Introduction

This book is an introduction to political science. What is it that political scientists are actually studying? Or, to put it another way: What do we mean when we talk about politics? Politics is about conflict, struggle, decision-making, power and influence. But not every conflict and not every situation in which power is exercised is widely regarded as politics. A football coach who decides to leave a player on the bench because he has given him a bit of lip, is exerting power – and there is conflict here, too. However, few people would consider this a political issue. The same applies to a mother who quarrels with her adolescent daughter about going to a house party, a schoolteacher who gives a student detention, and so on. But if we were to limit our understanding of politics to official decisions that are taken by governments, in parliaments or on municipal councils, we would fail to recognise the political meaning of trade unions, lobbyists, protest groups, corporations and other more-or-less organised groups that influence collective decision-making.

In the first part of this chapter, we will look at two key concepts in political science: politics and power. Although these are key concepts within political science, there are no generally accepted definitions of either term. Political scientists have defined them in various ways, so they have sometimes drawn different conclusions, such as about the distribution of power. Without wanting to take a position in these debates, we consider it useful to offer insights into the ways these two concepts have been defined by political scientists. That helps clarify the concepts themselves as well as their various aspects. In the second part of this chapter, we will give a brief outline of the research conducted on power. How is political and economic power distributed in democratic societies, and what are the different perspectives on this?

What is politics?

Although most people have an image of what politics is, it is not easy to define just what we mean by it. If we look at different definitions of the term, three elements are frequently mentioned: decision-making, conflict and power. Each is discussed here, starting with decision-making.

Introduction

In the pillarized Dutch society, election results were once very predictable. In the 1950s, a gain of two or three seats was celebrated as a resounding victory. Many people voted their entire lives for the party that represented the pillar they belonged to (see also Chapter 5). The shares of seats in parliament were therefore highly stable. The three major Christian parties, the Catholic People's Party, the Christian Historical Union and the Anti-Revolutionary Party, which later merged into Christian Democratic Appeal (CDA), the labour Party (PvdA) and the liberal VVD (People's Party for Freedom and Democracy) received between them more than 80 percent of the votes. From the middle of the 1960s, this began to change. Electoral swings were increasing, and over the last two decades we have seen more and more fragmentation of in the Dutch party-political landscape. After the 2017 elections, at least four parties were needed to form a coalition that had a majority in the Lower House of Parliament.

The results of those elections highlighted major changes, in particular because of the losses sustained by the two ruling parties, the VVD (8 seats lost) and the PvdA (that lost 29 seats). This was not the first election in which there were important electoral shifts. In 1994, the CDA lost 20 seats and D66 won 12. In 2002, the List Pim Fortuyn (LPF), the new party of the assassinated politician Pim Fortuyn, entered the Lower House with 26 seats. The PvdA lost 22 seats; the VVD, 14. It is no longer that surprising if a large gain for a party is followed at the next election by a big loss. While the fluctuations in electoral results in the Netherlands are quite considerable, this rise in instability is an international phenomenon. To an important degree, the higher fluctuations go hand in hand with the emergence of new parties, especially environmental and anti-immigration parties, and the declining support for Christian Democratic and social-democratic parties. Major electoral changes are taking place not only in many Western European countries, but elsewhere as well, like in Australia and Canada.

In this chapter, we will look first of all at the function of elections in a democratic system.

In the Introduction to this volume, we reported that instabilities in the results of past studies have led researchers to call for a methodological fix – a “trick” that would give stability to their findings. We believe that one way to find this trick is to employ an approach that solves the various problems outlined in Chapter 1. Whether such an approach would be the required trick – the ideal approach that would yield stable findings – remains to be seen. By addressing the problems we have enumerated, however, we will take a major step in the right direction. On the basis of our discussion of these problems, and summarizing the conclusions we reached in Chapter 1, we can reiterate some of the properties that an ideal approach would have.

The first property of an ideal approach is that it should allow us to combine information from different countries in a single dataset that contains data at the level of individual respondents. Such a dataset would permit us to use objective national economic indicators to measure economic conditions. In order to obtain sufficient variance in economic conditions for reliable effects to be measured, the number of different elections needs to be quite large. How many elections we need to study in order to acquire adequate variation in economic conditions is not something that can be definitively stated.

In this chapter, we will develop and test a model of the effect of objective economic circumstances on individual-level support for political parties. Most of the evidence previously found for effects of the objective state of the economy on party support has been found at the aggregate level (e.g., Hibbs 1977; Tufte 1978; Chrystal and Alt 1981; Fair 1988; Lewis-Beck 1988; Markus 1988, 1992; Erikson 1989; Clarke and Whiteley 1990; Mackuen, Erikson, and Stimson 1992; Powell and Whitten 1993; Price and Sanders 1993; Clarke and Stewart, 1995; Norpoth 1996; Sanders 1996; Clarke, Stewart, and Whiteley 1998; Whitten and Palmer 1999; Campbell and Garand 2000). It has been suggested (e.g., Jacobson 1983; Lewis-Beck 1988: 29–31) that these results may have reflected the ecological fallacy first identified by Robinson (1950), making it problematic to draw inferences from aggregate data about the behavior of individuals. More to the point, in our opinion, even assuming that effects found at the aggregate level are not spurious, aggregate analyses do not allow us to assess the importance of economic circumstances relative to other considerations that affect individuals' electoral choices, such as ideological predispositions or issue preferences, and give us no way of knowing under what circumstances these other considerations might trump the effects of economic conditions.

We will start by replicating with our data models of economic voting used in aggregate analyses, in which party choice is typically predicted only by the previous vote share of government parties (to control for time-serial dependencies) and by interactions of a government party dummy variable with measures of the state of the economy.

The models of economic voting presented in Chapter 4 hardly paint a picture of Homo economicus. Effects of the economy are small compared to other effects on party support. Of course, small effects can have large consequences if they lead to a substantial segment of the electorate changing its party choice from one party to another. So, in order to assess the political consequences of changing economic conditions, we should know which groups of voters are affected by such changes. Much of the existing research has treated voters as homogeneous in the way they reach their decisions. However, some recent studies have suggested that voters are rather heterogeneous in the way they reach electoral decisions (Sniderman, Brody, and Tetlock 1991; Bartle 1997; Krause 1997; Pattie and Johnston 2001; Dorussen and Taylor 2002; Kroh 2003; Zaller 2005). One purpose of this chapter is to assess whether specific groups of voters are more responsive to changing economic conditions than others. We need this information in order to assess properly the consequences of changing economic conditions on election outcomes, which we turn to in the next chapter. Another purpose is to explore in more detail the effects of economic voting at the individual level and to evaluate those effects in the context of other effects on voters' party preferences. We will start by rehearsing the individual-level implications of our findings so far.

Science is the product not only of reason, but also of sentiments and passions. This particular book is the offspring of two opposite sentiments: disenchantment and enthusiasm. Disenchanted though we were with much of the literature in the field of economic voting, our enthusiasm derived from newly emerging possibilities for systematic comparative studies of electoral processes.

Many of the publications in the field of economic voting are highly sophisticated in their modeling and analyses of empirical data, yet – in our view – often highly implausible when approached from first principles in the light of straightforward observations of the political world, and in the light of what we believed we had learned from our own earlier comparative analyses of voters and elections. As a case in point, analysts commonly group parties in two categories, those in government and those in opposition. This crude dichotomy obviously does not reflect the kind of choice that voters are faced with in multiparty systems. Moreover, when analyzing the effects of “the” economy on voters, categorizing the world in this way has the effect of discarding and disregarding all sorts of other characteristics of parties that are likely to contribute to party choice. Moreover, election results show that, more often than not, the electoral fortunes of coalition partners diverge considerably, thus calling into doubt the simple logic that government parties are rewarded or punished for the way the economy performed under their stewardship.

The empirical data for most analyses in this book derive from the European Elections Studies (EESs) from 1989, 1994, and 1999. These were survey-based studies of the electorates of the member states of the (then) European Community (EC) in 1989 and of the European Union (EU) in 1994 and 1999. In all years, the surveys were conducted immediately following the June elections to the European Parliament. In 1989 and 1994 all twelve member states were surveyed, and all fifteen in 1999. The member states were (in 1989 and 1994) Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, and Britain; and in 1999 additionally Austria, Finland, and Sweden. The questionnaires of the surveys focused in particular on the direct elections to the European Parliament, and contained a large number of items about political topics in the domestic and European political spheres.

The EESs were designed as stratified samples of the European population, in which each of the member states is a stratum. Although technical aspects of the sampling procedures varied somewhat between the various years and between the countries, all surveys were based on random samples from the electorates in each of the countries. Sample sizes varied likewise and are reported in Table A.1. Owing to missing data on the dependent variable (propensities to vote for parties), the number of cases included in the analyses is somewhat reduced, to 32,950 (in the weighted sample – about weighting, see later).

In this book, we have demonstrated that some of the ways in which economic voting is traditionally investigated are inappropriate and bound to yield biased results, inconsistencies, instabilities, and anomalies. More specifically, we demonstrated that among the major sources of error are:

1. The use of subjective indicators of economic conditions, which are fraught with endogeneity problems and which customarily yield unrealistically high estimates of what are purported to be the effects of economic conditions. More importantly, as argued in Chapter 1, if we are interested in the extent to which voters hold governments accountable for economic conditions, then it is the real economy that is at issue, not a subjective reality that may help to determine vote choice but whose link to normative concerns, such as government accountability, is questionable.

2. The use of a dichotomy between government and opposition parties when (as is most often the case) there are more than two parties in a political system. This dichotomy is of little relevance for voters' reactions to economic conditions in so-called low-clarity countries. Moreover, it incorrectly forces the analyst to estimate a single effect for all government parties and assume an equal but opposite effect for all opposition parties. A final source of error arises from the following.

3. […]

For ease of presentation, some of the very large tables of results were not presented in the text or were presented in a condensed form. This appendix presents the full tables.

Conventional wisdom asserts that economic conditions are closely linked to election outcomes. Bill Clinton turned this conventional wisdom into a cliché when, as a candidate in the 1992 American presidential election, he had his campaign staff put up a banner that hung across their campaign headquarters emblazoned with the words “It's the economy, stupid!” This conventional wisdom is supported by much academic research. At least since the 1930s, voters in a variety of democratic countries have tended to hold governments accountable for bad economic times, reducing their support for parties holding government office in conditions of high unemployment or inflation or of low economic growth (Tufte 1978; Chrystal and Alt 1981; Hibbs 1977; Fair 1988; Lewis-Beck 1988; Markus 1988, 1992; Erikson 1989; Mackuen, Erikson, and Stimson 1992; Nadeau and Lewis-Beck 2001; Dorussen and Taylor 2002). These general findings hold whether the effects of economic conditions are modeled in terms of votes for government parties (generally referred to as “vote functions”) or in terms of government standing (generally referred to as “popularity functions”).

But it is clear that the economy does not always determine either government popularity or vote shares. In the American presidential election of 2000, Al Gore failed to win decisively as the incumbent party standard bearer despite a booming economy. In the Netherlands in 2002, all three members of the governing coalition lost votes in similarly excellent economic conditions. In Britain and Ireland in 1997, ruling parties also failed to win reelection despite booming economies.

In this chapter, we develop our argument that previous investigations into economic voting have been hobbled by model specification problems. We believe that the main problem in previous work has been a failure to focus on political parties as actors that compete for votes. This failure means that previous research has been unable to take account of the extent of party competition (which can vary from election to election). We argue that this failure to take account of party competition leads to models that are intrinsically mis-specified and that this is one of the root causes for the unstable and sometimes incoherent findings in the literature.

In what follows, we diagnose the deficiencies we see as endemic in existing studies of economic voting (many of these deficiencies apply to other studies of electoral behavior as well) and set out our ideas about what a viable approach should look like in general terms. This will set the scene for Chapter 2, in which we present our own approach to the study of economic voting and describe the ways in which it mitigates or eliminates the deficiencies of past studies.

Conceiving the Dependent Variable

Studies of economic voting have been conducted at the aggregate level, treating the country or election as the unit of analysis, and at the individual level, treating the survey respondent as the unit of analysis.

So far, we have studied effects of economic conditions on the strength of voters' propensity to support each of the parties competing for their votes. In this chapter, we move on to the second stage in the Downsean calculus of voting and assess the consequences for party choice of these changes in party support.

Our findings so far indicate that economic conditions have significant though not especially powerful effects on support for particular political parties. But the fact that these effects are not very powerful does not mean that the impact of economic conditions on overall election outcomes is necessarily small. As pointed out in the previous chapter, most of the more powerful independent variables tend to be relatively stable over the short run. The alignments of parties with sociostructural variables such as class and religion have a character that is more or less given, especially in the short term. Political parties' ideological stances do change over time, but rather slowly. The same is true of their position on issues, particularly important issues. So, while these variables largely explain differences in the attractiveness of various parties as options to vote for, the same variables cannot explain short-term changes in party choice. Economic conditions, on the other hand, can (and occasionally do) change rapidly. An apparently booming economy may come to a grinding halt in a very few months, with immediate consequences for economic growth and often for unemployment as well, consequences that are widely and rapidly experienced.