in what generation you were born. Generation effects, also known as "cohort effects", depend on the date of birth, the time of birth and the events that affect the same generation. All the people who lived in May 1968, for example, all these people were permanently affected in their political socialization. It is possible that this generation effect will continue as people age, but they are still marked by the event that may have affected them when they were 20 years old. This is one way, for example, to explain the decline in participation over time. The continuous decline in elections and voting over time is probably the generational effect in large part, that is, the decline in civic sense as generations renewed themselves. Civic sense was much more developed at the beginning of the 20th century than at the beginning of the 21st century, it was almost natural to vote at the time, it was a duty, this sense of civic duty was gradually lost from generation to generation. It is also possible that this decline is due to the aging of the population. There are more and more very old people, but as a result there are more and more people who have left the political participation circuit.
  • Period effect: this period effect affects everyone at the same time, regardless of whether you were born in 1920 or 1950 or 1980, if there is a particular event, it can have an effect for everyone at the same time. Probably there is an example at this time of a period effect. We reached a low point in 1995 in Switzerland and after that it increased a little bit, we had a politicization effect for the whole period from 1995 to today. There has been a greater politicization of political life in Switzerland that has affected everyone, that has affected everyone, we have all participated a little more, whatever the age. In connection with the polarization of Swiss politics, this has become more competitive and conflictual.

It must be understood that it is very difficult to study the age effect because these three mechanisms are intertwined and combined. For example, it is impossible on the basis of a single election or a single opinion survey to be able to unite the three effects, it is impossible because they mix. If we want to identify the specific effect of the ageing of the life cycle, the generation effect, the period effect, we must have long time series in order to be able to control these different effects over time if possible and it is only in this way that we can identify the specific effect of each of the mechanisms.

There is a whole series of studies that focus on participation with much more complex models and more complex variables. Nevertheless, we must be aware of the effects of important variables such as age and gender.

Annexes

  • Bartolini, Stefano & Mair, Peter (1990). Identity, Competition, and Electoral Availability. Cambridge: Cambridge University Press.
  • Bornschier, Simon (2007). Cleavage Politics and the Populist Right. The New Cultural Conflict in Western Europe. Philadelphia: Temple University Press.
  • Brunner, Matthias & Sciarini, Pascal (2002). L'opposition ouverture-traditions. In Hug, Simon & Sciarini, Pascal (éds.), Changements de valeurs et nouveaux clivages politiques en Suisse. Paris: L'Harmattan, pp. 29-93.
  • Campbell, A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). The American Voter. New York: John Wiley.
  • Downs, Anthony (1957). An economic theory of democracy. New York: Harper and Row.
  • Duverger, Maurice (1951). Les partis politiques. Paris: Seuil
  • Inglehart, Ronald (1977). The silent revolution: Changing values and political styles among western publics. Princeton: Princeton University Press.
  • Kerr, Henri (1987). The Swiss Party System: Steadfast and Changing. In Daalder, Hans (ed.), Party Systems in Denmark, Austria, Switzerland the Netherlands and Belgium. London: Frances Pinter.
  • Kriesi, Hanspeter, et al. (2008). West European politics in the Age of Globalization. Cambridge: Cambridge University Press.
  • Lazarsfeld, Paul F., Berelson, Bernard & Gaudet, Hazel (1944). The People's Choice. New York: Columbia University Press.
  • Nicolet, Sarah & Sciarini, Pascal (2010). Conclusion. In Nicolet, Sarah & Sciarini, Pascal (éds.), Le destin électoral de la gauche. Le vote socialiste et vert en Suisse. Genève: Georg, pp. 439-467.
  • Oesch, Daniel & Rennwald, Line (2010a). La disparition du vote ouvrier? Le vote de classe et les partis de gauche en Suisse. In Nicolet, Sarah & Sciarini, Pascal (éds.) Le destin électoral de la gauche. Le vote socialiste et vert en Suisse. Genève: Georg, pp. 219-256.
  • Oesch, Daniel, & Rennwald, Line (2010b). Un électorat divisé? Les préférences politiques des classes sociales et le vote de gauche en Suisse en 2007. In Nicolet, Sarah & Sciarini, Pascal (éds.), Le destin électoral de la gauche. Le vote socialiste et vert en Suisse. Genève: Georg, pp. 257-291.
  • Petrocik, John, R. (1996). Issue Ownership in Presidential Elections, with a 1980 Case Study. American Journal of Political Science 40(3): 825-850.
  • Sciarini, Pascal (2011). La politique suisse au fil du temps. Genève: Georg.
  • Sciarini, Pascal, Ballmer-Cao, Thanh-Huyen & Lachat, Romain (2001). Genre, âge et participation politique: les élections fédérales de 1995 dans le canton de Genève. Revue suisse de science politique 7(3): 83-98.

References