![Publication Cover]({"type":"image" , "src":"/na101/home/literatum/publisher/tandf/journals/content/rrep20/2018/rrep20.v054.i03/rrep20.v054.i03/20181031-01/rrep20.v054.i03.cover.jpg"})
![Open access]({"type":"image" , "src":"/pb-assets/3rdPartyLogos/accessOA-1452596421933.png"})
![CrossMark Logo]({"type":"image" , "src":"/templates/jsp/images/CROSSMARK_Color_horizontal.svg"})
Abstract
This article provides the necessary tools to advance comparative research studying the substantive representation of ethnic minorities and women. Firstly, I clarify how the various indicators for individual representatives’ and parliaments’ considerateness of the interests of traditionally excluded groups used in earlier (mostly single-country) studies relate to each other and discuss the advantages and drawbacks of different measures for quantitative comparative research. Secondly, the present article introduces new data comprising three indicators for the substantive representation of ethnic minorities.
Introduction
The substantive representation of ethnic minorities and women—thus to what extent representatives and parliaments are considerate of these groups’ political interests—has barely been studied from a comparative perspective (four notable exceptions exist: Bernauer et al. [2015]; Hänni [2017]; Heath et al. [2005]; Schwindt-Bayer [2006]). Such an approach promises valuable insights into questions that are difficult if not impossible to answer for the large and growing number of single-country studies (see, for instance, Bird [2005, 2015]; Celis [2006]; Childs [2001]; Epstein et al. [2007]; Grey [2002]; Griffin and Newman [2007]; Griffin et al. [2012]; Lončar [2016]; Saalfeld and Bischof [2013]; Swers [2014]; Wüst [2014]). For example, we could learn whether or to what extent electoral incentives moderate the motivation of belonging legislators to advocate for their group’s political interests. Or, in which manner the presence of women’s or minority organisations outside parliaments promotes feminist or minority-friendly legislations. A major barrier to cross-country studies is the lack of a scientific debate about different measurement approaches for the substantive representation of traditionally excluded groups. Aiming to enhance comparative scholarly work in the field, this article provides an answer to the following question: How can researchers measure the substantive representation of traditionally excluded groups in cross-country studies?
For comparative research, indicators have to show ‘validity in terms of each social system and reliability across social systems’ (Teune and Przeworski 1970: 107).1 Measures that identify the level of substantive representation of traditionally excluded groups in a specific case and country, however, often alter their meaning or relevance in different institutional and cultural settings. Analyses of legislators, voting behaviour like those identifying differences by gender (see, e.g., Frederick [2009]; Swers [1998]) or ethnicity (see, e.g., Cameron et al. [1996]; Griffin and Newman [2007]) in the United States provide hardly any insights in parliamentary systems characterised by strong party discipline. Or, the policies under study matter only for specific minorities. For instance, the introduction of English-only laws helps to evaluate the representation of Latinos (Preuhs 2005)—an issue that does not matter for Roma in Central and Eastern Europe who care more about welfare policies. In parallel, not all women favour feminist policies such as liberal abortion rights and extensive childcare policies (Celis and Childs 2012). The problem of equivalence of indicators for substantive representation of traditionally excluded groups thus has many angles and the aim of this study is to provide theoretically grounded and systematic guidance in designing comparative research.
I argue that existing indicators for substantive representation of ethnic minorities and women vary along two attributes. Firstly, researchers use different levels of analysis looking at either the micro level—representatives’ behaviour—when asking questions about individual efforts to promote group interests. Or, scholarly works focus on the macro level—the output of the collective action of all legislators—if interested in the overall considerateness of parliaments of group interests. The room for manoeuvre for single legislators is subject to the restrictions set by the specific institutional context, making the micro level of substantive representation challenging (but not impossible) to study in cross-country comparisons. Secondly, researchers use different means to determine group interests: presuming to know or identifying them empirically. The first strategy risks essentialising group members, even though comparing representatives’ and parliaments’ activities in promoting feminist policies or minority language rights provides easily accessible shortcuts to group preferences. In contrast, empirical assessments of women’s and ethnic minorities’ interests enhance comparability across cases. Positioning indicators explicitly with regard to the level of analysis and the means to determine group interests reveals their equivalence in different country and group contexts and raises awareness for the implicit assumptions they require.
The article at hand further applies these insights by presenting new data for three comparative measures for ethnic minorities’ substantive representation: (1) the introduction of new minority rights, (2) policy congruence between minority citizens and parliaments, and (3) membership in minority committees.2 These indicators are considerate of differences in the interests of various minorities because they either rely on survey data and expert interviews, or focus on representatives’ active engagement in institutions promoting minority representation rather than particular policies. In this manner, the meaning of the three indicators is independent of the contextual setting. The data thus constitute a valuable resource supporting the emergence of the new research agenda in the field.
Existing Measures for Substantive Representation of Traditionally Excluded Groups
Researchers studying the substantive representation of ethnic minorities and women make use of the same set of indicators. Measurement approaches include representative claim analysis, self-reported preferences and activities based on surveys, committee membership, positions in committees, bill sponsorship, voting behaviour, or content analyses of parliamentary speeches, questions, and websites, laws passed, the size of public spending in certain sectors, or ideological congruence.3 While neither of these indicators captures the meaning of substantive representation in a comprehensive manner, they are proxy variables for different facets of the concept depending on the level of analysis and the mode used to identify minority interests.
The Level of Analysis: Individual Representatives or Whole Parliaments
The meaning of substantive representation might refer to two levels of analysis: the connection between the represented and single legislators—the micro level—or the relationship between citizens and whole legislatures—the macro level. Scholars thus either compare the preferences of ethnic minorities or women with individual legislators’ engagement into relevant policy areas or parliaments’ thoughtfulness of respective topics. This distinction differs from Franceschet and Piscopo (2008), who argue that women’s substantive representation consists of the inclusion of women’s interests in the legislative process and policy outcomes. While their typology focuses on the stages along the legislative process, the level of aggregation varies as well and, in my view, is decisive for the selection of comparative indicators.
At the micro level, behavioural and self-reported measures coexist (Celis et al. 2008: 102; Lovenduski and Norris 2003: 86). The former include all actions taken by representatives with the objective to promote group interests such as membership in relevant committees (Bratton 2005; Donovan 2012; Haynie 2001; Saalfeld and Kyriakopoulou 2011; Wüst 2011) and leadership positions in these institutions (Bratton 2005; Donovan 2012; Heath et al. 2005). Other approaches analyse the number of bills sponsored by individual legislators that serve the well-being of ethnic minorities or women (Baker and Cook 2005; Barnes 2012; MacDonald and O’Brien 2011; Wallace 2014). A large set of study emerges around the voting behaviour of office-holders, asking whether representatives tend to support or reject policy proposals in the interest of traditionally excluded groups (Baker and Cook 2005; Cameron et al. 1996; Hogan 2008; Lublin 1997; Saalfeld and Kyriakopoulou 2011; Swers 1998; Wallace 2014).
Self-reported measures of substantive representation evaluate all types of information provided by representatives about the citizens they aim to represent. For that purpose, researchers narrow down the issues that representatives deem important through quantitative content analysis of parliamentary speeches and questions (Baker and Cook 2005; Bird 2005, 2011; Celis 2006; Crisp et al. 2018; Donovan 2012; Grey 2002; Saalfeld and Bischof 2013; Saalfeld and Kyriakopoulou 2011; Wüst 2014) or websites (Saalfeld and Kyriakopoulou 2011). Another type of self-reported measure is representative claim analysis (Saward 2006), which looks into claims to promote minority or women’s interests brought forth by individual representatives again as part of parliamentary speeches and questions (Bird 2005, 2015; Hirschmann 2014; Lončar 2016), but also in interviews or surveys (Barrett 1995; Brown 2014; Fraga et al. 2007; Gidengil et al. 2003).
Both micro-level approaches to measuring substantive representation face a major hurdle to their application in a comparative setting: through which means representatives can act as advocates of traditionally excluded groups depends on the options and limitations laid out by the institutional design. Whether representatives vote for certain legislative drafts serves only as a meaningful measure of substantive representation in government systems with low party discipline. Analysing bill sponsorship implies that office-holders have the right to submit policy drafts independently rather than in groups. Since parliaments function differently across countries, indicators that inform about the efforts of members of parliament (MPs) to promote the interests of traditionally excluded groups in one country are often not transferable to other institutional settings.
At the macro level, substantive representation refers to the relationship of ethnic minorities and women to the performances of whole parliaments. To what extent are parliamentary debates and outputs considerate of their interests? Measures include the number of bills enhancing the well-being of traditionally excluded groups passed by parliaments (Bratton and Haynie 1999; Epstein et al. 2007; Franceschet and Piscopo 2008; Htun et al. 2013; Preuhs 2005; Reingold and Smith 2012) or the size of welfare, education, and health spending (Chattopadhyay and Duflo 2004; Owens 2005; Preuhs 2007). Further studies investigate the extent to which parliaments mirror the ideological preferences of women—so-called congruence (Giger et al. 2012; Ruedin 2013: 79–92).
In contrast to micro-level measures, those at the macro level carry the same meaning across different institutional contexts. The objective to represent all citizens and to create policy outputs does not depend on the available decision-making mechanisms (even though the efficiency certainly does). The latter level of substantive representation is therefore easier to access comparative research.
Means to Identify Group Interests: Assumed or Observed Group Interests
Identifying the interests of ethnic minorities and women is a prerequisite to study policy responsiveness and can be achieved through two strategies: pre-defining the interests of group members or measuring them empirically. In the first case, researchers assume that groups of citizens who share an identity also display similar policy preferences—at least concerning some issues. Many studies focusing on the United States, for instance, rely on the assumption that increases in welfare spending are an important dimension of African-American preferences and use higher or lower percentages of money flowing into the respective sections of state budgets as a measure of substantive representation (Owens 2005; Preuhs 2007). Traditional feminist policies that enhance women’s freedom and equality are policy outputs that researchers deem to be in the interest of women (Bratton 2005; Franceschet and Piscopo 2008; Htun et al. 2013; Smith 2014). A related strategy defines salient topics for traditionally excluded groups and analyses the number of relevant mentions in parliamentary questions and debates. For instance, immigrant minorities should care about immigration and multiculturalism policies (Donovan 2012; Saalfeld and Bischof 2013; Saalfeld and Kyriakopoulou 2011; Wüst 2014).4
This approach suffers from one major flaw: the policy preferences of women and minority members may differ systematically from men and the majority, but not necessarily in the same manner for all group members (Phillips 1995: 105). Policy preferences are diverse even within minorities. While members of traditionally excluded groups share common perspectives on some issues because they made similar life experiences, this does not necessarily translate into equal political interests (Mansbridge 1999; Modood 2000; Young 2000: 88). This problem is especially apparent looking at feminist policies such as abortion rights or childcare policies, which sizeable numbers of women reject (Celis et al. 2008; Celis and Childs 2012).
For comparative research, presumed policy preferences pose additional problems, since salient topics and relevant policies vary across groups and countries (Norris 2004: 220). The exact same activity or legislation does not involve an equal level of substantive representation in two cases. The source of the shared identity of ethnic groups determines relevant policies, e.g., language rights enhance the well-being of linguistic minorities only while religious minorities need religious rights. In addition, contextual settings also change the meaning of policies, so that for instance regulations concerning the headscarf have very different meanings for women in Iraq compared to France. It is difficult (if not impossible) to identify a substantial topic towards which ethnic minority members or women in different countries hold comparable policy preferences and which is equally salient for all groups.5
The second strategy uses empirical data to determine the interests of traditionally excluded groups. Researchers might look at actual preference distributions (Bernauer et al. 2015; Griffin et al. 2012) and—given there is no single interest of all group members—often equate the opinion of the majority of the minority with the group’s interest. For instance, studies make use of the mostly leftist policy positions of African-American citizens in the United States (Cobb and Jenkins 2001; Grose 2005). An inductive, empirical approach to identifying group interests entails comparing attitudes and working foci of majority and minority legislators, assuming that group members in parliament necessarily promote group interests. To give an example: if female legislators are more likely to speak about feminist policies than their male colleagues, this topic must be of interest to women (Barnes 2012; Celis 2006; Grey 2002).
The empirical approach is not without assumptions either because preferences have to be visible, exogenous, and stable. Opacity characterises citizens’ policy positions, given that people in general struggle to communicate their preferences and to demand concrete policies (Eulau and Karps 1977). At the very least, citizens have to engage in an active process of expressing their interests to avoid miscommunication (Plotke 1997: 30–2). Nevertheless, the interests of traditionally excluded groups often remain invisible to MPs. According to Mansbridge (2003: 518), representatives not only reproduce given preferences of constituents, but take part in creating them. Either way, legislators aim to demonstrate that they are considerate of their electorates’ interests.
In a comparative research design, identifying the policy preferences of women and ethnic minorities empirically reduces the risk to study topics that are of little real-life relevance for certain groups. Yet, inductive approaches focusing on differences in representatives’ attitudes and legislative behaviour only allow answering questions at the micro level of analysis. And the survey data scholars use to determine the policy preferences of citizens empirically is not available for all ethnic minorities (partly due to the low numbers of co-ethnics in representative samples and partly because it is prohibited (for researchers in general or publicly funded agencies in particular) to ask survey questions about ethnic identities in some countries including inter alia France and Luxembourg).
Three Comparative Measures for Ethnic Minorities’ Substantive Representation
The subsequent section presents three measures for the substantive representation of ethnic minorities in developed democracies.6 I describe how each indicator overcomes the obstacles outlined above and in which manner they promise insights into different facets of substantive representation. Even though the data cover only ethnic minorities, each section contains a clarification whether and how these indicators might be applied in the context of women’s representation. Given that there is only a single comparative analysis investigating to what extent laws mirror minority preferences (Hänni 2017), the present study satisfies the urge for comparative data.
Substantive Representation as the Introduction of New Minority Rights
A first measure for the substantive representation of ethnic minorities in a comparative setting is the adoption of new rights that benefit minorities according to experts. This approach follows the tradition of studying minority-supported legislation (see, e.g., Bratton and Haynie [1999]; Epstein and O’Halloran [1999]; Epstein et al. [2007]; Preuhs [2005]). It enables researchers to evaluate for instance how increases in descriptive representation or the seat share of ethnic parties impact the level of substantive representation.
The indicator is based on the World Directory of Minorities and Indigenous People (Minority Rights Group International [2013]), which provides expert-based information on the introduction of new laws in the interest of a group. Rather than identifying comprehensive lists of all relevant policies, specialists provide information on change in important legislation in a clearly defined time horizon. This ensures contextual sensitivity because the experts know which laws profit minorities in their area of expertise. In this manner, comparisons across linguistic, religious, and national minorities are possible.
Substantially, the directory contains texts with details about minorities’ characteristics and histories as well as updates on latest changes. In the section on recent developments, the experts describe whether there were important new laws introduced that improved the status of the group since the latest version of the directory. Making use of this information, I identify groups that profited from at least one new right between 1997 and 2012.7 This binary measure informs about the willingness of the majority of legislators to enhance minority well-being. Since groups included in the directory face discrimination according to the experts, all of them have policy requests that legislators have to consider and analysing the introduction of new laws reveals whether parliaments realised them. This simplification, however, also implies that the variable neither provides information on the quality of these legislative changes nor on improvement it entails given the status quo.
The directory lists detailed information on 106 disadvantaged minorities in the member countries of the European Union and the OECD. A total of 49 or 46.23% of these groups did profit from new laws in the period under study. As a validity check, I compare this information to the Multiculturalism Policy Index (MPI) from Banting and Kymlicka (2012) in Figure 1. The index assesses the degree to which a country achieves a pre-defined set of policies generally assumed to be in the interest of minorities. The targeted policies range from multicultural education over the minority language’s official status to the realisation of indigenous land claims.8 To identify recent changes in minority-supported policies, I calculate the difference between the level of minority protection in 2000 and 2010. Even though this information refers to the country, it is possible to identify the minorities it relates to through the evidence section at the MPI website, leaving me with data for 47 minorities covered by the MPI and my data.
Published online:
07 August 2018FIGURE 1 Box plots comparing the new rights indicator with changes in the MPI between 2000 and 2010
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/rrep20/2018/rrep20.v054.i03/00344893.2018.1504112/20190306/images/medium/rrep_a_1504112_f0001_ob.jpg"})
FIGURE 1 Box plots comparing the new rights indicator with changes in the MPI between 2000 and 2010
Figure 1 displays separate box plots for change in the MPI if the dummy variable indicates (1) no new rights or (2) new rights. It shows a higher median increase of Banting and Kymlicka’s (2012) measure for minorities that received new rights than for groups without new legislation in their interest according to the directory. Beyond that correlation, the new rights indicator is more sensitive to the group-specific context as a closer look at the outliers reveals. To start with, although Muslims in the Netherlands experience a decrease in the MPI, they were increasingly protected by anti-discrimination measures. Second, the MPI of Basques and Galician’s in Spain rose from 4.5 to 6 points even though the description in the World Directory of Minorities and Indigenous People does not mention the introduction of any new rights. The adoption of the statutes of autonomy by the central government leads to formal improvements for national minorities in this case, but the implementation remained vague. The MPI is thus sensitive to adjustments in legislations which the experts of the directory did not deem significant or effective.9
By showing the quantity of minorities who were granted new rights depending on the electoral system type, Figure 2 sheds light on the insights we might gain from studying this indicator. Parliaments elected through systems with single-member districts most successfully initiate new laws profiting ethnic minorities. This seems surprising given that majoritarian electoral systems create barriers for minority members and minority ethnic parties. Contrary to the common wisdom, the increased seat shares for minority-friendly actors in proportional electoral systems (Phillips 1995; Ruedin 2013) do not translate into the passage of relevant legislation. Powerful positions for minority legislators as well as the willingness of the majority to support relevant policy drafts are hence more important for creating policy outputs in the interest of minorities than the mere presence of minority members and parties during decision-making (Blalock 1967; Heath et al. 2005; Yoder 1991). It is important to note that this insight only relates to change in legislation, but not to the overall level of protection of ethnic minorities, which might still be higher in countries with proportional electoral systems compared to majoritarian ones.
Published online:
07 August 2018FIGURE 2 Proportion of minorities receiving new rights according to district magnitude in the country
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/rrep20/2018/rrep20.v054.i03/00344893.2018.1504112/20190306/images/medium/rrep_a_1504112_f0002_ob.jpg"})
FIGURE 2 Proportion of minorities receiving new rights according to district magnitude in the country
Applying a similar measure to women’s representation is feasible but all the more challenging. Often, women’s interests vary along social cleavages, making it difficult to disentangle whether legislation serves women’s well-being or the overall societal fragment. Identifying all types of legislative changes that are beneficial for women is therefore challenging and, to my best knowledge, no single source claims to fulfil this endeavour.
Substantive Representation as Policy Congruence
A second approach to substantive representation is identifying the level of correspondence between minority citizens ‘policy preferences and parliaments’ policy orientations. In a nutshell, measuring policy congruence means comparing the preference distribution of citizens as brought forth in population surveys to the policy positions of all representatives. This approach stands in the tradition of studies inquiring policy responsiveness of all citizens and scholars also made use of it to study women and the poor (see, e.g., Blais and Bodet [2006]; Giger et al. [2012]; Golder and Stramski [2010]; Powell and Vanberg [2000]). With some adaptions, this indicator can be applied to study women’s substantive representation as a recent study by Dingler et al. (2018) shows. Even though this indicator does not measure action ex-post, it investigates the policy agenda that parties communicate to voters and thus studies credible commitments to promote group interests. Analysing to what degree MPs mirror the policy preferences of minority citizens (relative to all other citizens), makes it possible to study the conditions under which parliaments become more or less considerate of the interests of minorities.
One strength of this measure derives from the empirical identification of minority interests through information provided by group members in surveys. Additionally, using the concept of ‘many-to-many congruence’ (Golder and Stramski 2010) allows for diverse opinions within the minority rather than making the assumption that all minority members favour the same policies. Covering a broad set of policy areas can ensure that the included issues are salient for minorities. After all, accurate representation of policy preferences only serves the minority, if group members care about the topic at hand. Although many studies on policy congruence of citizens focus on simple left–right placements, it is possible to take into account different policy fields (see, e.g., Stecker and Tausendpfund [2016]) such as immigration, multiculturalism, or redistribution policies.
While survey data are scarce for most minorities, some exemptions exist. These include historic minorities such as African Americans and Latinos in the United States (in the American National Election Study by Brader et al. [2017]) or Maori in New Zealand (in the New Zealand National Election Study by Vowler et al. [2011]), as well as most minorities of immigrant origin.
For example, comparativists might estimate policy congruence between representatives and citizens with and without immigrant backgrounds making use of the European Social Survey (2002–2014) and the Chapel Hill Expert Study (Bakker et al. 2015).10, 11 The population survey asks respondents for instance whether they think their ‘country’s cultural life (is) undermined or enriched by immigrants’. As part of the expert survey, scientists with country knowledge place all parties with regard to their ‘position on integration of immigrants and asylum seekers’ from ‘strongly favours multiculturalism’ to ‘strongly favours assimilation’. By multiplying each party position with its seat share in parliament, a frequency distribution of the policy orientations in the legislature emerges, which can be compared to the preferences of citizens with and without immigrant backgrounds (see Figure 3 for an example). The larger the gap between parliament and groups of citizens, the less accurately do representatives mirror the policy preferences of the respective sub-group. Calculating the area between frequency distributions provides a measure for the size of the deviation and allows for comparison across countries.
Published online:
07 August 2018FIGURE 3 Frequency distribution of preferences towards multiculturalism of representatives and citizens with and without immigrant background in Austria (2002–2016). Annotations: Own calculations based on data from the European Social Survey (2002–2014) and the Chapel Hill Expert Study (Bakker et al. 2015)
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/rrep20/2018/rrep20.v054.i03/00344893.2018.1504112/20190306/images/medium/rrep_a_1504112_f0003_ob.jpg"})
FIGURE 3 Frequency distribution of preferences towards multiculturalism of representatives and citizens with and without immigrant background in Austria (2002–2016). Annotations: Own calculations based on data from the European Social Survey (2002–2014) and the Chapel Hill Expert Study (Bakker et al. 2015)
Figure 4 compares the congruence (or rather deviation) measures for citizens with and without immigrant backgrounds concerning multiculturalism and redistribution in Europe. Larger values indicate more deviation between citizens and elected officials. Accordingly, the interests of citizens of foreign descent with regard to multiculturalism are represented more accurately by parliaments than those of the rest of the population. When it comes to this highly salient topic for immigrants, parliaments are considerate of their needs. By contrast, legislatures mirror the preferences concerning redistribution of both groups of citizens equally well.
Published online:
07 August 2018FIGURE 4 Scatter plot of many-to-many congruence of citizens with and without immigrant backgrounds for the policy areas multiculturalism and redistribution
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/rrep20/2018/rrep20.v054.i03/00344893.2018.1504112/20190306/images/medium/rrep_a_1504112_f0004_ob.jpg"})
FIGURE 4 Scatter plot of many-to-many congruence of citizens with and without immigrant backgrounds for the policy areas multiculturalism and redistribution
Substantive Representation as Membership in Minority Committees
My last suggestion for measuring substantive representation across countries is to look at membership in minority committees. This indicator follows the tradition of scholars studying the question as to which representatives join education, culture, or welfare committees (Bratton 2006; Donovan 2012; Haynie 2001; Saalfeld and Kyriakopoulou 2011; Wüst 2011). However, while committees’ substantial competences are subject to considerable variation across countries for these broader topics, minority committees always focus on a single policy area: minority politics. Analysing membership in these institutions thus allows uncovering the type of MPs who make minority representation the focal point of their parliamentary activity.
Parliaments of all developed Western democracies organise committees in which experts and interested MPs work and prepare policy drafts before they are submitted to the plenum (Longley and Davidson 1998; Strøm 1997). Generally speaking, committee membership indicates that representatives spent a large amount of their time working on issues related to the competences of the institution. Legislators aim to join committees according to their personal interests, since membership means they will spend most time working on related issues. Those who participate in minority committees thus care about minority representation.
Table 1 lists all minority committees in Western-developed democracies. These institutions conduct a broad set of tasks and obligations that—to some extent—vary by country. The Croatian Committee on Human and National Minority Rights, for instance, debated the proposed amendments to the ‘Law on the Use of Languages and Script of National Minorities in Croatia’, but also the state budget for the subsequent years (Croatian Parliament 2013, 2015). To give another example, the Maori Affairs Committee in New Zealand reviewed the financial situation of the Maori Broadcasting Funding Agency and discussed the settlement of Maori land claims (New Zealand Parliament 2012, 2013). In sum, such institutions possess four competences:
financial oversight over funding of minority cultural affairs,
introducing, discussing, proposing the adaption of all kinds of legislation relevant for ethnic minority groups,
the right to discuss and bring forth issues of matter for ethnic minorities,
or maintaining relationships with international bodies and institutions in other countries that aim to promote group interests.
Published online:
07 August 2018TABLE 1 List of minority committees by country
All kinds of issues that matter for minorities and minorities within minorities can be discussed and presented in such institutions. Legislators can set relevant topics on the agenda depending on the group-specific context. In this manner, studying membership in minority committees as an indicator of minorities’ substantive representation allows for inter- and intra-group differences in policy preferences.
While membership indicates that representatives spent a considerable amount of time working on issues of high relevance for ethnic minorities, it lies beyond the scope of this measure to identify how intensively legislators engage in these committees. One might argue that MPs join such institutions for symbolic reasons and barely participate actively. Yet, this still implies that representatives care enough to pretend to promote minority interests. While we cannot separate strong engagement and attempts to appear as advocates of minority interests, this measure informs about how legislators present themselves.
Membership in minority committees is equally relevant for the substantive representation of minorities in all countries that have such an institution in the first place. While not all issues are handled in this arena, they capture a large amount of topics that matter for the groups. Of course, native language teaching for linguistic minorities could as well be on the agenda of education committees, while the financial benefits for socially discriminated groups such as Roma are also on the agenda of welfare committees (for such an approach, see, e.g., Bratton [2006]; Wüst [2011]). Compared to minority committees, considering such a broad set of committees nevertheless bears two disadvantages: they are difficult to compare across a diverse set of minorities and they involve much time spent on issues that are irrelevant for minorities.
Of course, MPs are not entirely at liberty to select any committee: seats are often distributed according to party size. Legislators have to coordinate membership with party leaders and more senior actors are served first. Despite these constraints, Strøm (1997: 40) argues that representatives in all Western democracies enjoy far-reaching autonomy in selecting committees. At the end, they usually join institutions working on topics of their choice. What is more, the most intense competition over seats usually takes place in committees handling policy areas that receive publicity such as welfare or foreign policy. Minority politics is less prestigious and membership in respective committees is easier to achieve. In sum, it is therefore reasonable that representatives join minority committees by choice and as a consequence of their own desire to work on this topic.
In the sample of Western-developed democracies, 11 countries from Europe and other continents, without and with post-communist traditions, with very high or rather low levels of economic development, have minority committees. A total of 226 of 3501 representatives or 6.46% join these institutions. A first view at this indicator in Figure 5 reveals that representatives belonging to ethnic parties are considerably more likely to hold office in minority committees (21% compared to 6%). It does thus not lack efforts of minority party members to enhance group well-being—even though the first indicator revealed that they might barely succeed in creating new legislation.
Published online:
07 August 2018FIGURE 5 Proportions of MPs belonging to minority committee according to minority party membership
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/rrep20/2018/rrep20.v054.i03/00344893.2018.1504112/20190306/images/medium/rrep_a_1504112_f0005_ob.jpg"})
FIGURE 5 Proportions of MPs belonging to minority committee according to minority party membership
This indicator might as well be applied to study women’s representation by looking into membership in gender-specific committees such as the equal treatment committee in Austria or the gender equality committee in Denmark. While earlier studies already revealed which substantial committees women tend to join more frequently (Heath et al. 2005), focusing on the characteristics of representatives who become members of institutions specifically devoted to improving the situation of women might allow addressing an interesting set of comparative research questions: Which type of men join these institutions? Or how do contextual factors such as the electoral system type moderate incentives to join women’s committees?
Conclusion
This article enriched the toolbox of research designs available to scholars studying the substantive representation of traditionally excluded groups. I showed that the numerous approaches to measuring representatives’ efforts to promote the interests of ethnic minorities and women vary along two attributes: the level of analysis and the means to determine group interests. The micro level of analysis—thus legislators’ advocacy for groups’ interests—is particularly challenging to study from a comparative perspective, since the institutional settings must provide legislators with an equal room for manoeuvre in different countries. By contrast, the macro level of analysis focuses on the policy positions of parliamentary majorities and to what extent they mirror group interests—a factor that is not limited by institutions. The definition of group interests is in either case a challenging task. Empirical means to identify the policy preferences of traditionally excluded groups are preferable in a cross-country comparison, since pre-defining the objectives of ethnic minorities and women risks essentialising groups and group members. The choice of indicators selected by researchers thus influences to which nuances of substantive representation their study speaks, as well as the assumptions they have to make in order to ensure equivalence across countries.
In this manner, the conceptualisation of different approaches to measuring substantive representation presented in this enables two disjoint sets of literature to speak to each other. Research studying the representation of ethnic minorities on the one hand and women on the other hand tended to remain unconnected so far. Yet, when building on the knowledge from previous research, studies applying the same means to identify group interests might provide the most insightful grounds to formulate causal mechanisms, independent of the group under study.
In this study, I further provided data that allow studying ethnic minorities’ substantive representation in Western-developed democracies: the introduction of new minority rights, policy congruence between citizens of immigrant origin and parliaments, and minority committee membership of representatives. All three measures capture actual interests of the groups under study by identifying the preferences of minorities empirically. For that purpose, I make use of expert rankings and survey data, or a broader focus on the institutions promoting minority representation rather than specific policies. The substantial meanings of my three measures cover different facets of minorities’ substantive representation. Membership in minority committees, shows whether representatives invest a lot of time working on topics related to ethnic minorities. It exposes individual efforts to enhance minority interests. Policy congruence reveals information about the degree to which legislatures mirror the preferences of minority citizens, while the new minority rights indicator measures the capacity of parliaments to implement new minority-friendly legislation. The three measures hence enable researchers to investigate various aspects of responsiveness to minority interests and findings might vary depending on the applied indicator.
| Country | Period | Minority committees |
|---|---|---|
| Croatia | 2011 | Committee on Human and National Minority Rights |
| Denmark | 2011 | The Faroe Islands Committee; The Greenland Committee |
| Japan | 2012 | Okinawa and Northern Problems Committee |
| Mexico | 2012 | Indigenous Affairs Committee |
| New Zealand | 2011 | Maori Affairs Committee |
| Poland | 2011 | Commission on National and Ethnic Minorities |
| Romania | 2012 | Committee for Human Rights, Cults and National Minorities Issues |
| Slovakia | 2012 | Committee on Human Rights and Minorities |
| Slovenia | 2011 | Commission for National Communities |
| United Kingdom | 2010 | Northern Ireland Affairs Committee; Scottish Affairs Committee; Welsh Affairs Committee |
codebook_indicator_1.docx
Download MS Word (31 KB)ACKNOWLEDGEMENTS
The author is very grateful to Jessica Fortin-Rittberger, Zoe Lefkofridi, Sarah C. Dingler, and two anonymous reviewers for their helpful comments on her research and insightful discussions.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the author.
SUPPLEMENTAL DATA AND RESEARCH MATERIALS
Supplemental data for this article can be accessed on the Taylor & Francis website, https://doi.org/10.1080/00344893.2018.1504112, or via Harvard Dataverse, https://doi.org/10.7910/DVN/K1NUWL.
Notes
1 For a summary of the problem of equivalence in comparative research, see for example, Adcock and Collier (2001: 534–6).
2 All data presented in this article will be available online after publication via Harvard Dataverse, https://doi.org/10.7910/DVN/K1NUWL.
3 I do not consider measures of perceived responsiveness as introduced by Banducci et al. (2004) or Hodžić and Mraović (2015), since group members tend to overestimate the level of substantive representation as a consequence of descriptive representation (Banducci et al. 2004).
4 For a critical assessment of this approach, see Aydemir and Vliegenthart (2016). The ‘Pathways to Power Project’ (Cinalli et al. 2016) currently aims to provide a comparative data set analysing substantive representation as parliamentary questions for eight European democracies.
5 The Multiculturalism Policy Index (MPI) (Banting and Kymlicka 2012), the Migrant Integration Policy Index (Huddleston et al. 2015), and the Gender Inequality Index (United Nations Development Programme 2018) apply such an approach, just like the only comparative study on ethnic minorities (Hänni 2017).
6 Even though it is possible to gather data for a broader set of countries, I chose to limit the focus of this study to enhance comparability. All countries share a solid level of democratic experience and sufficient economic resources to promote minorities’ well-being.
7 A complete list of groups covered and details of the coding process are available in the codebook.
8 Banting and Kymlicka (2012) provide separate indices for immigrant and national minorities and indigenous people, given that not every policy applies to each type of minority, and further aggregate the information for each of the three types of minorities at the country level. The indices range from zero (no minority protection) to (a) six (perfect minority protection) for national minorities, (b) eight for immigrant minorities and (c) nine for indigenous people. I re-scaled all indices on a scale from zero to six.
9 Since the directory is provided by a minority-friendly organisation, the binary measure most likely underreports the amount of legislative changes, but there is no reason to expect systematic bias across countries.
10 The author applied this approach in a recent publication (Kroeber 2017).
11 A complete list of the countries and years covered are available in the codebook, just like a list of additional policy areas.