1
Joshua Stom 11002360
J.stom.edu@gmail.com Gordon Arlen (Supervisor) Eric Schliesser (Second Reader) BSc. Political Science Thesis June 25th, 2018
DESCRIPTIVE REPRESENTATION: MAKING THE LOWER EDUCATED PRESENT IN PARLIAMENT
ontemporary literature on descriptive representation is often motivated by rectifying historical injustices and marginalisation for groups defined by their gender or ethnicity. However, the arguments these authors invoke to address such injustices are also applicable to another marginalised group in society: the lower educated. Indeed education, more than gender or ethnicity, determines levels of political participation, trust, and apathy. Moreover, making the lower educated present in representative assemblies will enhance its quality of substantive representation from an ideal of aggregation and deliberation. Thus, if we seriously want to increase descriptive representation for the sake of its ideals, we must also reconsider the absence of lower educated people from representative assemblies.
1. Introduction.
Contemporary literature on descriptive representation (Mansbridge 1999; Conti; 2018; Phillips 1995) or its close adherent, group representation (Kymlicka 1995; Williams 2000 Young 2002), is mostly focussed on the rectification of historical injustices and marginalisation experienced by groups defined by their gender or ethnicity. Throughout history, women and ethnic minorities have experienced both institutional and cultural forms of discrimination. Although these groups are now formally equal to their white male counterparts, this history of marginalisation still impacts their presence in politics today. In the United States, slightly more than half of the electorate is female, whereas in March 2018, only 19.3% of its Lower Chamber’s seats were occupied by women. The statistics are moderately better for African Americans: 9.8% of the representatives in the Lower Chamber are African American, a group that composes around 14% of the US´ population.
2
If there were no obstacles in place to keep women or ethnic minorities out of the political realm, we would expect official political positions to be more randomly distributed among all gender groups and ethnicities. However, since the statistics show that this distribution is significantly distorted beyond some minor deviations, we must conclude that both groups are being structurally discriminated against (Phillips 1995: 63). In liberal democracies that regard men, women, and non-whites to be equally capable of occupying governmental positions, this situation is unjust and a cause for action.
When compared to the American case, ethnic minorities and women in the Netherlands are far more equally represented in representative assemblies. Its proportional electoral system, which simultaneously has no election threshold, allows for minority interests to be more easily represented. The Dutch Second Chamber contains an animal rights party, a highly conservative religious party, and since the elections of 2017, an immigrant party (the first ever in Western party politics). Moreover, party lists enhance the chance of electoral success for women and ethnic minorities, unlike in majority systems like the US, where it is difficult for such groups to top the list in their constituency (Philips 1995: 83).
Yet, despite the presence of a highly proportional electoral system in the Netherlands, another minority is completely absent: the lower educated. Recent works by Dutch scholars Armen Hakhverdian (2017) and Mark Bovens (2014) show how highly educated Dutch parliamentarians are. As of 2012, more than 90% of the Dutch representatives either finished Hoger Beroepsonderwijs (university of applied sciences) or university, while only 28% of the Dutch population is highly educated. (Idem: 53). The lower educated, citizens who finished their VMBO (preparatory secondary vocational education) or below, are completely absent (CBS 2008).
To be sure, no elected parliament is ever a true microcosm of the entire electorate. Parliamentarians will always differ from their constituency in some respect, be it gender,
3
ethnicity, education level, or even arbitrary characteristics like hair colour (Mansbridge 1999: 631). However, this fact does not necessarily imply that every bit of unlikeness between parliamentarians and the population is problematic. Within the tradition of descriptive representation, scholars have already argued why the underrepresentation of women and ethnic minorities is problematic (Phillips 1995: 27-57; Kymlicka 1995: 106-130). In this paper, I will problematise the descriptive underrepresentation of another group in society, namely the lower educated.
Like my predecessors who have theorised descriptive representation, I will present an argument that goes beyond the quality of substantive representation. Not only is the political absence of the lower educated detrimental to the substantive representation of representative assemblies, structural discrimination in the political (and social) realm causes the lower educated to dissociate from the political regime, threatening its very legitimacy. Like gender or ethnicity, education is an important explanatory variable of political apathy. Lower educated people participate less and are significantly less trusting of politics. Thus, I argue that if we seriously want to reconsider descriptive representation in our democracy, we must not only reconsider the role of women and ethnic minorities in politics but also the role of the lower educated. We must do so both from an ideal of substantive representation, of social recognition, and of democratic legitimacy.
Firstly, I will utilise contemporary literature on descriptive representation to determine to what extent this literature is applicable to the case of the lower educated. I will expose the ways in which lower educated people have been historically marginalised and excluded from the political decision-making table, and I will argue that this process of historical marginalisation has caused the lower educated to become alienated from the political regime.
Secondly, I will assess the substantive representativeness of the Dutch Second Chamber from an ideal of aggregation and deliberation. For contemporary scholars like Jane
4
Mansbridge (1999: 5), the quality of substantive representation is an even more important ideal than rectifying historical injustices. If it is found that an all highly educated Parliament does a better job at representing the interests of the lower educated and the electorate as a whole, this would be a major argument against descriptive representation. Consequently, I will analyse whether the absence of lower educated people from Parliament is detrimental to its substantive representation.
Scholars of descriptive representation often deviate from the aggregative model of democracy because it tends to reproduce inequalities that possibly have an oppressing effect on minorities (Young 2002: 26). For this reason, I will devote the third part of my argument to a justification of the assumption that enhanced representation from an ideal of aggregation and deliberation is both desirable in representative democracies.
Lastly, I will resolve a major issue that problematises the presence of lower educated people. The principle of competence implies that representative democracies should give as much weight as possible to superior intelligence and virtue in the political process (Thompson 1976: 54). In representative democracies, it is assumed that gender and ethnicity are arbitrary variables to determine such virtues. For education, however, this claim is more problematic since political competence is often solely prescribed to the highly educated. By offering a fundamental critique of this notion, I will argue against the principle of competence. Together, all parts will form a comprehensive argument why the lower educated should be made present in representative assemblies.
2. Descriptive representation
Descriptive representation is often defined as a democratic ideal. However, the political culture of Britain during the 1830s’ Reform era shows that supporters of descriptive
5
representation originally resisted democracy (Conti 2018: 1). Let us further elaborate on this history. Advocates of descriptive representation argued that a representative body can only be truly representative if its composition corresponds accurately to that of the entire nation (Pitkin 1967: 60). In Whig-liberal Bernard Cracroft’s words (1866): “The representation of the country would be perfect if every class, every interest, every opinion was represented in exact proportion to its weight and worth.”
The nineteenth-century adherents of descriptive representation were not considered democrats because they believed that democracy (understood as uniform universal suffrage), enabled the working class to outnumber the diversity of socio-economic interests in all electoral districts. The diversity of social interests, however, was deemed important from a deliberative ideal. It was believed that having all parts of society present, resulted in rich debates wherein citizens were enabled to transcend their own personal characteristics and encounter the nation as a whole (Conti 2018: 4). Thus, no matter how virtuous a Parliament was, it would still be considered non-ideal if not all sections of the community were present.
Yet, the democratic principle of one-man-one-vote won out over the ideals of descriptive representation. Advocates of descriptive representation wanted to give due weight to each interest in society but were never able to overcome its main objection: according to the democrats, no standard existed by which this ‘due weight’ could be measured (Dicey 1867: 71). They believed that the selection of groups who qualified for affirmative actions would be highly susceptible to arbitrariness (Conti 2018: 16). What makes the interests of left-handers more salient than the interests of redheads? Democrats converged on the idea that the only fair mechanism to determine the ‘due weight’ of each interest was to give every man an equal vote in the electoral process.
A similar debate was held during the ratification debate of the US’ Constitution (Manin 1997: 102). Anti-Federalists proposed that the ratio between voters and
6
representatives was too big to allow any likeness between representatives and the electorate. Anti-Federalists favoured smaller districts because they feared a process wherein only the natural aristocrats would get elected. On the other hand, James Madison and the Federalists argued that likeness would be detrimental to the assembly´s quality of representation. Rather, representatives should be superior to the electorate; they ought to exceed voters in talent, wealth, and virtue (Idem: 110). With these characteristics, the Federalists argued, representatives would be able to think beyond partial interests and reach agreements in a more competent manner. Meanwhile, representatives always remained accountable to the electorate because the continuity of their power ultimately depended on re-elections (ibid.).
We might distinguish two forms of descriptive representation: a microcosmic form, favoured by the Anti-Federalists (but also by American founding fathers John Adams and James Wilson), and a selective form (Mansbridge 1999: 631). In microcosmic representation, representatives form a true microcosm, this is to say, a true representative sample of the entire electorate (ibid.). The only way to achieve microcosmic representation, however, is by selecting representatives through lot. This implies that microcosmic representation cannot exist in representative democracies which utilise election (ibid.). The principle of one-man-one-vote inevitably leads to a bias towards candidates that possess characteristics the electorate deems important for running for office (Manin 1997: 134). These preferences are never randomly distributed but skewed towards characteristics like wealth and expertise, or even more arbitrary characteristics like gender or ethnicity.
Secondly, we have selective forms of descriptive representation. Here, certain groups in society are given more descriptive representation in assemblies than they would achieve through normal electoral selection. These affirmative actions are often employed to bring the proportion of these groups in the legislature closer to their proportion in the population (Mansbridge 1999: 632). Put differently, selective forms of descriptive representation
7
compensate for the effects of another process that interferes with an expected proportionality that would otherwise be achieved if the legislature was to be composed randomly (ibid.). Again, this form of representation is incompatible with the principle of universal suffrage and ‘one-man-one-vote’. In electoral democracies, individual voters can decide which characteristics are desirable for representatives to have, no matter how arbitrary these characteristics may be (Manin 1997: 136).
In the context of these historical debates, let us now turn to the contemporary debate. Like their early-Victorian predecessors, contemporary scholars still argue for more likeness in politics from an ideal of deliberative democracy. However, their motivation to develop such arguments differs significantly. Anne Philips (1995: 4) notices that the triumph of democratic ideals over the ideals of descriptive representation shifted the emphasis from who a politician is (in the Athenian democracy), to what a politician represents (policies, preferences, ideas). Hence, the quality of representation is now based on mechanisms of accountability that bind politicians more closely to the stances they claimed to represent during elections. However, Phillips argues that what is left out in this picture is a widely felt exclusion by historically marginalised groups characterised by their gender or ethnicity (Kymlicka 1995: 109; Young 2002: 37). Consequently, contemporary scholars are often motivated by a desire of rectifying these historical injustices and refer to them when justifying enhanced descriptive representation.
Approaching descriptive representation from within an ideal of rectifying historical injustices helps to overcome an important problem: how should democracies select the groups
which qualify for enhanced descriptive representation? The standard by which groups can
claim more presence in politics is whether they are disadvantaged in the electoral process due to historical marginalisation. Furthermore, analytical frameworks to reveal marginalisation have been employed, making it possible to decide which groups satisfy this criterium. In the
8
next section I will elaborate on one of such frameworks: Iris Young’s five faces of oppression (1990: 37).
Yet, even after this development, some fundamental critiques remain. The remaining paragraphs are dedicated to these critiques and their responses. Firstly, let us consider the problem of accountability. As I explained before, for descriptivists, the quality of representation depends on a representative’s descriptive characteristics; on what he is or is like. (Pitkin 1967: 61). However, the mere fact that a representative “looks” like you, does not guarantee that this person will similarly represent your substantive interests. How then do we keep representatives accountable to the group they ought to represent?
Although Anne Phillips (1995: 76) acknowledges that representation based on descriptive characteristics does not offer binding mechanisms through which the quality of substantive representation can be guaranteed, she goes on to argue that a representative’s descriptive characteristics remain important indicators of how one might act in the assembly. After elections, voters expect their representatives to pursue policy that is in line with the ideas they had when they asked for the electorate’s vote. However, four-yearly electoral competition offers only a limited mechanism through which the electorate can express its views (ibid.). This means that representatives cannot be mere cyphers of voters because they will come across issues that have not been dealt with during the elections. Thus, for representatives to act on our behalf, we must afford them some autonomy and trust, and here it matters whether our representatives are like us or not because it gives us an indication of how the representative might act (ibid.).
Secondly, let us consider the problem of essentialism. For Jane Mansbridge (1999: 637), the tendency towards essentialism, which is the assumption that members of certain groups have an essential identity shared by all members of the group and of which no others can partake, makes up a significant cost to the pursuit of descriptive representation. To say
9
that women should represent women, implies the existence of an essential quality of ‘womanness’ that all women share, an essential quality that men cannot adequately represent (Mansbridge 1999; Phillips 1995; Young 1990; Kymlicka 1995). This essential nature binds all women together, giving them a common interest. However, this essential nature ignores major lines of cleavage within groups and assimilates subordinate interests into those of the dominant group without recognising their existence (Mansbridge 1999: 637). Essentialism is a problem that afflicts every minority group trying to challenge the hegemony of the dominant group. For example, scholars of African matriarchy and black feminism have fiercely criticised the feminist notion of sisterhood for its domination by white middle-class women (Amadiume 1997).
Scholars who try to make oppressed minorities present in politics often mitigate these costs by focussing on non-essentialist reasons for selecting groups for descriptive representation. For Mansbridge (1999: 638), “descriptive representation most closely approaches normative ideals when it reflects the inner diversity of any descriptively denominated group”. Hence, she devotes special attention to historical contexts or characteristics of the electoral system suppressing certain voices in the political domain and how descriptive representation would make these voices heard again in representative assemblies (ibid.). Building on such contemporary theories of descriptive representation, I will develop a new argument that problematises the descriptive underrepresentation of the lower educated.
3. Justice and legitimacy
Following the previous conceptualisation, we must conclude that groups have a claim to enhanced descriptive representation when their members are (1) subject to systematic
10
disadvantage in the political process, or (2) have a claim to self-government (Kymlicka 1995: 145). Since the latter criterium mostly applies to national minorities (the Quebecois in Canada e.g.), I will focus on the former. Unlike national minorities, lower educated people do not share a common cultural consciousness that is distinct from middle or highly educated people for which they can claim a right to self-government.
Systematic disadvantage, however, is a more complicated criterium to expose because it is unclear how it should be measured; many groups claim they are being disadvantaged in some respects while they are simultaneously privileged in others (ibid.). Iris Young operationalises disadvantage with her five faces of oppression: exploitation, marginalisation, powerlessness, cultural imperialism, and violence. Either one of these faces prevents people from learning and using expansive skills in socially recognisable settings like the political domain or expressing their perspectives on social life during political processes of deliberation (Young 1990: 38).
Powerlessness is one face of oppression that applies precisely to the case of the lower educated. In both the political and social sphere, lower educated people lack the authority, status, and a sense of self that the highly educated tend to have (Aarts & Thomassen 2000: 50;
SKON 2017: 44)1. With the establishment of representative democracy, the founders
deliberately introduced the feature that representatives should be socially superior to those who elect them (Manin 1997: 94). Representatives should exceed their voters in wealth, talent, and status. Thus, the idea that lower educated people are unfit to occupy political positions, is embedded in our political culture and systematically prevents the lower educated from learning political skills or expressing their perspectives in political deliberative organs.
1 The most recent study by Dutch Parliamentary Electoral Studies (NKO 2017) states that 72% of the lower
11
Meanwhile, the privileges of the highly educated advance progressively over time (Young 1990: 56). Firstly, the highly educated obtain specialised knowledge through education which grants them the expertise necessary to become a political professional. Thereafter, the highly educated expand their knowledge by putting it into practice in the political domain. Because of their articulateness, people are more likely to listen to the social perspectives of the highly educated while the lower educated remain powerless (ibid.). The comparative study by Verba et al. supports these findings by showing that the extent to which people get involved in politics, and thus gain access to the decision-making table, is directly correlated with the resources they have at their command. For the highly educated this is often a broader network, better literacy skills, and a higher income (Verba et al. 1978).
However, just the fact that lower educated people are systematically disadvantaged in the political realm is not enough to justify affirmative actions. If we solely were to follow Young’s analytical framework in identifying oppressed groups in the US, we would have to consider affirmative actions for around 80% of the population, making a solution to the problem impracticable (Kymlicka 1995: 145). We have to come up with yet another argument why the characteristic of education should be made salient in the academic tradition of descriptive representation. I will deliver such arguments by addressing its importance for a democracy’s legitimacy and substantive representativeness.
Scholars concerned with enhanced representation for women (Phillips 1998), ethnic minorities (Mansbridge 1999; Kymlicka 1995; Williams 2000), or social groups in general (Young 2002) often argue that the absence of certain social groups from political processes lowers a democracy´s legitimacy. When certain social groups have been historically excluded from the political decision-making process, its members are likely to dissociate themselves from it. Indeed, women and ethnic minorities exhibit lower level of political participation than their white male counterparts (Van der Meer & Hakhverdian 2017: 93). Consequently, they
12
become apathetic, lower their trust in the regime, and feel less motivated to take part in it (Young 2002: 144). In short, they lose their allegiance to the regime, which will decline in legitimacy. Conversely, the inclusion of such social groups enhances the de facto legitimacy of political regimes (Mansbridge 1999: 650). The knowledge that one’s interests are being heard, and that one’s descriptive characteristics are no reason for exclusion from the political process, will increase the legitimacy of the democratic process in one’s eyes.
For Phillips (1995: 40) it goes beyond legitimacy: “it is not about stabilising democracies and preventing mass alienation of social groups who might take it to the streets.” There are also psychological goods of fair representation that are attached to what Charles Taylor (1992) calls the ‘politics of recognition’. To see that your descriptive characteristics are no reason to exclude you from the political decision-making process also constitutes a kind of recognition; a public acknowledgement of equal value.
Let us again focus on the Netherlands. Its fully proportional electoral system could not prevent the total exclusion of lower educated people from Parliament (Hakhverdian & Schakel 2017: 53). In the Netherlands, 96% of the parliamentarians are highly educated while this group only makes up 28% of the Dutch population (ibid.). Here, the same delegitimising effects of underrepresentation occur among the lower educated, the very same delegitimising effects scholars address when they argue for more descriptive representation of racial, ethnic, or gender minorities (Bovens 2006: 15).
When the Netherlands changed from a majoritarian to a proportional electoral system in 1917, it also implemented compulsory suffrage to make sure that election outcomes accurately reflected the party preferences of all voters (Aerts et al. 2013: 165). As a result, approximately 95% of the Dutch electorate voted in parliamentary elections. Yet, compulsory suffrage was always contested because people believed it could not be enforced. Moreover, liberals argued that the individual should be free to decide if he (or she after the introduction
13
of female suffrage in 1919) wanted to vote or not (Lijphart 2012: 3). Ultimately, compulsory suffrage was abolished in 1970.
Before then, education was of no significance in predicting voter turn-out. This changed, however, with the abolishment of compulsory suffrage in 1970. Now, education is an important explanatory variable of political participation. Also, consider that voting is just one form of political participation, and a very non-demanding one as well (Hakverdian 2017: 52). For high-demanding types of participation, like attending consultation meetings or other interactive decision-making forums, the explanatory power of education is even stronger. Compared to the lower educated, highly educated people are six times more likely to attend
such meetings (SKON 2018: 62)2. Lower levels of cognitive skills and articulateness put the
lower educated at a disadvantage, preventing them from getting the confidence necessary to engage in politics or express their social perspectives in political or civic debates (Hakhverdian et al 2012).
Following the earlier mentioned scholars on group representation, this fact has detrimental effects on both the de facto legitimacy of the Dutch representative democracy, and on the psychological well-being of the lower educated. Lower educated people dissociate themselves from the political regime because they feel excluded from its decision-making processes. They cannot identify themselves with the representatives taking part (SKON 2017:
44)3. Consequently, this leads to a situation wherein the lower educated do not feel that their
equal political worth is being acknowledged by the representatives.
Not only does disengagement manifest itself through levels of participation, levels of political trust show the same trend. In the Netherlands, education explains variances in
2 In 2017, around 12% of the highly educated had ever participated in such consultation meetings whereas for the
lower educated, this was only 2%.
3 In 2017, more than half of the lower educated declared that their ‘kind of people’ cannot have a significant
impact on political decision-making and that their opinion is irrelevant to representatives. For the highly educated this was 24% and 37% respectively.
14
political trust levels better than gender or ethnicity (Van der Brug 2007: 42). Moreover, when we look at levels of trust in political institutions alone, gender and ethnicity are non-significant determinants of political trust, whereas ‘medium’ educated and especially lower educated people are significantly less trusting. These findings hold in comparative studies that study the levels of political trust in a broader European context (Van der Meer 2017).
To conclude, advocates of enhanced descriptive representation for women and ethnic minorities often argue that these social groups deserve affirmative treatment in politics because they have been historically excluded from the political decision-making table. This historical exclusion structurally disadvantages women and ethnic minorities in politics which prevents them from participating commensurately with their share in the population. Consequently, this results in disengagement from the political regime, lowering its de facto legitimacy.
However, as I have argued, the same logic applies to the lower educated in society. The notion that lower educated people should refrain from politics is embedded in our political culture and puts them at a disadvantage in the political decision-making process. Consequently, the lower educated dissociate themselves from the political regime, as manifested in lower rates of political trust and participation. Both effects are supported by empirical evidence extracted from the Dutch case. Thus, if we want to take descriptive representation seriously, we must not only consider rectifying historical marginalisation and dissociation from the political regime for women and ethnic minorities, but also for the lower educated citizens in our representative democracies.
15
The previous section provided an argument for enhanced descriptive representation of the lower educated by focussing on the assembly’s role of constructing social meaning and legitimacy. This argument, however, failed to discuss the most important function of representative assemblies: representing the electorate’s substantive interests (Mansbridge 1999: 360). For the lower educated, it is most important that their substantive interests are being represented. If it is found that an exclusively highly educated Parliament does a better job in representing these interests than a Parliament including lower educated people, it would be a major argument against the introduction of lower educated people in Parliament.
From an aggregative conception, democracy is regarded as a process of aggregating the preferences of citizens in choosing public officials and policies (Young 2002: 19). In the decision-making process, individuals have varying policy preferences, and political parties and candidates offer a platform to accommodate the largest number of preferences. Citizens with similar preferences can organise themselves into interest groups to affect the actions of elected political parties or candidates. Leaders, parties, and social groups compete over their preferences with other political actors, and as long as this competition is transparent and fair, the outcomes of both elections and parliamentary decisions will reflect the most widely held preferences in society (Dahl 1998).
In the Netherlands, electoral campaigns are publicly funded and electoral seats proportionally distributed. Hence, the threshold of getting elected and the influence of money in electoral competition is relatively low. In this situation we would expect all preferences to be more proportionally reflected in representative assemblies (Gilens 2015: 226). To see whether this claim holds, I will compare survey data on the policy preferences of both national and local representatives with the policy preferences of the Dutch electorate. It will allow us to measure if representatives truly reflect the policy preferences of the population.
16
The NKO (Dutch Parliamentary Electoral Research) measured Dutch MPs’ policy preferences on income distribution, European integration, multiculturalism, criminality, and euthanasia. Dutch scholars Armen Hakhverdian and Wouter Schakel (2017: 38) compared the outcomes of this survey with voter preferences on the same issues. All preferences were measured on a scale from 1 to 7, making the levels of congruence more robust (ibid.). The congruence scores are 75.3%, 75.9%, 78.4%, 54.3%, 57.9% respectively for every issue. A comparison between the mean scores of voters and representatives for each issue shows that voters are more progressive on income equality and euthanasia, and more conservative on European integration, multiculturalism, and criminality (a shown in Table 1) (Schakel & Hakhverdian 2018: 10).
However, since there is no objective scale indicating which scores imply good or bad representation, the numbers alone mean nothing. It only shows that representatives do not perfectly reflect the policy preferences of the electorate. However, this is true for every other representative democracy and it does not tell us how the quality of substantive representation may be enhanced. We can
17
only say something of the quality of congruence scores by relating them to different congruence scores. For this reason, the MPs’ levels of ideological congruence will be compared to those of local representatives (Schakel & Hakhverdian 2018: 17).
The results show higher congruence scores for local representatives: 90.6% vs. 78.5% for income inequality, 83.7% vs. 75.9% for European integration, 83.0% vs. 75.3% for multiculturalism, and 76.9% vs. 54.3% for criminality (euthanasia was not measured on a local level). Therefore, we may conclude that, compared to national representatives, the substantive interests of the Dutch electorate are better reflected by local representatives (Hakhverdian & Schakel 2017: 43). Somehow, local representatives differ from their national counterparts in a way which is beneficial for the quality of substantive representation.
One of the differences that possibly explains the increased congruence scores, is the presence of local parties that are not represented on a national level. Like the Dutch electorate, local parties tend to be more Eurosceptic and in favour of stricter migration policies. However, when we leave local parties out of the analysis, congruence scores of local representatives only drop marginally (ibid.). Thus, there must be another, more important difference between local and national representatives that makes up for the varying congruence scores.
For Hakhverdian and Schakel (2017: 44), the most important difference between local and national representatives is their level of education. Both argue that local representatives reflect the electorate’s preferences better because their level of education approximates more closely to the electorate’s level of education (ibid.). This directly implies that MPs will reflect the electorate’s policy preferences better when their level of education is more similar (drops) to the electorate´s level of education. Although it is likely that a lower level of education will increase the quality of substantive representation from an aggregative ideal (especially since the enhanced congruence scores on a local level cannot be explained by the presence of local
18
parties), it cannot be confirmed empirically. In the Netherlands, similar studies have never
been conducted with relatively lower educated Parliaments.4
To clarify the importance of education for ideological congruence, I will compare the congruence scores for both higher and lower educated citizens using Wouter Schakel and Armen Hakhverdian’s analysis of the same NKO data in 2018. Furthermore, I will disentangle the effects of education from income by presenting congruence scores for all possible combinations of income and education categories for two issues: income inequality and multiculturalism. This will allow me to measure whether policy preferences of local representatives approximate the preferences of the highly educated more closely because their level of education is more similar and if these effects occur independently from the effects of income.
4 Theoretically, one could compare the findings from the Dutch case with ideological congruence scores in
countries where its representatives are relatively lower educated. This, however, goes beyond the scope of this article.
Graph 2.1
19
In graph 2.1 & 2.2, Schakel and Hakhverdian (2018: 15) show that when income is kept constant, the congruence scores of the higher and lower educated differ 13 percentage points for the issue of income inequality and 36.6 percentage points for the issue of multiculturalism. Both differences between the higher and lower educated are statistically significant. These statistics show that the policy preferences of the highly educated are significantly better reflected than the policy preferences of the lower educated, even when controlled for income. Furthermore, when education is kept constant, it becomes clear that income fails to explain variances in congruence scores for salient cultural issues. “Keeping education constant, there is a difference of 5.9 percentage points between low and high incomes, which is also no longer statistically significant” (ibid).
With this data, I also reject the possible objection that we are actually concerned with social class rather than education-level; that education-level is simply a proxy for socioeconomic status. Indeed, from a perspective of historical injustices, suffrage was also
restricted on the basis of financial prerequisites (Manin 1997: 98)5. I argue, however, that the
variable of education deserves to be treated separately because its effects on ideological congruence occur independently from the effects of income. Moreover, it is now clear that income has an insignificant effect on the congruence scores of cultural issues. Thus, if we want to enhance substantive representation from an aggregative ideal by making the lower educated present in representative assemblies, these lower educated people are not necessarily
required to have a low level of income as well.6
Lastly, I will deal with some of the graphs’ major deficiencies that have not been properly dealt with in the analysis. Firstly, the earlier presented graphs do not take other
5 In the UK, formal property qualifications for MPs were established in 1710. The qualifications were different
from and higher than qualifications for electors. Also, in late eighteenth-century France, one had to pay the equivalent of three day’s wages in direct tax to obtain the right to vote.
6 Though clearly, since education and income are strongly correlated, some form of class-based representation
20
control variables into account (like political participation, political knowledge, age, or gender). Schakel and Hakhverdian (2018: 16), however, do control for these variables by dividing them into several categories and calculating the congruence scores of all possible combinations. For cultural issue-stances, we may solely focus on the educational effects since we found that the effects of income are insignificant.
Indeed, they find that the educational effects are weakened by the effects of the given control variables. However, control variables can only explain about one third to a half of the effects found so far:
“For ideological congruence on the issue of income inequality, the effect of education is reduced from 20.4 to 12.9 percentage points by age and income, a reduction of 37%. The educational effect on European integration is lowered by 24% by political knowledge and income. The same two variables have the biggest impact with regard to crime, where they explain 28% of the educational effect. Political participation and age reduce the effect of education by 33% for multiculturalism” (Idem: 17).
This means that the introduction of control variables does not undermine the fundamental claim that policy preferences of the highly educated are significantly better reflected than the policy preferences of the lower educated. In short, even though control variables mitigate the effects of education on ideological congruence, most gaps between higher and lower educated groups remain in double digits.
Another main point of critique is the time-period in which the data is gathered. Unfortunately, the most recent data on policy preferences among national representatives dates from 2006 (Hakverdian 2017: 33). This year was a turning point in Dutch politics because the figurehead of populism changed from earlier assassinated Pim Fortuyn to PVV party-leader Geert Wilders. It is most likely that the electoral gain of the PVV brought the policy preferences of representatives on multiculturalism and European integration closer to
21
the preferences of the population (Van Spanje 2010: 579). However, it is unlikely that an electoral occupation of twenty seats has diminished educational congruence gaps entirely, especially because the rise of Pim Fortuyn forced established political parties to move closer to the policy preferences of the population way before Geert Wilder’s party-leadership (Ibid.). To conclude, from an aggregative conception of democracy, the substantive interests of the highly educated are significantly better reflected than those of the lower educated. This finding holds when tested for the given control variables and is explained by the representatives’ high level of education. Moreover, a comparison between policy preferences of national and local representatives further indicates that education is an explanatory variable of ideological congruence. Local representatives reflect policy preferences of the population better because their level of education is relatively lower and thus more similar to that of the population. Since the presence of local parties cannot explain enhanced congruence scores on a local level this conclusion is all the more likely. Consequently, this would imply that the introduction of lower educated people in Parliament (thus lowering its level of education), will bring the policy preferences of the representatives closer to those of the lower educated and the population as a whole. However, this claim cannot be tested empirically under current conditions.
4.2. Substantive Representation: Deliberation
The aggregative conception of democracy remains contestable, however, presenting many problems to authors who stand for the interests of minorities. Namely, this model does not provide a strong motivational basis for accepting parliamentary decisions as legitimate (Young 2002: 21). If decisions are only justified by reflecting the most widely preferences held, minorities have no reason not to abide by such decisions, no matter how oppressive they
22
are. For this reason, theorists of descriptive representation often deviate from an aggregative conception of democracy and try to enhance the quality of substantive representation from a deliberative one.
In a deliberative democracy, deliberating parties debate on a basis of inclusion and political equality to agree on collective decisions justified by reasons that are mutually acceptable and generally accessible for everyone to understand (Gutmann & Thompson 2004: 10). Thus, the source of legitimacy, which is the collective judgement of the people, follows from a disciplined set of practices defined by the deliberative ideal (Ibid.). Deliberative democracy, therefore, is fundamentally different from an aggregative conception of democracy because policy preferences are not simply given but ought to be justified. Deliberative ideals require that all actors affected by collective decisions should be included in the decision-making process (Young 2002: 23). Not only should people be included, they should be included on equal terms too. Everyone ought to have an equal opportunity to express their interests and perspectives on social life (ibid.).
Thus far it has become clear that the lower educated are excluded from the decision-making table and that their perspectives on social life are not as much appreciated as those of the highly educated. However, perhaps on salient issues the preferences of the lower educated are represented by a broad spectrum of political parties present in Parliamentary debates. This would mean that the presence of the lower educated is no prerequisite for proper substantive representation because the lower educated often have highly educated party members defending an argument on their behalf (Thompson 1976: 91). This fact may even be desirable since highly educated people more often satisfy deliberative norms of articulateness (Williams 2000: 133).
In some situations, however, highly educated MPs insufficiently represent the interests of the lower educated and it is in these situations that the presence of lower educated people
23
enhances the quality of substantive representation from a deliberative ideal. The lower educated always were (and still are) considered unfit for politics. This oppressive relationship does not only result in lower levels of political trust among the lower educated, it simultaneously affects how the highly educated judge their perspectives. In turn, the lower educated might be overlooked or not taken seriously at the decision-making table (Mansbridge 1999: 641). When a social group’s relationship with its representatives is so fiercely damaged by oppression, descriptive representation helps by facilitating vertical communication between the two (ibid.). People sharing the same experiences of belonging to a subordinate group often share bonds of trust based on these experiences (Mansbridge 2000: 99). Enhanced vertical communication between lower educated voters and representatives should consequently increase the quality of substantive representation by making the perspectives of the lower educated present in deliberation.
The goods of making lower educated people present in deliberative processes are found in yet another situation. Even though political parties already cater a to a broad range of policy preferences of the lower educated, these preferences only concern salient issues of which party preferences are well crystallised and made public in party manifestoes. After elections, policy issues will arise that have not been thoroughly discussed during the campaign. For these policy issues, the lower educated have to allow a certain space of autonomy for representatives to reach agreements (Pitkin 1967: 163). It is in this space of autonomy that the quality of substantive representation may depend on a representative’s descriptive characteristics (Philips 1998: 76).
Especially for non-salient issues of which its policy-agreements will heavily impact the well-being of the lower educated, it is important that the lower educated themselves are present during the decision-making process. Being part of the lower educated group in society will bring along a unique set of experiences and perspectives which may contribute
24
significantly to the quality of deliberation in a decision-making process (Williams 2000: 129). Not just because this perspective can only be properly expressed by the lower educated, but
also because highly educated MPs might not be aware of its existence.7 After all, highly
educated representatives lack the experiences necessary to develop this perspective.
5. An unlikely concurrence
As I pointed out earlier, authors concerned with descriptive representation often expose the deficiencies of the aggregative model to pave the way for an argument solely based on the deliberative conception of democracy (Young 2002; Pitkin 1995; Kymlicka 1995). I go beyond these scholars by making the ideal of aggregation salient alongside the ideal of deliberation. Because both ideals have a prima facie conflicting relationship, the underlying assumption that satisfying both ideals is beneficial to the quality of substantive representation needs to be justified. First, I will elaborate on the conflicting relationship between the aggregative and deliberative model of democracy. Thereafter, I will reconcile the two and justify why I use the aggregative model in arguing for the presence of lower educated people in representative assemblies.
For deliberative democrats, the aggregative assumption that representatives ought to reflect the policy preferences of the electorate is false and problematic. In a representative democracy, they argue, representatives are entrusted with an open mandate, so they may form
their own judgements and be convinced by the better argument (Thompson 1976: 112)8.
7 In May 2018 for example, Dutch Minister Van Engelshoven initiated a law that changes the appreciation of
MBO-students (intermediate vocational education). By law, MBO-students were first referred to as ‘participants’, depriving them from privileges regular students enjoy. Van Engelshoven initiated an amendment after non-highly educated people explained how this law marginalises and disadvantages them in every-day life (Trouw 2018).
8 This ideal is still enshrined in the Dutch Constitution: Members of Parliament vote without ‘last’ (translated as
‘burden’) (Article 67:3). Originally, Members of Parliament also ought to vote without ‘ruggespraak’ (without consulting parties or voters). This last part changed after a Constitutional amendment in 1983.
25
When representatives are deprived of their open mandate and solely ought to recite their voters’ preferences, this would be impossible.
Furthermore, entrusting representatives with an open mandate is beneficial to the electorate because the agreements that follow from the deliberative model are superior to those following from the aggregative one. Firstly, the deliberative process is bound to varying communicative and procedural ideals like reason-giving, inclusion, and political equality, making agreements more just and reasonable (Gutmann & Thompson 2004: 7). Secondly, the division of labour representative democracy entails, allows representatives to engage with fact, interests, and perspectives full-time, making them more capable deliberators than voters (Dahl 1989: 215). In short, discrepancies between the preferences of voters and representatives are unproblematic since deliberative norms and the superiority of representatives ensures that agreements will be better than those reached in an aggregative model.
Now we are left with the question why we would still want our representatives to reflect the policy preferences of their voters. I will answer this question as follows: indeed, the aggregative model inadequately describes some of the fundamental aspects of democracy for which the deliberative model constitutes a supplement. However, in some respects the deliberative model suffers from severe deficiencies which require supplementation by goods following from the aggregative model. Both models are thus mutually complementary.
In the deliberative model, the basis of legitimacy for representative assemblies lies in its ability to reach agreements reasonable and acceptable to all. This presumption is fulfilled when decisions follow from an appropriate process of deliberation based on the norms and ideals sketched earlier (Gutmann 2004: 9). However, when applied in the empirical world, enforcing such norms is problematic in two ways: one politically and one normatively.
26
Firstly, the political reality of especially multi-party democracies disseminates the quality of deliberative democracies to reach agreements better than policy agreements reached by merely aggregating policy preferences. Representatives are no longer free to vote according to their own conscience because their open mandate has been appropriated by political parties (Manin 1997: 211). Hence, party discipline precludes representatives from being convinced by the better argument because policy preferences are now constituted on a party-level. These preferences turn rigid once they enter Parliament. Consequently, Parliament has become an instrument of measurement that weights the relative forces of conflicting social interests, and agreements are only reached on the basis of compromise (Idem: 212). In short, the political reality forces the aggregative model to play a more important role by ensuring that Parliament accurately weights the social interests of society.
Secondly, there are several issues that cannot be solved by the force of the better argument. Decisions on such matters can therefore not be objectively better than decisions reached in an aggregative model. This is particularly true for conceptions of the good life (Mouffe 1999: 751). With different conceptions of the good life, the common good is bound to mean different things and therefore cannot be discussed inside the arena of rationality (Schumpeter 1943: 251). When engaging in debates outside this arena, representatives are unable to fully distantiate themselves from their pre-deliberative policy preferences. Thus, for voters’ substantive interests to be fairly well represented in such debates, it matters a great deal whether representatives and voters share the same policy preferences. For these two reasons, I argue that enhanced representation from an ideal of aggregation (alongside enhanced representation from a deliberative ideal) is a good, making the presence of lower educated in representative assemblies more desirable.
27
The introduction of education as a salient characteristic for the theory of descriptive representation resurfaces an important debate, a debate that engenders one of the largest obstacles in making the political presence of the lower educated feasible. Earlier, I concluded that for liberal democracies, gender and ethnicity are arbitrary variables in determining political competence. For education, however, the story is entirely different.
The principle of competence expresses the belief that a democracy should give as much weight as possible to superior intelligence and virtue in the political process (Thompson 1976: 54). Originally, John Stuart Mill employed two kinds of competence that governing minorities should possess: instrumental and moral. Instrumental competence lies with the representative’s ability to look for the best means to certain ends whereas moral competence allows representatives to discern sinister interests and aim at serving the general interest instead (Mill 1861: 107). Democratic elites must utilise these competencies to combat the dangers of the social tyranny of the majority and the influence of interests not identical with the common good.
Relating to the previous chapter, representatives should not only be entrusted with an open mandate because their official position allows them to be concerned with facts full-time but also because they have more instrumental and moral competence. Invoking a Platonic analogy, Mill argued that “a representative should not act according to the judgements of his voters, any more than a representative requires a physician to prescribe for them according to his own notions of medicine” (Duncan 1973: 261). The increasing complexity of our society further increases the risks political decisions entail (Beck 1992). Consequently, the principle of competence is resonated by scholars who point towards these risks using a similar analogy:
28
“In life-and-death medical decisions, what could be more stupid than holding a vote” (Estlund 2008: 62)9.
Some scholars go even further by arguing that suffrage should be restricted to the politically capable only because it is unjust to significantly alter a citizen’s life prospects by the force of decisions made in an incompetent way (Brennan 2011: 704). For these so-called epistocrats, suffrage should be restricted to those who combine extensive knowledge and analytical sophistication with open-mindedness; those who do not let emotion or bias cloud their knowledge (Brennan 2016: 5). These epistocratic virtues, however, are only prescribed to the highly educated, problematising the presence of lower educated in representative assemblies. The same applies to virtues important from a deliberative conception, like proper articulateness and open-mindedness (Young 2002: 38). Yet, while I do not gainsay the dangers of incompetence in political decision-making, I argue that the introduction of lower educated people in representative assemblies will not affect the likelihood that these dangers become reality. To reaffirm, this argument is not aimed towards the competence principle per se, but to the epistocratic convention that the lower educated cannot satisfy its expectations.
Firstly, to say that prescribing policy is the same as prescribing medicine, implies that politics is solely a matter of technical expertise (Philips 1995: 28). These Platonic analogies, invoked by philosophers to justify the notion that representatives should be technical experts, are questionable. Let us focus on an example of Plato himself: We might prefer to see experts navigating the seas, as Plato suggests in his famous “ship of state” analogy. However, managing a ship entails more than just technical knowledge about water and wind currents. In case of a passenger’s ship, important decisions must be made about inclusion: where the ship will dock and who is allowed on board? For such decisions, fundamental disagreements are
9 Here, Estlund is solely referring to the ancient analogy invoked by critics of democracy. Estlund himself is a
29
likely to persist over priorities, principles, and goals (ibid.). Consider, for example, recent debates over whether migrant ships should be allowed to dock at European ports (Wheaton 2018). These issues must be debated not simply within a realm of technical expertise, but within a broader realm of morality.
Yet, this fact does not make technical expertise entirely irrelevant. Technical expertise might still be required in arguing for the reasonableness of certain decisions. When issues outside a realm of rationality still require technical knowledge, a Parliament that includes lower educated people can still utilise this knowledge. However, rather than being on top, experts are now on tap, serving the assembly (Bovens 2006: 71).
Secondly, the conventional notion that highly educated people are more capable of weighing facts and arguments is questioned by several empirical studies. It is found that even highly educated people, including academics, extensively exhibit human traits like accepting and exposing oneself to arguments more often when they reinforce their own predispositions, often described as confirmation bias and selective exposure (Mahony 1977; Oswald & Grosjean 2004). Human fallacies in rational debates are thus not solely prescribed to the lower educated.
I do not deny that the lower educated occupy a disadvantaged position at the decision-making table. Centuries of marginalisation have prevented the lower educated from fulfilling the norm of articulateness necessary to be taken seriously in deliberation (Williams 2000: 133). However, with a proper Parliamentary training, and through participating in deliberative processes, the lower educated should be able to somewhat mitigate these effects. John Stuart Mill, an architect of the competence principle himself, favours a broad scope of political participation due to its educational effects (Thompson 1976: 32). For Mill, participation is associated with the political knowledge we deem important for wise political judgement. Although it does not make a person more educated, active participants are nevertheless more
30
likely to be better informed about politics, better able to perceive differences between parties and citizens, and more likely to have sophisticated opinions (Idem: 40).
Ultimately, by stepping away from the Platonic conception in which formal reason in intrinsic to political judgement, and instead perceiving it as an artful political practice, I challenge the epistocratic notion that only highly educated experts are capable of debating effectively (Barber 1988: 211). It is through participation that the lower educated will acquire the civic competences we associate with wise political judgement, as well as specialised
knowledge we associate with expertise (Ibid.).10 Consequently, the political capability of
representative assemblies will not be affected by the introduction of lower educated people.
7. Conclusion
Contemporary scholars of descriptive representation are often motivated by an ideal of rectifying historical injustices for social groups defined by their gender or ethnicity. Indeed, the long-held belief that women and non-whites are inferior to their white male counterparts negatively affects the political presence of these groups in contemporary times. Both groups exhibit lower levels of political participation and trust, lowering the de facto legitimacy of democratic regimes.
I have argued, however, that the same mechanisms of marginalisation apply to the lower educated in society. Despite a highly proportional electoral system in the Netherlands, the lower educated are entirely excluded from its representative assemblies. Consequently, education is an even larger determining variable of political trust and participation. In short, if we seriously want to enhance descriptive representation from an ideal of justice and de facto
10 Possibly a certain paradox arises: lower educated representatives will become educated by virtue of being in
parliament. In response, I think it is important to emphasise the fundamental differences between institutional education for which one gets a diploma and civic education which one will acquire through participating in politics.
31
legitimacy, the presence of lower educated people in representative assemblies should also be considered.
Furthermore, I have shown that the introduction of lower educated people in representative assemblies would increase the substantive representativeness of Parliament from an ideal of aggregation and deliberation. While acknowledging some fundamental critiques of the aggregative model, I have argued that satisfying its ideals is a good that requires the presence of the lower educated in representative assemblies. This aggregative justification can coexist alongside a deliberative democratic argument for descriptive representation. The aggregative model constitutes a supplement in areas where the deliberative model has insufficiently dealt with reaching agreements based on reasonableness.
The deficiencies of the deliberative model are further exposed in my fundamental critique of the principle of competence. By rejecting the Platonic conception of formal reason as intrinsic to political judgement, I have challenged the epistocratic convention that only the highly educated are capable of reaching wise political judgement. I contend that the lower educated are kept from enacting wise political judgements precisely because they do not have a seat at the decision-making table. By granting the lower educated a seat, I wager that they will acquire the civic competence we associate with wise political judgement, as well as some of the specialised knowledge we associate with expertise. It is for all these reasons that we ought to make the lower educated present in representative assemblies: not only in the Netherlands, but in all advanced Western democracies.
References
Aarts, K. & Thomassen, J. (2000). “Belangstelling voor Politiek en Politiek Zelfvertrouwen”, in: Politieke veranderingen in Nederland 1971-1998: Kiezers en de Smalle Marges van de Politiek, ed. Thomassen, J. Aarts, K. & Van der Kolk, K. (SDU), pp: 35-55.
32
Aarts, R. et al. (2013). Land van Kleine Gebaren: Een Politieke Geschiedenis van Nederland 1780-2012. Amsterdam: Uitgeverij Boom.
Amadiume, I. (1997). Re-inventing Africa: Matriarchy, Religion and Culture. London: Zed Book Ltd.
Barber, B. (1988). The Conquest of Politics: Liberal Philosophy in Democratic Times. Princeton: PUP
Beck, U. (1992). Risk Society: Towards a New Modernity. New Delhi: Sage.
Birch. A.H. (1967). Representation and Responsibility in Politics: An Essay on the British Constitution. London: George Allen & Unwin Ltd.
Bovens, M.A.P. (2006). De Diploma Democratie: Over de Spanning Tussen Meritocatie en Democratie, Bestuur en Beleid, 33(4): 1-27.
Bovens, M. and Wille A.C. (2010). “The education gap in participation and its political consequences”, Acta Politica 45(4): 393–422.
Bovens, M. and Wille A.C. (2017). Diploma Democracy: The Rise of Political Meritocracy. Oxford: OUP.
Brennan, J. (2011). “The Right to A Competent Electorate”, The Philosophical Quarterly, 61(245): 700-724.
Brennan, J. (2016). Against Democracy. Princeton: PUP.
Conti, G. (2018). “Democracy Confronts Diversity: Descriptive Representation in Victorian Britain”, Political Theory, 0(0): 1-28.
Dahl, R. A. (1989). Democracy and It’s Critics. Yale: YUP. Dahl, R. A. (2000). On Democracy. Yale: YUP
Dicey, A.V. (1867). “The Balances of Classes”, in: Essays on Reform, 67-85. London: Clay, Sons, And Taylor
Duncan, G.C. (1973). Marx and Mill: Two Views of Social Conflict and Social Harmony. Cambridge: CUP
Gilens, M. 2015. “Descriptive Representation, Money, and Political Inequality in the United States.” Swiss Political Science Review, 21(2): 222-228.
Gutmann, A. & Thompson, D. (2002). Why Deliberative Democracy? Princeton: PUP
Hakhverdian, A., Van der Brug, W. and De Vries, C. (2012), ‘The emergence of a “diploma democracy”? The political education gap in the Netherlands, 1971-2010’, Acta Politica 47(3): 229–247.
Hakhverdian, A. (2015). “Does it Matter that Most Representatives are Higher Educated?”, Swiss Political Science Review 21(2): 237-245.
Hakhverdian, A. and Schakel, W. (2017). Nep Parlement? Een Pleidooi voor Politiek Hokjesdenken. Amsterdam: AUP.
Kymlicka, W. (1995). Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford: Clarendon Press
Lijphart, A. (1997). “Unequal Participation: Democracy’s Unresolved Dilemma”, The American Political Science Review, 91(1): 1-14.
Madison, J. (1787). “Federalist No.10”, in: The Federalist Papers, ed. Clinton Rossiter (New York: New American Library, 1961), pp: 77-84.
Mahoney, M. J. (1977). "Publication prejudices: an experimental study of confirmatory bias in the peer review system", Cognitive Therapy and Research, 1(2): 161–75.
33
Manin, B. (1997). Principles of Representative Government. Cambridge: CUP.
Mansbridge, J. (1999). “Should Blacks Represent Blacks and Women Represent Women? A Contingent "Yes", The Journal of Politics, 61(3): 628-657.
Mansbridge, J. (2000). “What does a Representative Do? Descriptive Representation in Communicative Settings of Distrust, Uncrystallized Interests, and Historically Denigrated Status”, in: Citizenship in Diverse Societies, ed. Kymlicka, W. and Norman, P. (Oxford University Press), pp: 99-123.
Mill, J. S. (1861). Considerations on Representative Government. London: Parker, Son, and Bourn.
Mouffe, C. (1999). “Deliberative Democracy or Agonistic Pluralism?”, Social Research 66(3): 745-758.
Oswald, M. E. & Grosjean, S. (2004). "Confirmation Bias", in Pohl, Rüdiger F., Cognitive Illusions: A Handbook on Fallacies and Biases in Thinking, Judgement and Memory, Hove, UK: Psychology Press, pp. 79–96.
Phillips, A. (1994). “Democracy and representation: or, why should it matter who our representatives are?”, SVPW Jahrbuch, 34: 63-76.
Phillips, A. (1995). The Politics of Presence. Oxford: Clarendon Press. Phillips, A. (1998). Feminism and Politics. Oxford: OUP.
Pitkin, H. F. (1967). The Concept of Representation. Berkeley: University of California Press. Schakel, W. & Hakhverdian. A. (2018). “Ideological congruence and socio-economic
inequality”, European Political Science Review, 1-25.
Schumpeter, J. A. (1942). Capitalism, Socialism, and Democracy. New York: Harper & Brothers.
Stichting KiezersOnderzoek Nederland (SKON). (2017). Aanhoudend Wisselvallig: National Kiezersonderzoek 2017, Retrieved from: http://www.uva.nl/binaries/content/document s/personalpages/m/e/t.w.g.vandermeer/nl/tabbladdrie/tabblad/drie/cpitem%5B3%5D/a sset?1519485269416
Kleinjan, G. J. (2018). “Mbo'ers zijn officieel studenten. 'Onze opleiding is echt geen
cursusje'”. Trouw, June 1.
Thompson, D. F. (1976). John Stuart Mill and Representative Government. Princeton: PUP. Van der Brug, W. and Van Praag, P. (2007). “Erosion of Political Trust in the Netherlands:
Structural or Temporarily? A Research Note”, Acta Politica, 42: 443-458.
Van der Meer, T. and Hakhverdian, A. (2017). “Political Trust as the Evaluation of Process and Performance: A Cross-National Study of 42 European Countries”, Political Studies, 65(1): 81-102.
Van Spanje, J. (2010). “Contagious parties: anti-immigration parties and their impact on other parties’ immigration stances in contemporary western Europe”, Party Politics 16(5): ` 563–586.
Verba, S. & Nie, N.H. & Kim, J. (1978). Participation and Political Equality: A Seven-Nation Comparison. Chicago: CUP
Wheaton, O. (2018). “Boat with Hundreds of Refugees on Board Arrives at Spanish Port
After Being Turned Away by Other European Countries”. Independent, June 17. hundreds
34
Democracy”, in: Citizenship in Diverse Societies, ed. Kymlicka, W. and Norman, P. (Oxford University Press), pp: 124-147.
Young, I. M. (1990). Justice and the Politics of Difference. Princeton: PUP Young, I. M. (2002). Inclusion and Democracy. Oxford: OUP.
