Recruitment and Representation of Ethnic Minorities under Proportional Representation

Data and Measurement

The scholarship on legislative representation of ethnic minorities has few extensive and reliable databases at its disposal. Moser highlights this problem as a serious limitation to advancing the research agenda in this particular area. ¹⁰ Due to data limitations, many theoretical propositions regarding the factors that influence minority representation tend not to be directly tested or are examined through the use of questionable proxy measures such as the proportion of women elected to the legislature or the electoral success of ethnic parties.

The Bulgarian dataset we have assembled is based on individual-level ethnic, social, and political background data for all the deputies elected into the Bulgarian parliament during the past six consecutive parliamentary terms. The dataset includes observations on both deputies that served a full parliamentary term and those that served a part of the term. The dataset thus includes all deputies that entered the parliament since 1990 up to the end of the 2005-2009 parliamentary term. The dataset has 1,768 observations, where the unit of observation is a deputy/parliamentary term.

The coding of data was based primarily on information that was self-reported by the deputies and published in the official publications of the Bulgarian parliament. ¹¹ This data was supplemented by information from a scholarly work and other published sources produced by a number of commercial and nongovernmental organizations. ¹² The information on the ethnic affiliation of deputies was compiled in cooperation with Bulgarian specialized institutions on minority issues, whose experts were recruited to help ensure the accuracy of the ethnic affiliation data. ¹³

Proportionality of Ethnic Representation

There is a considerable degree of variation in how successful different minority groups are in securing legislative representation in Bulgaria. Although the analysis of data on the ethnic composition of the entire corps of parliamentary deputies points to a significant presence of minorities in the legislature, this presence was much less than proportional in the case of large minority groups. Both of the two largest ethnic minorities, Turks and Roma, were underrepresented in the national parliament. Similar to other East European cases, the Roma community in Bulgaria has experienced the highest degree of underrepresentation. Ethnic Bulgarians, the country’s majority group, enjoyed representation on a level that is slightly higher than proportional.

Data on ethnic composition of the Bulgarian parliament is summarized in Table 1. The table lists population and parliamentary shares of all minority groups represented in the parliament and provides frequency information on a number of deputies of each ethnic background. The last column gives scores for the proportionality of representation index, which is calculated by dividing an ethnic group’s proportion in the parliament by its proportion in the population. This provides a single summary figure where 1.0 symbolizes “perfect” proportional representation, more than 1.0 designates a degree of “overrepresentation” and less than 1.0 indicates “underrepresentation”.

OPEN IN VIEWER

Table 1

Ethnic Background of Bulgarian Legislators, 1990-2008

Ethnicity	Population Share (%)	Population in Absolute Figures	Legislative Share (%)	Legislative Frequency Count (N)	Proportionality of Representation Index
Bulgarian	83.94	6,655,210	91.46 (91.17)	1,617 (1,424)	1.09 (1.09)
Turk	9.42	746,664	6.73 (7.17)	119 (112)	0.71 (0.76)
Roma	4.68	370,908	0.62 (0.51)	11 (8)	0.13 (0.11)
Russian	0.197	15,595
Armenian	0.137	10,832	0.34 (0.26)	6 (4)	2.48 (1.87)
Vlachs	0.133	10,566
Macedonian	0.064	5,071
Greek	0.043	3,408
Ukrainian	0.031	2,489
Jewish	0.017	1,363	0.62 (0.70)	11 (11)	36.19 (40.97)
Romanian	0.014	1,088
Other	0.237	18,792
Pomak			0.23 (0.19)	4 (3)
No ethnic self-identification	0.783	62,108
Not indicated	0.313	24,807
Total	100	7,928,901		1,768 (1,562)

Source: Population data is from the 2001 national census; legislative data is based on authors’ calculations.

Note: Numbers for term starters only appear in parentheses. There was no category Pomak in the 2001 national census.

Two numbers are reported in each of the last three columns of the table: the first number refers to data from the entire dataset of parliamentary deputies; the number in parentheses is based on calculations that include only deputies who entered the parliament at the start of each term. These so-called term starters are distinguished from the deputies who entered later in a term as substitutes for deputies who left parliament for executive government positions or returned their mandate because of any other reasons. This was done to control for the possibility that different parliamentary turnover rates exist for deputies from ethnic majority and minority groups and that these differences affect the value of measurements reported in the table. As the results indicate, legislative shares of ethnic groups and proportionality of representation scores are rather similar for both types of deputy counts.

The table also indicates that a number of deputies from smaller ethnic groups also served in the parliament. The high degree of overrepresentation in the case of Armenians and Jews is a product of their small population shares. The population size of neither of the smaller ethnic minority groups in Bulgaria is even theoretically sufficient for filling a single parliamentary seat on the basis of demographic calculations—parliamentary representation is based on a norm of approximately thirty thousand citizens per seat.

The table includes the ethnic category Pomaks. Pomak (Muslim Bulgarian) is not an official ethnic category in Bulgaria, and therefore no census data are reported in the table. Estimates of the Pomak population, as well as population estimates for some other ethnic groups—the official census results are questioned in Bulgaria to a somewhat greater extent than in other Eastern European countries—are reported in Appendix A. Pomaks have a rather ambiguous and contested self-identification. ¹⁴ The reported number of Pomak deputies most likely underestimates the actual number of Pomak deputies. Deputies of Pomak background are more likely than deputies of other ethnic background, with the possible exception of Roma deputies, to withheld information about their ethnic origins due to stigmatization and lack of acceptance.

With regards to the patterns of underrepresentation of large minority groups, which were identified at the start of this section, a number of considerations and factors should be taken into account. The vast majority of deputies reported in Table 1 as having ethnic Turkish background entered the parliament through the list of the Movement for Rights and Freedoms (MRF). While ethnic parties are explicitly prohibited by the Bulgarian constitution, the MRF is de facto an ethnic party whose continuing electoral success has been secured by the considerable demographic weight and a high level of mobilization of the ethnic Turkish community. One important feature of the MRF’s candidate selection strategy, which will be discussed in greater details in the section on minority organizations, is multiethnic recruitment. The level of underrepresentation of the Turkish community reported in Table 1 is somewhat reduced when the total share of deputies elected on MRF’s ticket is taken into consideration.

The level of Roma descriptive underrepresentation is much higher and is not readily offset by any substantive representation of Roma interests by some electorally successful party in the Bulgarian case. Similarly to the Pomak numbers, the size of the Roma population is a disputed issue. ¹⁵ As the data in Appendix A indicates, some estimates assume the group size to be twice as large as reported in the last census. Regardless of the population estimates used, the underrepresentation of the Bulgarian Roma is severe even by the low standards of neighboring countries with a significant share of Roma population. In the Romanian case, for example, the degree of Roma underrepresentation is somewhat lower than in Bulgaria. This is, however, largely a product of positive discrimination measures (reserved seats provisions) rather than the willingness of Romanian parties to recruit Roma in significant numbers. ¹⁶

The general situation concerning the political participation of Roma in Eastern Europe has received considerable attention in the literature that deals with particular challenges this minority group faces in terms of social stigmatization and collective action problems. ¹⁷ How structural characteristics and the electoral behavior of Roma community affect the willingness of mainstream parties to include Roma candidates on their electoral lists and how these factors shape the ability of Roma political organizations to mobilize community support are among the key issues addressed in the following sections of the article.

While the persistence of PR rules in the Bulgarian case does not allow systematic evaluation of the effects that different electoral rules would have had on minority inclusion, the founding 1990 elections provided a setting for one natural experiment on the effects of electoral rules. The 1990 elections to a Grand National Assembly were held on a mixed electoral formula, with a half of seats elected according to PR and SMD rules, respectively. ¹⁸ The results allow the comparison of the effects of SMD and PR on minority representation, with an important caveat that this was an electoral contest at the very start of transition and electoral interactions were not repeated under similar rules in the subsequent rounds of electoral competition.

Table 2 indicates that the share of ethnic Turks elected in the 1990 assembly was slightly higher under SMD than PR. There were ten ethnic Turkish deputies elected to the parliament under the PR segment of electoral competition, and there were eleven ethnic Turks who won elections in SMDs. In the PR segment, nine out of the ten ethnic Turks were elected on the MRF’s electoral list. A lower share of ethnic Turks in the PR segment is, however, a function of MRF’s decisions to include a number of ethnic Bulgarians in their list. The winning portion of the MRF’s 1990 electoral list included three ethnic Bulgarians. The MRF recruited ethnic Bulgarians in subsequent elections as well—this practice is discussed in some details later in the text.

OPEN IN VIEWER

Table 2

Minority Representation across Parliamentary Terms (in percentages)

	Term 1
	PR	SMD	Term 2	Term 3	Term 4	Term 5	Term 6
Bulgarian	90.5	93.5	91.7	93.8	92.9	89.2	86.3
Turk	5.0	5.5	7.9	5.4	6.3	8.8	10.8
Roma	1.5	0.5	0	0.4	0	0.8	0.4
Others	3.0	0.5	0.4	0.4	0.8	1.3	2.5
All ethnicities	100	100	100	100	100	100	100

Note: Terms: 1 = 1990-1991; 2 = 1991-1994; 3 = 1995-1997; 4 = 1997-2001; 5 = 2001-2005; 6 = 2005-2009. PR = proportional representation; SMD = single-member district.

Roma and other minorities did significantly worse under SMD rules in the 1990 elections, which confirms the usual expectations about the effects of SMD on minority representation. ¹⁹ In all subsequent legislative elections—to a regular parliament rather than a constitutional assembly—a closed-list PR system was used. As Table 2 indicates, the overall share of minority deputies varied across the terms. The largest value of this share was recorded in the latest parliamentary term. An increase in the number of deputies both from ethnic Turkish community and smaller minority groups contributed to the growth of this share in the latest term. In neither of the terms, however, has the minorities’ legislative share approximated their population share.

Minority Inclusion in Mainstream Political Parties

Recruitment decisions are an important aspect of party activities. ²⁰ In the case of the closed-list PR electoral system, which has been in place in Bulgaria for almost the entire post-communist period, party leadership exercises considerable power over who is put on the list by controlling appointment procedures. In most of the cases the party leadership has also reasonably accurate expectations about how many candidates from their lists are likely to enter parliament in any given elections—these expectations are usually formed on the basis of preelectoral polling, which is an increasingly widespread and frequently used procedure. The composition of party factions in parliament therefore reflects party leadership priorities in terms of candidate selection.

Mainstream parties’ decisions to recruit ethnic minority representatives can be seen as driven by different types of considerations. Under any of them, however, parliamentary seats constitute a scarce and highly valuable set of prizes for political parties. Allocating even a single seat to any specific social group is a costly act for the party; it is also a meaningful signal of the party’s commitment to the group. Drawing on general literature on party behavior, ²¹ it might be analytically beneficial to differentiate between electoral and programmatic reasons for recruiting minorities. Electoral reasons are based on vote maximization considerations. From this perspective, including minority representatives on party lists makes sense when it has a promise of vote gain. Programmatic reasons are grounded in party ideology and do not carry the same instrumentalist connotation. Thus, for example, left-wing parties are sometimes described in the literature as minority-friendly due to ideological preferences for diverse and inclusive membership.

These premises help to generate the following simple propositions, support for which our dataset allows us to examine. First, electoral considerations should make mainstream parties especially interested in the recruitment from the large minority groups, which, in the Bulgarian case, are represented by ethnic Turks and Roma, rather than from demographically smaller groups. Second, programmatic reasons should lead the Bulgarian parties of the left to be more inclusive than parties representing other ideological families.

The inspection of the data reveals that neither of these propositions holds entirely true in the Bulgarian context. Table 3 provides details on the distribution of minority deputies across three major party families plus the ethnic minority party family. Given that political contestation on socioeconomic issues constitutes the basis for the main dimension of party competition in Bulgaria, ²² all electorally successful political parties and their deputies can be reasonably grouped in one these major party families.

OPEN IN VIEWER

Table 3

Distribution of Minority Deputies across Party Families

	Left	Center	Right	Minority	Total Number
Turk	1	0	1	117	119
Roma	6	2	3	0	11
Armenian	0	5	1	0	6
Jewish	4	3	4	0	11
Pomak	0	0	1	3	4
Minority total	11	10	10	120	151
All ethnicities	694	239	697	136	1,766
Minorities share (%)	1.59	4.18	1.43	88.24

Note: There were two independent deputies in the 7th Grand National Assembly, elected in single-member districts (SMDs), none of them from a minority group. The table includes all term starters and late comers. Classified according to parliamentary groups of parties and electoral coalitions. Right-wing: Ataka, BANU, CO BPU, DSB, FPL, PU, PU UDF, UDF; left-wing: BSP, CB, DL, EL, PU BSP, UF; center: BBB, LC, NMS2, UNS without MRF deputies; minority: MRF, MRF deputies in the UNS. Appendix C contains a coding sheet with full names and abbreviations.

The data indicates that there are practically no differences in terms of overall rates of minority inclusion between parties of the left and the right. The share of minorities in the groups of the political left and right was miniscule: 1.59 percent and 1.43 percent, respectively. The share of minorities was significantly higher for centrist parties that brought into parliament a group of deputies numerically smaller than groups of deputies elected on the lists of the left and right parties but that nevertheless had more than two hundred deputies in their ranks. The share of minority deputes was, naturally, the largest in the minority party family, which was composed of MRF deputies elected either on the individual party ticket or on the ticket of a preelectoral coalition that included the MRF. As mentioned above, the MRF’s ranks include also a number of deputies of ethnic Bulgarian background.

The noninclusion of ethnic Turks in the mainstream parties is an important fact, which can not be explained away by simply citing the strong electoral presence of the MRF. The latter’s electoral performance has been strong but so exceptionally strong at the very start of transition as to serve as an ultimate deterrent for other parties’ attempts to court the Turkish vote. The MRF vote also varied considerably across the post-communist period. Furthermore, the party’s political behavior, as the further discussion will indicate, has caused a considerable amount of protest inside the Turkish community. Given the high magnitude of the PR electoral system in the Bulgarian case, even a small increase in a vote for mainstream parties delivered by minority communities can have a considerable effect on these parties’ share of parliamentary seats. Explaining why such considerations of potential electoral benefits from appealing to the ethnic Turkish group have not been realized in the actual recruitment policies of the mainstream parties requires consideration of the parties’ programmatic positioning on issues other than socioeconomic.

Positions of the leading parties of the Bulgarian left and right on another important ideological dimension, ethnic culture, have been quite close, especially at the early stages of transition. Both political camps positioned themselves near the integrationist pole on the multiethnic-integrationist ideological dimension. This decision was heavily influenced by the difficult history of relations between the Bulgarian majority and the ethnic Turkish community, in which politicians from both camps had been involved.

The ability of the post-communist left to engage with this minority has been especially compromised due to such nationalistic policies of its communist predecessor as a “rebirth” campaign directed towards Bulgarization of the ethnic Turk community in the second half of the eighties. ²³ But proponents of forced assimilation occupied prominent positions also in the right-wing party camp dominated by the Union of Democratic Forces (UDF). One illustration of the strength of nationalist lobbies in the UDF is the fact that the leader and two of the most prominent figures of the extreme nationalist and xenophobic Ataka party, which has risen since 2005, occupied leading positions in the UDF in the early 1990s. Volen Siderov, Petur Beron, and Pavel Shopov were, respectively, editor in chief of the UDF newspaper Democracia, secretary and chair of UDF until November 1990, and deputy chair of the Christian Democratic Union in the UDF. As one account of early party formation in Bulgaria reports, a significant number of UDF founders and members were either “turncoats” from the Communist Party or “social outsiders.” ²⁴ Furthermore, the UDF’s ability to formulate any coherent policy positions or specific recruitment initiatives that would target the Bulgarian Turks was severely hampered by the structure of the union. Consisting of at times eighteen parties and organizations, the UDF was internally divided, and its decision making on controversial issues was severely hampered. Overall, the Turkish community was distrusted by the main political forces that came to dominate the post-communist transition and that continued to exploit the antagonism between the majority and the country’s largest minority group. Additionally, prevailing public attitudes can be seen as contributing to the lack of mainstream parties’ willingness to put Turkish candidates on their electoral lists. Doing so might have been perceived by mainstream politicians as carrying the risks of losing the support of their parties’ core constituencies.

There was, however, a considerably larger presence of members of the country’s second largest minority group, the Roma, on the lists of mainstream parties. A detailed examination of these instances of Roma recruitment provides considerable evidence for the assertion that these decisions by mainstream parties were electorally motivated. Although Roma political participation in Bulgaria, as in other Eastern European countries, has been characterized by low voter turnout, securing Roma support still had a promise of considerable electoral gains for the mainstream parties. Unlike ethnic Turks, the Roma community was neither politically mobilized at the start of transition nor perceived as a threat by majority politicians.

As Table 3 indicates, the largest number of Roma entered parliament through the parties of the left. This reflects the traditional sympathy of the Roma community for communist successor parties. Barany identifies this pattern of Roma support for communist successor parties across Eastern Europe and generally attributes it to the Roma perception of the communist period as the one that provided them with a relatively stable and prosperous socioeconomic environment. ²⁵ Table 3 indicates, however, that centrist and right parties have also engaged in Roma recruitment efforts, thus denying the left of a potential monopoly on the inclusion of Roma candidates in highly visible political positions.

The overall presence of Roma in parliament, as was demonstrated in Table 1, is still well below their population share. This high level of underrepresentation reflects the Roma’s inability to mobilize sufficient electoral support for any of the Roma parties to enter the parliament on its own terms. The inability to deliver votes also weakens Roma candidates’ bargaining with mainstream political parties. The limited number of Roma candidates included in the winning portion of electoral lists of these parties reflects the outcomes of bargaining between political parties and potential Roma candidates in which both sides are aware of the rather limited ability of Roma activists to secure Roma votes.

The important piece of evidence for the electoral motivations in the mainstream parties’ decisions to recruit Roma candidates comes from our detailed analysis of social and political background of all minority deputies found on the list of mainstream parties. This data allows to test one implication of the proposition that electoral calculations dominate mainstream parties’ decision about recruitment of minorities only in the case of large demographic groups. If this proposition is true, minority deputies from small and large ethnic communities could be expected to differ in terms of the strength of their organizational ties to the minority community. Minority deputies from smaller ethnic communities should have fewer significant ties with their communities than representatives of large groups. This is because the former might be selected on grounds other than ethnic affiliation while the latter are valuable for the parties because of their connections to the large ethnic community and their ability to help secure community electoral support. We provide an example of how organizational ties of individual deputies were coded in Appendix D. The data is summarized in Table 4.

OPEN IN VIEWER

Table 4

Community Links of Minority Deputies in Mainstream Parties

Deputies’ Ethnicity	Total in Mainstream Party	Deputies with Organizational Links
Turk	2	1
Roma	11	5
Armenian	6	1
Jewish	11	1
Pomak	1	0
Minority total	31	8

Note: There were two independent deputies in the 7th Grand National Assembly, elected in single-member districts (SMDs), none of them from a minority group. The table includes all term starters and late comers.

Our expectations find support in the case of Roma deputies. As the table indicates, five out of eleven Roma who served on behalf of mainstream parties in the Bulgarian parliament occupied leadership positions in various Roma organizations prior to being put on the parties’ electoral lists. In contrast, only one out of eleven deputies with Jewish background is reported as having occupied leadership positions in Jewish organizations prior to entering the parliament. The case of ethnic Turkish deputies found in the rosters of mainstream parties provides mixed support for our expectation. Only one of them had significant standing in the Turkish community. The fact that there were only two deputies of ethnic Turkish background on the deputy rosters of mainstream parties highlights our earlier point about these parties’ decision to abstain from targeting Turkish voters.

The results from Table 4 suggest that most deputies from smaller minority communities that are found on the lists of mainstream political parties should be viewed as individual political entrepreneurs whose personal resource endowment—either in the form of membership in influential policy networks, private wealth, or established professional reputation—makes them attractive candidates for political parties. Recruitment of deputies with membership in the Jewish community, which is numerically a very small group in the case of Bulgaria, most closely approximates this model of candidate selection. As was earlier reported in Table 3, deputies of this ethnic background were almost equally distributed among the three major party families, which testifies to the presence of opportunities for members of this minority group to enter politics through different party channels.

This analysis points to one important feature that is sometimes lost in discussions of minority representation: minority presence in the parliament is not always a product of the political salience of ethnicity. The appearance of minority group members on the lists of political parties might have little to do with electorally or ideologically motivated targeting of minority community members in the recruitment process. A number of deputies of the Bulgarian parliament selected on other than ethnic grounds just happened to be members of a minority group. This insight becomes important when issues of substantive representation of minority interests in the legislative process are discussed.

Ethnic Representation through Minority Organizations

Ethnic representation is further a product of the electoral performance of minority organizations. And group-specific characteristics affect the ability of minority organizations to succeed under the PR electoral rules. These characteristics include both structural and behavioral factors. Structural factors refer to variables such as demographic size and regional concentration of minority groups. Discussions of behavioral and attitudinal factors usually deal with vote participation rates, levels of group identification, and stability of electoral preferences. These types of issues received some amount of attention in the general literature on party politics in Bulgaria, ²⁶ but have not been extensively treated in the writings on minority participation. ²⁷

The role of institutional factors and, first of all, the nature of impact that a choice of specific PR provisions have on the electoral success of minority organizations and on subsequent patterns of minority representation, has also been discussed in this literature only in passing. In the Bulgarian case, two important institutional features, which are either directly or indirectly linked to the design of PR electoral system, have affected the minority representation outcomes. These are a relatively high electoral threshold for entering the parliament and a constitutional ban on ethnic parties. The constitutional ban on ethnic parties, as the constant presence of the MRF in national politics indicates, has been only partially enforced, yet, as the following discussion will indicate, it affects calculations of political actors and minority representation outcomes.

Comparing the outcomes of the efforts of Bulgaria’s two largest minority groups to self-organize and to secure parliamentary representation helps to illuminate the effects of institutional and structural variables on the levels of minority presence in national politics. The early adoption of PR rules in the Bulgarian case had little effect on improving the levels of Roma inclusion. As the aggregate data in Table 1 indicates, Roma have been the principal losers in terms of proportionality of representation. The data on changes in the levels of Roma parliamentary representation over time, which was reported in Table 2, also points to the fact that there have been no improvements in Roma representation in the course of the two decades after the start of transition.

Structural and attitudinal characteristics of the Roma group go a long way in explaining the Roma’s inability to take advantage of opportunities presented by the PR electoral system. In terms of attitudinal and behavior characteristics, Barany describes Roma electoral behavior in the following terms: a tendency to vote for the party in power or for the party that is expected to win, significant support for the successors of former communist parties, low support for ethnic Roma candidates due to “little confidence in their own,” and voting participation rates that are far below that of the rest of the population. ²⁸ Obviously, there is a significant cross-national variation in terms of empirical support for these patterns identified by the author on the basis of a combination of quantitative data and numerous qualitative interviews with experts and practitioners.

In terms of structural factors that complicate efforts to construct Roma organizations that could be electorally viable on the national level in Bulgarian politics, Barany’s list could be amended by considering factors such as demographic size and concentration. Table 5 provides one measure of the territorial concentration of minorities. The calculations are made using Bulgaria’s administrative organization, which includes thirty-one districts, as a basis for the analysis. The table lists the largest minority share in the population of a single administrative district for each of the country’s five largest minority groups. The table also reports the size of this share relatively to the total size of a minority group.

OPEN IN VIEWER

Table 5

Regional Concentration of Ethnic Groups

Ethnic Minority	Population	Largest % of the Population of a Single Region	Region’s Minority as % of Total Minority Population
Turk	746,664	61.65	13.54
Roma	370,908	12.5	6.14
Russian	15,595	0.31	2.83
Armenian	10,832	0.48	20.68
Vlach	10,566	0.78	34.26

As the table indicates, the levels of concentration for the Roma community are much below the levels of concentration for the Turkish group. Geographic dispersion of the Roma population means that Roma activists face considerably larger organization-building and coordination costs in their efforts to participate politically than their Turkish counterparts. The resulting fragmentation of Roma political organizations translates into very modest electoral returns for each of the several Roma parties that try to compete in national elections. To date, the best performance for a Roma organization in the parliamentary elections was the 2005 election result for the “Evroroma” party, which received 45,637 votes or 1.25 percent of the national vote. The situation is slightly better at local level, and the total vote for Roma parties in local elections has been growing. Roma parties received 84,044 and 90,116 votes in the 2003 and 2007 local elections, respectively. ²⁹

Institutional features of the Bulgarian PR system, however, also play a role in explaining the fragmentation of Roma political efforts to gain parliamentary representation. The existing electoral threshold provision makes it very difficult for the Roma community, which accounts only for 4.68 percent of the population, to build an electorally successful party. Given that the group’s size only minimally exceeds the 4 percent threshold and Roma have a well-known record of low voter participation, Roma politicians have few incentives to invest in the construction of a unified party capable of winning parliamentary representation through regular electoral channels. A set of beliefs about potential gains from pursuing different electoral strategies that Roma politicians have formed is heavily influenced by the existing institutional provisions and encourages proliferation of Roma political organizations.

Individual Roma leaders’ payoffs from a “go alone” strategy are presented, on the national level, in the form of possibilities to strike a deal with mainstream parties about the inclusion of minority organization leaders into the winning portion of mainstream parties’ lists. As the previous discussion of the organizational ties of minority deputies revealed, Roma representation in the Bulgarian parliament has been facilitated by informal coalition-building between mainstream parties and the representatives of Roma organizations. These representatives were put in the winning portions of party electoral lists in exchange for their organizations’ support in campaigning for the Roma vote in election periods. The terms of this exchange—as revealed by a very limited number of Roma on the lists of electorally successful parties—is rather unfavorable for Roma.

Unlike the Roma community, ethnic Turks were successful in building an organization, which, despite the existence of numerous legal and political challenges, was consistently able to secure parliamentary representation. The patterns of ethnic Turkish politics provide a good illustration of the impact of institutional settings on minority participation and representation. One peculiar feature of the Bulgarian institutional design is the constitutional ban on ethnic parties. The behavior of an electorally successful Turkish party, the MRF, reflects the party’s efforts to avoid risks of being banned. Attempts to use this constitutional norm to prohibit the activity of minority organizations have actually been made, which somewhat increases credibility of the norm. ³⁰ The MRF’s strategy for dealing with these risks involved making adjustments both in the programmatic profile and the organizational structures of the party. The candidate selection practices reflect party efforts in this respect.

From the start of its operations the MRF has made conscious efforts to use recruitment as means of avoiding a status of a mono-ethnic minority organization. The MRF consistently recruited and promoted a small number of ethnic Bulgarians through the ranks of their organization. As Table 4 indicates, ethnic Bulgarians were given places on the winning portion of the MRF’s electoral lists in all but one parliamentary election. Given that parliamentary seats are a highly scarce and valuable commodity for a party, the MRF’s decision to allocate a number of these seats to ethnic Bulgarians was a costly one. The party’s willingness to bear these costs is likely to be motivated in part by the existence of constitutional restrictions on political activity of minority organizations. A genuine desire to increase its appeal outside the Turkish community might also play a role in party recruitment decisions, which are a part of a general strategy of targeting nonminority vote. As Hajdinjak reports, there is evidence that this strategy has brought some modest electoral benefits for the party. ³¹

The view that the inclusion of ethnic Bulgarians on the MRF’s list is a move intended to placate the titular group rather than a genuine commitment to establishing and running a multiethnic organization, however, is given some support by the absence of ethnic Roma from the MRF’s parliamentary rosters. Another piece of evidence is constituted by the absence of ethnic Bulgarians from the MRF’s electoral list when things started getting tough for the party and opportunities existed for somewhat concealing the composition of the list. The decline in voter support and some other factors forced the MRF to join the preelectoral coalition in the 1997 parliamentary elections rather than to contest elections, as it did before and after, on its own. The MRF’s candidates were included into the general electoral list composed by the Union for National Salvation (UNS) coalition. As Table 6’s data for the fourth term indicates, neither of the MRF candidates that entered parliament on the UNS’s list was of other than Turkish ethnic background.

OPEN IN VIEWER

Table 6

Ethnic Composition of MRF Deputies

	Deputy Count (N)
Ethnicity	Term 1	Term 2	Term 3	Term 4	Term 5	Term 6	Total	Share %
Turk	20	21	13	14	23	26	117	86.03
Bulgarian	3	5	2	0	1	5	16	11.76
Pomak	0	0	0	0	0	3	3	2.21
Total	23	26	15	14	24	34	136

Note: Included are the deputies of the Movement for Rights and Freedoms (MRF) parliamentary groups and the MRF party members in the UNS coalition, both term starters and late comers.

Table 6 also shows that three out of a total four Pomak members of the Bulgarian parliament identified in the course of our research have entered parliament on the ticket of the MRF. This reflects the existence of ties between the Muslim Bulgarian and Turkish communities, which are reflected in a considerable electoral support for the MRF in Pomak-dominated regions. Pomaks do not constitute a community with an unambiguous identity and clearly defined membership in the Bulgarian context, yet the attempts to engage in independent political campaigning on behalf of the Pomaks have been undertaken. A number of politicians were linked to attempts of establishing a Pomak party. For example, in 1993, Kamen Mitkov Burov from Zhultusha village in the Rhodope mountains registered the “Democratic Party of Labour” and claimed that the party would represent the interests of the Pomak population, especially in the Rhodope mountains. ³² The MRF’s recent efforts to put prominent members of the Pomak community in positions of power and prestige—through their inclusion in the winning portion of the party’s electoral list—could be seen as a response to ethnic entrepreneurs’ attempts to politically activate the Pomak community.

Internal developments inside the Turkish community appear more prominently in Bulgarian politics than the sporadic attempts to politicize Pomak identity. Another feature of Bulgarian institutional design, the already mentioned 4 percent electoral threshold, plays some role in explaining a major trait of minority intragroup politics, the MRF’s monopoly on representation of the Turkish community at the national level. Despite the fact of very significant levels of community dissatisfaction with the MRF’s performance, causes of which will be summarized below, the party was able to maintain its dominant position throughout the entire post-communist period. Neither of the political alternatives—other ethnic Turkish parties that have been registered and have been allowed to contest elections—has been successful in establishing itself on the political scene and in gaining a considerable electoral following.

The presence of a relatively high electoral threshold is one of the problems that politicians interested in establishing alternative political projects face. Although the group’s demographic size, which is 9.42 percent of the total population, is theoretically large enough to accommodate the existence of two electorally successful parties, the 4 percent electoral threshold provision imposed practical constraints on the possibility of coexistence of two electorally successful ethnic Turkish parties. Appendix B provides a graph of the MRF’s electoral results. In none of the rounds of parliamentary elections, except the 2005 elections, was the MRF able to receive even 8 percent of the total national vote, which would constitute some minimal theoretical requirement for successful operation of two ethnic Turkish parties. ³³ This information is a piece of common knowledge that shapes both ethnic Turkish political activists’ and voters’ expectations about electoral outcomes and makes it very difficult for ethnic Turkish politicians to construct viable political alternatives to the MRF.

While the existing literature has pointed to the benefits of the MRF’s monopoly on group representation, which, in view of some authors, helped the MRF’s leadership to sustain an accommodationist course in relations with the titular group, ³⁴ this monopoly has had considerable costs. The primary component of these costs comes in the form of a lack of accountability of the MRF’s leadership. The main criticisms against the MRF, which have been leveled by a wide range of individuals and organizations inside the Turkish community, were not cast in terms of demands for a more radical policy stance vis-à-vis the titular group. These criticisms primarily deal with what is perceived as the persistent use of public office for private gains by the MRF leadership and officials. The MRF’s representatives have been accused of prioritizing not service to the community but the achievement of such other goals as political career advancement, accumulation of personal wealth, or securing economic gains for narrowly defined interest groups. ³⁵ The MRF’s activity has been plagued by numerous accusations of personal corruption and self-serving links between the party and a small group of businesses. The MRF leadership has been, for example, implicated in operating a business center from the premises of their political headquarters. Another major allegation against the MRF’s leadership concerns its ties to the socialist security apparatus. These ties are alleged to continue to operate in the after-socialist period, benefiting the MRF’s political and business operations. ³⁶ This constitutes a typical case of clientelistic exchange between politicians and interest groups in which the former use their access to government decision making to exchange procurement decisions or regulatory favors for financial contributions.

The other component of these costs is directly relevant to legislative recruitment issues. The MRF’s monopoly on representation has made the party nonresponsive to the needs of guaranteeing internal democracy and social inclusiveness inside the organization. For example, the MRF is not inclusive in gender terms. The overall share of women in the Bulgarian parliament during the 1990 to 2008 period was 14.7 percent. For the same period, the share of female deputies in the MRF’s parliamentary group was only 5.8 percent. The MRF is also routinely criticized for the lack of turnover in leadership positions and authoritarian management both by political and nongovernmental organizations working in the Turkish community. ³⁷ According to the party statute, the chair of the MRF has exceptional power and confirms the candidate deputies for the national parliament and the candidate mayors. On this power and the governance style of the party elite, Mehmed Dikme, a former MRF minister of agriculture, said, “Because the way in which the party is governed now—the introduction of authoritarian, totalitarian mechanisms of government—takes the party back to the times even before 1990.” ³⁸ In an interview in the Bulgarian newspaper Monitor, Sezgin Miumiun (founder of Federation Justice Bulgaria, a Turkish NGO) explains the reasons for the success of the MRF and the lack of interest of the MRF leadership in an official investigation for the culprits of the “Rebirth” campaign: “The MRF claims the authorship on the issue of the victims, that have to be protected and avenged by the permanent participation of the movement in government…. In their structure both the MRF and ATAKA are totalitarian parties. There is no chance that in this state they will create a fertile space for the search for truth. They exploit the problem and cultivate opposition. And that is why the society is full of aversion to hear what exactly has happened.” ³⁹ Ahmed Dogan—the chair of the MRF—shines brightest in the focal point of the criticism targeting the MRF. As cited, he is charged for his excessive party power and authoritarian leadership style but also for his dubious role in the “Rebirth” campaign when he was arrested, imprisoned, and sentenced for the creation of an antistate organization, while at the same time being an active state security agent. In fact, the lustration committee of the Bulgarian parliament found out that Dogan had started work for the security apparatus in 1974 while studying philosophy and continued his collaboration during his career in the philosophy department at the Bulgarian Academy of Sciences until at least 1988, when his dossier was closed. ⁴⁰ His popularity does not profit either from his continued self-portrayal as the only decision taker and leader in a movement that is the sole guarantee for the security and prosperity of the Bulgarian Turks and the fact that he publicly admits but also arrogantly defends the operation and support of business enterprises linked with the party, his person and the remaining party leadership. ⁴¹

Overall, a large number of facts from different facets of the MRF’s political behavior come together to form a strong case against the MRF. The monopoly on group representation, facilitated by the design of the country’s electoral institutions, resulted in many abuses of public trust, which calls into questions usual assumptions about the independent value of descriptive representation for securing an adequate substantive representation of minority interests.

Conclusion

Bulgaria’s choice of closed-list PR electoral rules from the very start of the post-communist transition has provided researchers with an opportunity to examine the effects of these rules on minority representation in a transitional country with a stable institutional setting. The article demonstrated that representational demands of main minority groups were processed under these rules in different ways. Party ideologies were argued to be an important intervening variable in explaining the mainstream parties’ decisions about minority recruitment. In the case of the Turkish minority, exclusionist policy stances of the leading Bulgarian parties outweighed any considerations of potential electoral benefits from inclusive recruitment. The mainstream parties’ initial choice not to target the Turkish community has been subsequently reinforced by the electoral successes of the ethnic Turkish party. These successes served to further deter any mainstream party from engaging with the Turkish community.

In the case of another large minority group—the Roma—the PR rules did indeed have an effect of providing mainstream parties with sufficient electoral incentives to recruit members of that minority. These incentives were responsible for the fact that the mainstream parties across the ideological spectrum sought Roma candidates for their electoral lists. The slightly higher share of Roma deputies on the lists of parties of the left suggests that disadvantaged minority groups might have a special interest in establishing political ties to parties with a redistributive agenda. The extent of Roma inclusion was, however, very minimal in all cases. The number of Roma deputies who served in the Bulgarian parliament through two decades of post-communist transition suggests only a nominal presence of Roma representatives in the national legislative process.

The examination of the Bulgarian experience suggests that the adoption of the PR system should not be automatically assumed to lead to representational gains for minority communities. The fact of the adoption of PR rules tells us little in itself about how and to what extent minorities will be included in the political process. Turks and Roma have been included into politics in a very different way: the former through monopolized representation by one ethnic party, the latter through cooptation of its representatives in different mainstream parties.

While minority group characteristics and party ideologies are important in explaining representational outcomes, the details of the PR system matter as well. In terms of relatively alterable elements of electoral design, the article identified electoral threshold provisions as a main culprit. While the use of high thresholds is justified by the need to balance general political system concerns about representativeness and effectiveness, the article pointed to a number of adverse and largely unintended effects of this provision on minority representation. The irony of the application of PR electoral rules in the Bulgarian case is that they did little to improve representation of the group that needed it most—the Roma. The introduction of a relatively high threshold provision from the very start of transition denied Roma political organizations any prospects of entering the parliament on their own. The same provision had also an effect of stifling political competition inside the Turkish community and reducing the accountability and responsiveness of the MRF’s political leadership. The questions of when the benefits that much of the literature associates with PR are actually forthcoming, or what the implications of having this type of electoral design are for accountability and responsiveness of minority elites, constitute important topics for further research.