U
NEQUALR
EPRESENTATION IN THEN
ETHERLANDS THE ROLE OF SOCIO-ECONOMIC BACKGROUNDWouter Schakel (10469729) dr. Armèn Hakhverdian (supervisor) Research Master Social Sciences Amsterdam, 24 June 2015
Abstract
This study examines the existence of unequal representation on the basis of socio-economic characteristics in the Netherlands. It does so by comparing the views of different income and education groups to the views of parliamentarians. The analyses reveal stark differences in the representation of different socio-economic groups. The political views of members of parliament are far more similar to those of rich and/or high educated citizens than to those with less income and/or education. Both income and education contribute to unequal representation overall. The underrepresentation of lower socio-economic strata remains intact when controlling for factors like political participation and age. Finally, an analysis of developments over time shows that inequalities in political representation have grown in recent years. I conclude that even in a country with one of the most proportional electoral systems in the world, the views which are represented in parliament contain major distortions of views within the wider electorate.
1
Introduction
In 1961, V.O. Key wrote: “Unless mass views have some place in the shaping of policy, all the talk about democracy is nonsense” (Key, 1961: 7). In straightforward terms, this oft-quoted sentence summarizes an argument that can be traced at least back to Aristotle, namely that political equality (“one person, one vote”) forms the backbone of democracy (Dahl, 1971). Equally old is the concern that political equality is potentially undermined by socio-economic inequality (Hacker and Pierson, 2010: 75; Chomsky, 2007: 207-208). If rich and high educated citizens have a disproportionate influence on government, can we still speak of democracy? And how do we assess whether this is really the case?
These questions come at a time when socio-economic inequality is once again high on the public agenda. From the success of books like Capital in the Twenty-First Century (Piketty, 2014) and The Spirit Level (Wilkinson and Pickett, 2010) to the Occupy movement and discussions surrounding the economic crisis, it is clear that many are concerned with the causes and effects of growing inequality. In this debate, thorough and rational analysis is needed to sort through the jungle of opposing claims. The study of socio-economic biases in representation not only illuminates an important potential consequence of inequality, it can also shed light on its sources. After all, unequal influence on policy can reinforce the very inequalities that formed the basis of this influence, raising the possibility of a vicious circle (Winters, 2011). In short, the subject’s relevance extends far beyond academic circles.
Recent studies in the United States have contributed to our knowledge on the topic by comparing the preferences of low and high income groups to roll-call voting in Congress (Bartels, 2008) and policy outcomes (Gilens, 2013; Gilens and Page, 2014). They conclude that political influence is sharply stratified along income lines in the US. In the Netherlands, like elsewhere in Europe, the issue is more underexplored. To the extent that the effect of socio-economic inequality on representation has been studied here (e.g. Bovens and Wille, 2011), it has focused on education but not on income as the determinant of inequality. Moreover, and more importantly, it has not directly compared the views of income and education groups to those of political elites. Nevertheless, the Dutch case deserves to be studied, since it has one of the most proportional electoral systems in the world. Such a system is generally conducive to the representation of marginal groups in society, including the poor and low educated (Iversen and Soskice, 2006). In this sense, the Netherlands provides a least likely context for unequal representation to arise (Hakhverdian, 2015). If it exists here, it probably exists elsewhere too.
2
This study fills this empirical gap by examining the political representation of socio-economic groups in the Netherlands. Representation is interpreted here as congruence, meaning the overlap in views on various policy issues between citizens and parliament. The analyses reveal stark differences in the representation of income and education groups. The views of parliamentarians are far more similar to those of rich and/or high educated citizens than to those with less income and/or education. Both income and education contribute to unequal representation overall. On economic issues, the two factors are equally important, while education trumps the effect of income on cultural issues. These gaps remain intact when controlling for factors like political participation and age. Finally, an analysis of developments over time reveals that inequalities in congruence have grown in recent years.
These results show that even in a country with one of the most proportional electoral systems in the world, the views which are represented in parliament contain major distortions of views within the wider electorate. Returning to Key’s statement, we have to ask ourselves to what extent we can still call such a system a democracy. Although it would go too far to proclaim the Netherlands to be an oligarchy just yet, it should make us more modest when talking about ‘our’ democratic values and institutions. To anyone who values democracy, not just as a buzzword but as a practical concept, these findings should be a cause for concern.
Theory
At its core, this study is about the effect of socio-economic inequality on political representation in the Netherlands. Here, representation refers to the views that are heard in parliament. I focus on income and education as the potential causes of political inequality, for two reasons. Firstly, both income and education bring about a number of advantages that can significantly amplify one’s voice in the political arena, as I will discuss in more detail below. Secondly, it is often claimed that party positions in Western Europe are structured by two dimensions, one economic and one cultural (Van der Brug and Van Spanje, 2009; Pelikaan, De Lange and Van der Meer, 2007). The economic dimension concerns issues like redistribution and taxation; the cultural dimension includes multiculturalism and European integration, among other things. Income is an important predictor of one’s position on the economic dimension, while education is an important predictor of the cultural dimension. Hence, we would expect income and education groups to have different views on these respective dimensions, an important requirement for unequal representation to arise (Soroka and Wlezien, 2008), and both factors to illuminate different domains in which political inequality can manifest itself.
3
To understand the relationship between socio-economic inequality and political representation, we need to address the following question: how does the unequal distribution of resources flowing from education and income affect the kind of government that ends up in power? Or to put it in terms of congruence, the measurement of representation used in this study: what are the mechanisms that exacerbate or limit unequal congruence in the Netherlands on the basis of socio-economic characteristics? Although the list of potential mechanisms is surely very long, perhaps the most prominent mechanisms are descriptive representation and lobbying. These refer to the possibility that the views of rich and high educated citizens are similar to those of government actors because (a) representatives are overwhelmingly rich and high educated themselves and they act on their personal views in parliament and (b) rich and high educated citizens possess the resources needed to pressure government into acting according to their interests. Both are about political participation in the broad sense of the term. They will be explored in more detail below.
Descriptive (mis)representation
As an ideal, descriptive representation refers to the idea that “the legislature be so selected that its composition corresponds accurately to that of the whole nation” (Pitkin, 1967: 60). It is concerned with the personal characteristics of representatives and the way these characteristics relate to their actions (Pitkin, 1967: 89). The literature surrounding this subject has largely focused on the numerical underrepresentation of women and ethnic minorities (Mansbridge, 1999; Broockman, 2013), but the theme can also be applied to socio-economic variables (Hakhverdian, 2015). Among citizens, it is well known that different socio-economic groups have different political preferences and priorities (Dekker et al., 2009: chap. 6-7). If the same link exists on the level of political elites, this is the key to understanding (un)equal representation. In this view, unequal representation of socio-economic strata follows from the fact that representatives are overwhelmingly rich and high educated themselves (Carnes, 2015). But is this really the case in the Netherlands, and if so, why?
The first question is relatively easy to answer. The vast majority of parliamentarians in the Netherlands have high socio-economic status. Their educational backgrounds, in particular, have received a lot of attention in recent years. As Bovens and Wille (2011: 66-67) note, about 90% of all members of parliament belong to the group with the highest level of educational attainment, with a large majority having a university degree. Moreover, this share has increased since the 1970’s, while other distortions in descriptive representation, including those based on gender and ethnicity, have shrunk (though they have not disappeared)
4
(Van den Berg and Van den Braak, 2004: 71-72). The income and/or wealth of MP’s is harder to assess, not least because they all have more or less the same income while serving in parliament. Their economic position before and after their time in parliament is still significant, but data are not easily available for the Dutch case. Nevertheless, we know that very few members of parliament hold blue-collar positions before and after being a representative, and we know that few of them come from a poor or working class family, though the importance of social milieu seems to be declining (Secker, 1995; Van den Berg and Van den Braak, 2004: 74-78; Bovens and Wille, 2011: 69-73). In short, then, a large part of Dutch parliamentarians are high educated and well-to-do. The descriptive representation of socio-economic classes is therefore not very equal, a conclusion that generalizes around Europe (Best, 2007).
The most obvious explanation for this is that both income and education are important predictors of political participation (Lijphart, 1997; Bovens and Wille, 2011; Gilens, 2013: 239-241; Bartels, 2008: 275-276). “This systematic class bias applies with special force to the more intensive and time-consuming forms of participation” (Lijphart, 1997: 1), and holding office is perhaps the most intensive of all. Other potential explanations include the fact that wealthy and higher educated people are more likely to be a part of social networks that are important for political recruitment, through work, leisure and voluntary associations, and that they more often possess the skills and confidence to succeed as a candidate (Bovens and Wille, 2011: 79-82). Many of these skills relate to verbal and written communication. In his research on descriptive misrepresentation in the United States, Carnes (2013) emphasizes the importance of structural factors, like “the practical burdens associated with holding office and the gatekeeping decisions of party leaders and interest groups” (2013: 12). This might also apply to the Netherlands. However, one of the most prominent of these factors in the US, the fact that political campaigns are so expensive that candidates with great personal wealth are at an advantage, does not seem to apply in the Netherlands. I will return to campaign finance shortly.
All in all, there are many possible explanations behind the descriptive misrepresentation of income and education. So far, however, we simply do not know which of these explanations are the most important and how individual behavioral and psychological circumstances interact with structural factors to create socio-economic skews in descriptive representation (Carnes, 2015).
A question that remains is how we can disentangle the effect of a representative’s personal characteristics from that of external constrains, like their party and constituents. For instance,
5
wealthy members of parliament might have more right-wing views on economic issues, but they are most likely also members of a right-wing party. This makes it hard to separate the effects of personal wealth and party membership on their views (Gilens, 2013: 235). To date, Carnes (2013) has undertaken the most serious attempt to disentangle these effects, and he concludes that “the unequal social class makeup of our political institutions affects who wins and loses in the policy-making process” (2013: 3; cf. Gilens, 2015). However, these findings are limited to the United States. In the Netherlands, Hakhverdian, Van Ditmars and De Lange (2014) find that the views of less-educated representatives on the local level are more in line with less-educated citizens than the views of high educated representatives, at least on some issues. This effect remains significant when controlling for such factors as party affiliation. This provides some evidence for the conclusion that a representative’s socio-economic characteristics matter in determining their views, regardless of external constraints.1
Lobbying
Lobbying is a broad term that encompasses “all efforts outside of elections to influence government policy, through contact and communication with government officials” (Winters and Page, 2009: 740). It may be seen as a collection of several forms of political participation, most of which require a lot of time, money and other resources. Lobbying contributes to a long-term relationship with representatives (Snyder, 1992), which increases the chance that they will share the views of the lobbying party. Here, the systematic class bias in participation again applies (Schlozman, Verba and Brady, 2012). This is most obvious on the level of individuals, where rich and higher educated citizens are, on average, at an advantage due to their increased financial resources, political engagement and knowledge (Bovens and Wille, 2011: 42-46; Gilens, 2013: 239-240).
Beyond this, lobbying largely takes place on the level of institutions, of organizations like interest groups, corporations and unions. Within interest groups, wealthy and high educated activists usually dominate, a phenomenon that is reminiscent of Robert Michels’ iron law of oligarchy (Bovens and Wille, 2011: 71-73). Furthermore, the need for large financial resources means that interest groups representing moneyed interests have an important advantage over others (Van Waarden, 1992). This ties into the influence of corporations, which can manifest itself either directly or indirectly through interest groups. Their power is potentially checked by labor unions, which have traditionally represented the interests of
1 An obvious and crucial question here is whether these views affect policy. I will return to this link in the
6
lower and middle class citizens (Hacker and Pierson, 2010: 56-61). However, it is doubtful whether this is really an equal balance (Cohen and Rogers, 1983). Even if it once used to be, the decline of unions in recent decades makes it increasingly uneven (CBS, 2014).
Expert organizations can also have an important influence on the policy process. This mainly manifests itself through agenda setting and framing of issues by, among other things, “inspiring politicians [and] providing ammunition in political debates” (Scholten and Timmermans, 2010: 541). In other words, they can affect the way a politicians looks at an issue (Jacobs and Page, 2005). Since experts are usually high educated, their influence is likely to increase the congruence of high educated citizens with representatives. Furthermore, Jacobs and Page (2005) suggest that experts in the US might be influenced by business interests, which would further exacerbate their unequal influence. To my knowledge, this possibility has not been explored in the Netherlands.
The so-called revolving door between business and government can be seen as a particularly intensive manifestation of lobbying (Domhoff, 2005: 165-171). In that case, lobbyists and corporate executives have such strong ties to the government that they move freely between one and the other. Even if representatives do not directly reciprocate any past or future employment in the business world, this employment makes it likelier that they will share the mindset of their employer (Buiter, 2008). This mindset is most likely closer to that of the rich and high educated than the poor and low educated, if only because such jobs are usually highly skilled and well-paid. Since the mid 1990’s, at least half of all Dutch ministers have worked in the private sector after their time in government, showing that the revolving door is a real phenomenon in the Netherlands (Hollanders, 2015). Furthermore, members of parliament now move on to other positions much more quickly than they did before the 1990’s (Thomassen, Van Ham and Andeweg, 2014: 185-188), which could well increase their dependence on potential future employers.
Closely related to lobbying are political donations to parties and candidates. These have become increasingly prominent in the US, with the cost of presidential elections rising to billions of dollars (Center for Responsive Politics, 2012). Unsurprisingly, influential studies have identified financial contributions as the most likely mechanism behind the unequal influence of income groups in the US (Bartels, 2008; Gilens 2013; cf. Ferguson, 1995: 41). However, like most Western European societies, the Netherlands has much tighter limits on the amounts, transparency, and kinds of money that can be given by companies or private individuals to politicians, parties and campaigns (Tjernström and Katz, 2003). This rules out political donations as a likely mechanism behind unequal representation. Incidentally, an
7
intriguing possibility is that lobbying in its personal form becomes more important in settings, like the Dutch one, where rich and higher educated actors cannot use ‘impersonal’ donations to amplify their voices, but this possibility has not yet been explored.
To sum up, descriptive misrepresentation and lobbying are potentially very influential forces behind the unequal representation of socio-economic groups, even in a country with a highly proportional electoral system like the Netherlands. At the same time, there are numerous factors that could limit or work against unequal representation (see Soroka and Wlezien, 2008), including the possibility that members of parliament feel the responsibility to act on the views of poor and low educated citizens, even if it contradicts their own. I will return to this in the discussion. So far, we simply do not know the relative importance of such factors, and this lies beyond the scope of the current study as well.
Given the large gaps in our knowledge on the Dutch context, I cannot formulate many detailed hypotheses. Broadly, I expect congruence to increase with both income and education, given the previous discussion about descriptive representation and lobbying. That is, there is some degree of unequal representation based on socio-economic characteristics in the Netherlands. Beyond this, the distinction between economic and cultural issues can provide some guidance. I expect income to be a more important determinant of unequal congruence than education on economic issues and vice versa on cultural issues. The reason was stated above: opinions on economic issues are probably more strongly polarized on the basis of income than education, while opinions on cultural issues are more strongly polarized on the basis of education.
Data and methods
The lion’s share of the analysis is based on two data sources: the Dutch Parliamentary Election Studies (DPES) and the Dutch Parliament Study (DPS), both conducted in 2006. The DPES is a survey that is conducted at every parliamentary election among a random sample of Dutch citizens, focusing on a host of political issues. In 2006, 2806 respondents completed the survey. The DPS asks many of the questions from the DPES to members of the Dutch Lower House of Parliament. 114 members of parliament, 76% of a total of 150, completed the survey in 2006. An examination of this sample, presented in appendix A, shows that it is broadly representative of parliament as a whole, with some minor distortions.
The data from 2006 will receive the most attention, simply because they are the most recent and relevant to understanding contemporary Dutch politics. However, the DPS was
8
also administered to members of parliament in 2001 and 1990, which can be paired with data from the DPES of 1998 and 1989, respectively. Using these sources, I will also shed some light on developments over time in unequal representation.
Congruence
This study uses ideological congruence as a method of studying representation.2 Congruence can itself be operationalized in various ways. Until quite recently, most studies measured congruence by reducing the stance of representatives and citizens to a single point on a left-right scale and calculating the distance between them (Huber and Powell, 1994; Budge and McDonald, 2007). Despite its strengths, this approach has some major shortcomings. First and foremost, it ignores the distribution of both representatives’ and citizens’ views on the scale. As a result, this measurement cannot distinguish a centrist electorate from a highly polarized one, assuming that both extremes of the distribution are roughly equal in size. In both cases, one would conclude that a centrist government produces a high level of congruence, even though in the latter case, it would be far removed from most citizens.
This problem was remedied by the introduction of the many-to-many measure of ideological congruence by Golder and Stramski (2010). Instead of a single point, the many-to-many measure uses a distribution of preferences for both citizens and representatives on, for instance, a left-right scale. This is a better approximation of how congruence is usually conceptualized (Golder and Stramski, 2010: 105). In concrete terms, congruence is calculated by overlaying the distributions of citizens and representatives. At each point of the distribution, “we compare the percentage of voters positioning themselves at that point with the percentage of MPs positioning themselves at that same point, and we take the lower of these two percentages. If we sum the resulting (…) percentages, we have a measure for the overlap between the two distributions (‘the common area under the curve’)” (Andeweg, 2011: 43). The measure ranges between 0% and 100%. The higher the percentage, the bigger the overlap. The bigger the overlap, the better the electorate is represented.
An example of this is provided in figure 1, which presents many-to-many congruence for the entire electorate on the issue of income inequality (see appendix B for the survey question). On the far left of the figure, which indicates a preference for bigger income
2 A note on terminology: the approach is usually labeled as ‘ideological congruence’. The adjective ‘ideological’
is important to indicate its primary focus on views instead of policy outcomes. However, this study mostly casts aside broad ideological indices in favor of preferences regarding specific policy issues (see below). Such views are not necessarily part of an ideology; in themselves they are simply policy preferences. The term ‘ideological congruence’ is therefore not completely appropriate here, and most of the time I will simply call it ‘congruence’.
9
differences, the line for members of parliament lies slightly below the line for citizens, so we use the percentage of parliamentarians who place themselves here (0.9%). Moving one step to the right, the line for citizens is lower, so their relative frequency (3.6%) is added to the first number. Continuing like this for the rest of the distribution, we arrive at an overlap of 78.5%. In this study, the electorate will be divided into groups based on income and education, making it possible to see whether congruence increases as income and education go up.
Figure 1. Many-to-many congruence for all citizens on income inequality
A second problem with the conventional measurement of congruence is that it uses the left-right scale (Powell, 2009). The left-left-right scale is attractive for its seemingly all-encompassing meaning and its availability, but at the same time there is a great deal of uncertainty about what the left-right item really measures in surveys (De Vries, Hakhverdian and Lancee, 2013; see also Broockman, forthcoming). In addition, a major danger for present purposes is that citizens with low levels of income and education are more left-wing on economic issues and more right-wing on cultural issues than the average citizen. A broad index like the left-right scale would cancel out these differences and give the impression that preferences do not diverge across income and education groups (Gilens, 2013: 5). Given its prominence, I will use the left-right scale as well, to enable comparisons to previous studies. However, I will also use as many specific policy domains as possible, and I attach more importance to the latter when the two offer different findings.
Using the many-to-many measure of congruence, the analysis will not involve building any kind of statistical model. In fact, the analysis is very simple, which is simultaneously its
10
greatest strength. I will calculate many-to-many congruence for each subgroup on each issue and compare the outcomes. As a measure of inequality in congruence, I simply take the congruence of the top income and education tertiles with parliament and subtract the congruence of the bottom income and education tertiles with parliament. It is not possible to test whether this gap is statistically significant in the conventional way. Still, the estimates include sampling and measurement error, so it is important to test their significance. To do so, bootstrapping will be used to estimate a confidence interval around each congruence figure, and thus to test whether the top tertiles are significantly more congruent with parliament than the bottom and middle tertiles. Appendix C provides more information on this procedure.
Variables
In 2006, the DPS included five of the items from the DPES about specific political issues. These focus on income inequality, European integration, multiculturalism, crime and euthanasia, all measured on a seven point scale (see appendix B for the survey questions). The first of these clearly belongs to the economic dimension of the Dutch policy space, while the others belong to the cultural dimension. It might be argued that euthanasia is not a part of either dimension and is no longer a relevant issue in the Netherlands (Pelikaan et al., 2007), but if so, this will become clear in the analysis. In that case, congruence will be equal between socio-economic strata. Left-right self-placement was measured on a scale of eleven points, which was converted to seven points (see Andeweg, 2011: 42).
Income is measured using 20 categories. These categories are collapsed into three groups, which contain roughly the same number of respondents: low (<€25.155), middle (€25.155-€39.540) and high incomes (>(€25.155-€39.540). It has 18 missing values, less than 1% of the total sample. For 1989 and 1998, the threshold for income categories was changed so as to create three groups of roughly equal size in those years as well. Education is measured with five possible levels: elementary, (lower) vocational, secondary, middle level vocational / higher level secondary and higher level vocational / university. The first three of these are grouped together and the last two are left as they are, again creating three groups: low, middle and high education (Bovens and Wille, 2011: 24-25). Education has 188 missing values, which is 6.7% of all respondents. Unlike for income, the education categories were kept the same for 1989 and 1998. As a result, there are few respondents who are higher educated in 1989 (less than 5% of the sample).
11
Since there is no statistical model, it is not possible to include control variables in the conventional way. However, it is easy to think of factors that could explain potential gaps in congruence between income and education groups, and that therefore need to be controlled for. Perhaps the most obvious are political participation, political knowledge, age, ethnicity and gender (Ellis, 2012). As Ellis points out when discussing income, “much prior discussion (…) implicitly or explicitly point to these sorts of resource- or engagement-based factors as the driving forces behind biases in representation” (Ellis, 2012: 940). In other words, many assume that it is not income itself, but relevant correlates of income that are responsible for the gap. The same reasoning could apply to education.
Luckily, it is possible to include control variables ‘by hand,’ again using a very simple method. To investigate to what extent participation, knowledge, age, ethnicity and gender can explain the previous findings, I divide all these variables into two or three categories and calculate congruence for all possible combinations of, say, participation and education. I can then calculate the gap in congruence between education groups while participation is held constant. Due to this old-fashioned way of controlling for external factors, it is not possible to control for more than two variables at once; this already creates 18 to 27 separate distributions, some of which contain less than a hundred respondents.
Political participation is measured with a large number of questions, focusing on voting, joining demonstrations and contacting members of the government, among other things. Using these items, I construct three categories of participation: low, middle and high. If respondents reported voting in the parliamentary elections of 2006, the most recent municipal elections and the referendum on a European constitution, they are put into the middle category. If they did not vote in all three elections, this is coded as low participation.3 If they did vote in all three and undertook other forms of participation besides that, whatever this might have been, they are put into the highest category.
An index of political knowledge is already present in the DPES, based on the recognition of several Dutch politicians, along with their party and function. The variable with five categories is recoded into three categories by grouping together the lowest two and the highest two values. Age is also recoded into three categories. Respondents under 40 are grouped into the first category, respondents between 40 and 60 into the second, and respondents of 60 or older into the third. These values are fairly arbitrary; they are mainly chosen to create groups with comparable sizes. Next, ethnicity is divided into two categories:
3 Given the well-known overreporting of voting behavior in surveys (Schmeets, 2007), this is a less stringent
12
Dutch and Dutch. It would be interesting to distinguish Western immigrants from non-Western immigrants, but unfortunately these groups are not large enough to do so. Lastly, gender is divided into female and male.
Results
The results are structured as follows: first, congruence will be presented for two of the five policy issues, income inequality and multiculturalism, by income and education groups, respectively. This will hopefully allow the reader to understand the main analysis and its results. This picture will then be extended to the other policy issues. Afterwards, I focus on separating the effects of income and education on congruence, again using income inequality and multiculturalism as examples. Finally, two extensions to the main analyses will be discussed, namely control variables and developments over time.
Before turning to that, let us briefly look at a number of descriptive statistics for the five policy issues and left-right self-placement for both citizens and members of parliament (table 1). For the moment, these variables are treated as if they were measured on an interval level. The means show that citizens are, on average, clearly removed from parliamentarians on all issues, with left-right self-placement being an exception. With regard to income inequality, the average citizen favors smaller income differences than the average member of parliament. The average citizen also favors less European integration and multiculturalism, tougher action on crime and more freedom regarding euthanasia than the average MP. Although table 1 presents means rather than median values (which are less informative for smaller scales), it already suggests that parliament does not primarily represent the preferences of the median voter.
Income inequality and multiculturalism
The notion that parliamentarians do not primarily represent the median voter receives much stronger support when we look at congruence between groups of citizens and parliament on the issues mentioned above. We will start by looking at income inequality. Figure 2a-b displays many-to-many congruence for low and high incomes regarding income inequality. These graphs clearly show that the overlap between citizens and MP’s increases with income (69.1%versus88.8%). Theyalsoshowwherethedistributionoflowincomecitizens differs
13
Table 1. Descriptive statistics for five policy issues and left-right self-placement
Mean S.D. Min. Max. N
Citizens Income equality 5.23 1.60 1 7 2733 Eur. integration 4.69 1.82 1 7 2649 Multiculturalism 5.21 1.61 1 7 2790 Crime 6.00 1.23 1 7 2753 Euthanasia 5.90 1.66 1 7 2725 Left-right scale 4.17 1.36 1 7 2345 Parliament Income equality 4.64 1.49 1 7 108 Eur. integration 3.99 1.66 1 7 110 Multiculturalism 4.38 1.43 1 7 111 Crime 4.84 1.00 2 7 111 Euthanasia 4.74 1.60 1 7 109 Left-right scale 3.95 1.29 1 6 108
from that of parliamentarians. As could be expected, the poorest tertile of the electorate is clearly tilted towards the right of the scale, expressing a preference for smaller income differences. On the other hand, most members of parliament are found in the middle of the scale. In other words, low incomes are much more egalitarian in their preferences than parliament.
What is more surprising about figure 2 is that even the richest tertile has slightly more egalitarian views than members of parliament. This can be deduced from the fact that the line of parliamentarians lies above that of rich citizens on the left side of figure 2b, which denotes a preference for bigger income differences. On the right side, which denotes egalitarian preferences, the reverse is true. If the variable is treated as if it were measured on an interval level, its mean would be 5.51 (S.E. =0.05) for low incomes, 5.29 (S.E.=0.05) for middle incomes, 4.87 (S.E. = 0.05) for high incomes and 4.64 (S.E. = 0.14) for parliamentarians. This, too, suggests that MP’s are less egalitarian than even the richest tertile of citizens.
An obvious possible explanation of this is that parliament represents a smaller economic elite than the highest third of the income distribution. In that case, there would be inequality in congruence within this group. This is somewhat supported by the available evidence.
14
Overall, the original income variable, which has 20 categories, has a rank correlation of ρ=0.82 (p<0.001) with ideological congruence. The highest level of congruence is achieved by the top category, with a score of 90.2%. This category represents people with an income above €66.501. Obviously, the truly rich earn much more than this and it is possible that congruence is even higher for them. What speaks against the possibility of unequal congruence within the highest third of the income distribution is that these differences are not statistically significant. The topic of statistical significance will receive more attention shortly.
15
As an example of a cultural issue, figure 3a-b presents many-to-many congruence of low and high educated citizens on multiculturalism. Here, the difference between the lower and upper tertile is even bigger than in the figures related to income inequality. For low educated citizens, congruence with parliamentarians is mediocre at best (58.7%). The former clearly have more monocultural views than the latter. For high educated citizens, on the other hand, congruence is almost perfect (94.1%). Seeing as minor variations will always exist between the two distributions, for all intents and purposes, parliamentarians and high educated citizens have identical views on multiculturalism.
16
All issues
The findings so far generalize to the other issues (European integration, crime and euthanasia). Figure 4 presents ideological congruence by income tertiles for all five issues, and figure 5 does the same for education tertiles. Error bars indicate the 95% confidence interval. The striking conclusion from these figures is that, in terms of views, the representation of low educated citizens is significantly worse than that of high educated citizens on every issue. The same goes for income groups, with the exception of euthanasia, where the difference between low and high income is not significant (t= 1.30, p=0.193). Furthermore, the middle tertile is between the other two for all five issues in both graphs. In the case of education, this middle group scores significantly lower than the highest group on European integration (t=2.55, p=0.011), multiculturalism (t=3.30, p=0.001) and crime (t=3.05, p=0.002). In the case of income, the difference between middle and high incomes lies on the edge of significance for income inequality (t = 1.92, p =0.055) and European integration (t=1.84, p=0.066). Given the strict nature of the significance test, it might well be the case that these values indicate real effects.4
Overall then, these findings point to a clear, positive relationship between income and education on the one hand and congruence on the other. If we take the average of all five policy issues, many-to-many congruence increases from 61.4% for low incomes to 75.5% for high incomes. This difference is even bigger in the case of education, where average congruence increases from 56.7% for less educated citizens to 81.9% for high educated citizens. For all citizens and all issues combined, congruence equals 68.3%.
This average is arbitrary in the sense that it depends on the choice of issues in the surveys. Hence, what is important here is not that the average difference between education groups is bigger than that between income groups; this reflects the fact that only one of the five issues is clearly economic in nature. What is important is simply that, according to this measure, there are clear inequities in congruence on the basis of both income and education.
This conclusion is very different when we look at left-right self-placement, which differs from the previous measure in two crucial regards: congruence is much higher overall (89.4% versus 68.3%) and the differences between groups are much smaller. The latter is especially true for income, which shows no difference between tertiles whatsoever (comparing low and high incomes, t = 0.19, p= 0.847). There are differences between education groups, with congruence increasing from 80.0% for low educated citizens to 96.0% for high educated
4 All other t-values are provided in table C1 in appendix C. This appendix also explains why the current
17
Figure 4. Congruence for five policy issues by income tertiles
Figure 5. Congruence for five policy issues by education tertiles
citizens (t=3.00, p=0.003), but the distance is much smaller than before. Furthermore, even the preferences of less educated citizens appear to be represented quite well in parliament if we look at the left-right scale. These results are presented graphically in appendix D.
Since the policy issues arguably provide a less ambiguous and abstract basis for the measurement of congruence, I conclude that using the left-right scale would lead to an overly rosy picture of representation. It would suggest that congruence is “nearly perfect” (Andeweg, 2011: 47) and “better than ever” (Thomassen et al., 2014: 60), and that there are only limited gaps between groups of citizens. In fact, my analysis offers less ground for optimism. From here on out, the left-right scale will be put aside.
0 10 20 30 40 50 60 70 80 90 100
Equality Eur. int. Multicult. Crime Euthanasia
C ong rue nc e Low income Med. income High income 0 10 20 30 40 50 60 70 80 90 100
Equality Eur. int. Multicult. Crime Euthanasia
C ong rue nc e Low educ. Med. educ. High educ.
18
Disentangling income and education
So far, it seemed like income and education both contributed to inequalities in congruence. This is not a conclusive answer though, because income and education partially overlap and the graphs do not control for this. In the DPES, they share a rank correlation of ρ = 0.26 (p < 0.001) (or ρ = 0.28 (p < 0.001) if the original variables are used). To separate the effects of income and education, I make nine categories for every possible combination of income and education and calculate congruence for each category. Income inequality and multiculturalism are again used as examples.
Figure 6 presents congruence with parliament on income inequality for all nine combinations of income and education. Congruence increases along both the income axis and the education axis, providing further support for the claim that both factors are important here. Keeping education constant, there is a difference of 13.5 percentage points between low and high incomes. If income is kept constant, the difference between low and high education is 13.4 percentage points. Income and education therefore have very similar effects on congruence regarding income equality. For both, the difference between lower and upper tertiles remains within the range of statistical significance.
Figure 6. Congruence on income inequality by income and education tertiles
Low educ. Med. educ. High educ. 0 10 20 30 40 50 60 70 80 90 100 Low income Med. income High income C ong rue nc e
19
For multiculturalism, we can also calculate congruence for all nine possible combinations of income and education. The resulting bar graph is shown in figure 7. Unlike before, income and education clearly do not contribute equally to differences in congruence. There is hardly any increase along the income axis, while the increases along the education axis are very large. Keeping education constant, there is a difference of 5.2 percentage points between low and high incomes, which is no longer significant. If income is constant, the difference between low and high education is 30.9 percentage points, which is still highly significant. This means that education can explain almost two thirds of the income effect, whereas income can only explain about a tenth of the education effect. Education therefore has an extremely strong effect on congruence regarding multiculturalism, while the effect of income is negligible.
Additional analyses show that the same is true for the other cultural issues. On crime, low and high incomes differ by 7.2 percentage points when controlling for education, compared to a difference of 30.7 percentage points between education groups. These figures are 4.0 and 14.7 percentage points, respectively, for euthanasia. Of the four cultural issues, income has the strongest effect on European integration, but even here, the effect of education is more than twice as large (9.2 versus 21.3 percentage points). In all cases, the education effect remains significant, while the income effect does not.
Figure 7. Congruence on multiculturalism by income and education tertiles
Low educ. Med. educ. High educ. 0 10 20 30 40 50 60 70 80 90 100 Low income Med. income High income C ong rue nc e
20
The explanation for this is quite simple and in line with what was expected: education groups differ much more in their opinions on cultural issues than income groups. Put differently, education is a more important determinant of one’s views on these issues than income. This can be shown in different ways. For instance, the average scores for education tertiles differ by almost an entire point on an issue like European integration (5.11 versus 4.17), on a scale that ranges from one to seven. For income tertiles, this gap is less than half a point (4.88 versus 4.43). Another way to show it is to look at congruence between income and education groups, so the overlap in opinion between low and high incomes, and low and high educated voters. The congruence between low and high incomes on European integration is 84.2%. Between low and high education, it is 75.4%. On the other hand, the gap in opinion between income and education groups is very comparable when it comes to income inequality. The average of the lowest and highest group differs by 0.64 and 0.61 points, respectively, and congruence between the lower and upper tertiles is 79.5% and 74.9%.
It should be noted that euthanasia differs from the other issues in the sense that unequal congruence is not accompanied by differences in the average of lower and upper tertiles. In fact, these averages are almost identical for all groups; it seems that everyone is quite progressive. The finding that the views of low educated citizens are nevertheless less congruent with parliament than the views of high educated citizens results from the fact that the former place themselves in the highest category more often than the latter. Still, they are not any more progressive, as the averages show, so on closer inspection the differences in congruence are not very meaningful for this issue. The simple truth is that all groups are represented quite poorly when it comes to euthanasia. I will return to this in a moment.
The above discussion carries an important implication with it: the views of the bottom two thirds of the electorate in terms of income and education are only reflected in parliament insofar as they overlap with the views of rich and higher educated citizens. In other words, those with low levels of income and education are not represented directly but only indirectly to the extent that they happen to agree with those above them.
This conclusion mostly follows naturally from the previous findings. A more explicit way to show it is to look at the difference in congruence with parliamentarians between the bottom two tertiles and the highest tertile. For instance, congruence with parliament regarding income equality is 69.1% for low incomes, 77.5% for middle incomes and 88.8% for high incomes, so the difference in congruence with the highest tertile is 19.7 percentage points for low incomes and 11.3 percentage points for medium incomes. If this is compared to congruence between groups, for instance, the extent to which low incomes overlap with high
21
incomes on the issue of income inequality, we find a near perfect relation (r = 0.96,
p<0.001, n=20). This means that, the more low incomes disagree with high incomes, and low educated citizens disagree with high educated citizens, the less their views are reflected among parliamentarians.
Control variables
A reason to be cautious of these findings is that none of the analyses have included control variables so far. I will turn to that now. As was described earlier, it is possible to control for political participation, knowledge, age, ethnicity and gender by dividing all these variables into two or three categories and calculating congruence for all possible combinations of, say, participation and education. For the cultural issues, I only focus on the educational effect, since the income effect lost significance when controlling for education.
Even if we assume that all the above variables are completely exogenous to income and education, which is surely an unrealistic assumption (as the theory section has reviewed), most of the effects remain significant. At most, the control variables can explain about a third to half of the effects found so far. With regard to income inequality, political knowledge and education reduce the income effect from 19.8 percentage points to 11.1 percentage points, a reduction of 44%. This gap is on the edge of significance. For the same issue, the effect of education is reduced from 17.5 to 9.3 percentage points by age and income, a reduction of 47%. The educational effect on European integration is lowered by 32% by political knowledge and income. The same two variables have the biggest impact with regard to crime, where they explain 30% of the educational effect. Political participation and age reduce the effect of education by 35% for multiculturalism. Finally, ethnicity explains 43% of the educational effect for euthanasia. All other combinations of control variables have a smaller impact on the effects of income and education. The full results are presented in appendix E.
Hence, the introduction of control variables does not fundamentally change the main finding so far, which is that there are clear gaps in congruence between income and education groups. The control variables obviously reduce the size of the effects, but even when they are taken into account, the gaps between lower and upper tertiles are in double digits, with the exception of the educational effect on income inequality and euthanasia. The latter two are the only two which clearly turn non-significant after introducing control variables.
A more realistic and less stringent interpretation of these results, however, acknowledges that some of these control variables are endogenous to education and income. This is obviously not the case for age, ethnicity and gender, but it is true for political participation
22
and knowledge. That is, the effects of education and income on congruence partly work
through political participation and knowledge. In that case, I conclude that 15% to 30% of the
effects are mediated by participation and knowledge, and exogenous factors explain less than 20% of these effects for most issues.
Developments over time
Finally, the current data makes it possible to sketch developments over time in ideological congruence, albeit in a limited way. The same survey was administered to members of parliament in 2001 and 1990, which can be paired with data from the Dutch Parliamentary Election Studies of 1998 and 1989, respectively. Of the five issues discussed above (income equality, European unification, multiculturalism, crime and euthanasia), crime is the only one that was not included in previous surveys. The other four will be discussed in order. Figure 8a-d summarizes the findings by presenting many-to-many congruence for low and high incomes on the four issues over time. Figure 9a-d does the same for low and high education. Error bars indicate the 95% confidence interval. All figures, including those for the middle groups, are provided in appendix F. Appendix G provides the average scores for the same issues and time points.
Congruence regarding income equality did not display wide gaps between income and education groups before 2006. In 2006, low and high incomes were separated by 19.7 percentage points. By comparison, the gap was only 5.4 percentage points in 1998/2001 and 10.3 percentage points in 1989/1990. Even more remarkable is the finding that the middle tertile had the highest congruence of all groups at both points in time, though the differences are small and non-significant. The same pattern emerges when we look at education groups.
These developments cannot be explained by the fact that low and high incomes were closer to each other before 2006. In fact, they were not. In 2006, the gap in their average position was 0.61 points (t = 8.38, p < 0.001), compared to 0.87 points (t = 7.26, p < 0.001) in 1989. The average scores show that both parliamentarians and the public became slightly more egalitarian in their views between 1989/1990 and 1998/2001. This development continued for citizens after that but was reversed for members of parliament. This explains why congruence became clearly unequal in 2006. Congruence remained high for the top tertiles of income and education, while it dropped for the middle and bottom tertiles.
23
Figure 8a-d. Developments in congruence over time for income groups
Figure 9a-d. Developments in congruence over time for education groups
50 60 70 80 90 100 C ongr ue nc e 1989/1990 1998/2001 2006
(a) Income inequality
50 60 70 80 90 100 1989/1990 1998/2001 2006 (b) European integration 50 60 70 80 90 100 C ongr ue nc e 1989/1990 1998/2001 2006 (c) Multiculturalism 50 60 70 80 90 100 1989/1990 1998/2001 2006 (d) Euthanasia
Low income High income
40 55 70 85 100 C ongr ue nc e 1989/1990 1998/2001 2006
(a) Income inequality
40 55 70 85 100 1989/1990 1998/2001 2006 (b) European integration 40 55 70 85 100 C ongr ue nc e 1989/1990 1998/2001 2006 (c) Multiculturalism 40 55 70 85 100 1989/1990 1998/2001 2006 (d) Euthanasia
24
European integration was measured at one previous point in time, in 1998/2001. In this year, congruence was less unequal than in 2006, but there were still clear gaps between income and education groups. Low and high educated citizens were separated by 15.4 percentage points in 1998/2001, low and high incomes by 13.0 percentage points. By comparison, these gaps were 24.6 and 14.9 percentage points, respectively, in 2006. The average scores show that both parliamentarians and the public became more skeptical towards European integration between 1998/2001 and 2006. The increase in unequal congruence for education groups in particular is due to the fact that low and high educated citizens drifted further apart in their views (a gap of 0.48 points (t=4.94, p<0.001) versus 0.93 points (t=9.94, p<0.001)). Here, too, congruence remained high for the top tertiles of income and education, while it decreased for the middle and bottom tertiles. On average, parliament is more in favor of European integration than even higher educated citizens at both points in time.
Turning to multiculturalism, we find that gaps between education groups in particular became bigger over time. In this sense, it is similar to European integration. In 1998/2001, the gap in congruence was 17.1 percentage points for education and 15.3 percentage points for income. By 2006, this had become 35.4 and 14.3 percentage points, respectively. All groups became more monocultural in their views between 1998/2001 and 2006, except for parliamentarians, whose views remained unchanged. In 1998/2001, they were more monocultural than high educated citizens. By 2006, they had become relatively more multicultural. As a result, congruence sharply dropped for low educated citizens, while it actually increased for those with a high level of education.
Finally, euthanasia is interesting more for the development of overall congruence than for the development of inequalities in congruence. Like income inequality, its measurement goes back to 1989/1990. This reveals, firstly, that the average member of the public was always somewhat more progressive on this issue than parliament. Secondly, all income and education groups became more progressive from 1989 to 2006, which does not come as a surprise. Thirdly, high educated citizens used to be slightly more progressive than low educated citizens (t=-3.20, p=0.001), and high incomes used to be more progressive than low incomes (t=-2.80, p=0.005), but this gap had closed by 2006 (t=0.27, p=0.789 and
t=0.38, p=0.706 respectively). Essentially, everyone had become very progressive by then. This is in line with the view that moral issues like euthanasia no longer form a conflict line in the Dutch electorate (Pelikaan et al., 2007). Fourth and most interesting is the finding that, on average, parliament did not become more progressive between 1989 and 2006. As a result,
25
congruence displayed a dramatic drop between 1998/2001 (84.4%) and 2006 (57.9%). As was said earlier, everyone is represented quite poorly, in this sense of the word, when it comes to euthanasia. Perhaps this is a legacy of the Christian parties.
Overall then, congruence became more unequal through time for three of the four issues considered here, although European integration and multiculturalism still displayed clear inequities in 1998/2001. In the case of income inequality and multiculturalism, parliament failed to follow the developments of broad sections of the electorate, which mainly affected congruence with low educated and poor citizens. In the case of European integration, the views of education groups diverged between 1998 and 2006.
Conclusion and discussion
A place where every interest and shade of opinion in the country can have its cause even passionatelypleaded,inthefaceofthegovernmentandofallotherinterestsandopinions(…), is in itself, if it answered no other purpose, one of the most important political institutions that can exist anywhere, and one of the foremost benefits of free government.
– John Stuart Mill (1861: 105)
This study has focused on the unequal representation of socio-economic strata in the Netherlands, as measured by congruence with parliamentarians. The analyses have revealed clear and significant differences between income and education groups. In terms of congruence, rich and/or high educated citizens are better represented than the rest of the electorate. Essentially, the only way people with low socio-economic status can have their views reflected in parliament is to agree with those who have more status. To the extent that they disagree, the former are not represented.
Both income and education contribute to unequal representation overall, which was expected. On economic issues, the two factors are equally important, while education trumps the effect of income on cultural issues. The latter was expected; the former was not. These effects mostly remain intact when controlling for factors like political participation and age. Alternatively, participation and knowledge can be seen as mechanisms through which income and education manifest itself, in which case they explain a modest part of the effects. Finally, a discussion of developments over time shows that inequalities in congruence have grown in recent years.
26
By Mill’s standard, parliament in the Netherlands therefore fails to live up to one of its most important roles. Not every “interest and shade of opinion” is represented and “passionately pleaded”; in fact, some views are all but absent. This is troubling news to any proponent of democracy. Even though these findings do not necessarily apply outside of the Netherlands, the fact that unequal representation can arise in such a proportional system does not bode well for other countries.
The results are made even more significant by the fact that they are based on data from 2006, that is, after the rise of Pim Fortuyn and populism in the Netherlands. A common narrative maintains that, after the established parties had ignored various and growing concerns on the cultural dimension for years, Fortuyn and others like him formed a “self-correction of the representative democracy” (Thomassen et al., 2014: 64). However, even after this supposed correction, the views of parliamentarians mostly reflect those of high educated citizens. In fact, this has only become more apparent over time. One could argue that 2006 is an exceptional year, because it comes after the height of Fortuyn and before the height of Wilders, the new figurehead of populism in the Netherlands. However, the rise of populist parties forced established parties to move closer to their positions on issues like multiculturalism, an effect that should already have taken hold by 2006 (Van Spanje, 2010).
As for the economic dimension, Bovens and Wille (2011: 99-101) suggest that congruence is high and equal here, because it concerns ‘easy issues’ (see also Thomassen et al., 2014). This is not supported by my analysis. Perhaps this strange claim, along with their sole focus on education, is part of a broader development among academics, what might be called an overreaction: after not paying much attention to cultural issues before the turn of the century, it was virtually all they could see for more than a decade. However, economic issues and characteristics are still as important as ever. This is even truer when we consider the situation after 2006. While I suspect that congruence has become somewhat less unequal on the cultural dimension since then, due to the continued drift of most parties towards conservative positions, I expect congruence to have become more unequal on the economic dimension, because most parties have shifted to the right in the past decade (Van Spanje, 2010).
Needless to say, this study is not without limitations. The most important of these is that the exact nature of the causal relationship behind unequal representation, of which this study has explored only a part, remains unknown. Descriptive representation and lobbying are among the most prominent potential explanations of the effects found here, but we can only speculate at this point. However, if we do not know what causes inequalities in congruence, it
27
is nearly impossible to figure out how to combat them. This should therefore be a priority in future research on the topic.
Another part of the causal chain, and another caveat, concerns the link to policy. This study has focused on the link from socio-economic inequality to the views that are represented in the government, but an obvious and crucial question remains: if the unequal distribution of material resources affects the kind of government that ends up in power, does this actually make a difference for the policy that is enacted? Another way to put the question is in terms of Pitkin’s (1967) discussion of representation in its various dimensions: if socio-economic inequality affects descriptive representation, does it also influence substantive representation (Pontusson, 2015)?
One reason to ponder this question is that it is not completely clear under which dimension of representation ideological congruence should fall. The overlap in ideology or views between the electorate and political elites is too close to policy to fall squarely under descriptive representation, yet too far removed from it to be labelled substantive representation. It is probably closer to the former than the latter, especially as it has been measured here. Despite this, several prominent studies imply that there is no difference between congruence and substantive representation (Huber and Powell, 1994: 292; Golder and Stramski, 2010: 105). It is, however, not self-evident that different levels of congruence produce different policy outcomes. Could it be that the views of representatives do not decisively affect their actions? And if so, are inequalities in congruence between income and education groups really a cause for concern? Although equal congruence is a good in itself, as Mill argued, we are mainly concerned with representatives’ views because we assume these views will guide their actions (Pitkin, 1967: 89).
We can imagine several reasons why a politician’s actions could diverge from their personal views. The first is party discipline. Although the party line might differ from the representative’s views on certain issues, we should expect the two to coincide on most issues. Otherwise, it is hard to imagine why someone would represent that party. This is even truer for parliamentarians, who are presumably long-time members of their party and strongly identify with it. The second reason is that other actors, including interest groups, local politicians and international organizations, constrain members of parliament and force them to act against their own views. If anything, however, such less visible actors and stages of the policy process are likelier to exacerbate unequal representation, in the substantive sense of the word, than diminish it (Bartels, 2008; Gilens, 2013). Finally, some politicians might feel the responsibility to act on the views of the electorate, even if it contradicts their own. This
28
cannot be ruled out. What speaks against it is the fact that only a small minority of parliamentarians in the Netherlands see themselves as delegates – representatives bound by a strict mandate from the electorate which leaves little room for their own views (Andeweg and Thomassen, 2007). The trustee role is far more common, and here it is much likelier that representatives will act on their own views, judging it to be for the best in the long run, as we all do with our convictions.
Even though no solid evidence exists of a link between representatives’ views and actions, certainly not for the Netherlands, unequal congruence therefore provides a major indication of inequalities in substantive representation (Hakhverdian, 2015). At the same time, the policy link is another important avenue for future research.
Finally, a part of the message behind this paper concerns the study into political inequality and elites in the Netherlands. At the end of The Power Elite, C. Wright Mills writes: “Were we to select our field of study according to the ready availability of much unworked material, we should never choose the elite. And yet, if we are trying to understand something of the true nature of the society in which we live, we cannot allow the impossibility of rigorous proof to keep us from studying whatever we believe to be important” (Mills, 2000: 382). With this in mind, the scant attention given to unequal representation by scholars in the Netherlands should be a source of shame. The mere fact that the analysis in this paper had not been done before, despite it requiring no original data collection or advanced knowledge of statistics, is telling. Mills hit the nail on the head, as he often did, and his argument should be taken to heart.
References
Andeweg, R.B (2011), ‘Approaching Perfect Policy Congruence: Measurement, Development, and Relevance for Political Representation’ in M. Rosema, B. Denters and K. Aarts (eds.), How Democracy Works: Political Representation and Policy Congruence in Modern Societies, pp. 39-52. Amsterdam: Amsterdam University Press.
Andeweg, R.B. and Thomassen, J.J.A. (2007), Binnenhof van Binnenuit: Tweede-Kamerleden over het Functioneren van de Nederlandse Democratie. Den Haag: Raad voor het Openbaar Bestuur. Bartels, L.M. (2008), Unequal Democracy: The Political Economy of the New Gilded Age. Princeton:
Princeton University Press.
Best, H. (2007), ‘New Challenges, New Elites? Changes in the Recruitment and Career Patterns of European Representative Elites’ Comparative Sociology, 6 (1): 85–113.
Bovens, M. and Wille, A.C. (2011), Diplomademocratie: Over de Spanning Tussen Meritocratie en Democratie. Amsterdam: Bert Bakker.
