Political dejection in a divided society: a challenge for Latvia’s democracy?

ABSTRACT

Ethnically divided societies often seek political cohesion by pursuing nationalizing policies. Latvia has had some experience of this, for example, with the notion that citizens should speak the national language and respect government institutions coincides with portrayals of ethnic minorities as a challenge to democracy. The focus on nation-state building has resulted in persistently low levels of political engagement and public confidence among the citizenry. Public survey data shows that many voters believe their concerns have been neglected by elected representatives, who have fixed their attention on ethnopolitical issues. While we see no demonstrable correlation between ethnicity and levels of political dejection, ethnonational politics explains unstable voter-representative relations and electoral volatility during the Saeima elections from 2010–2018. These unanticipated effects of nation-state building on Latvia’s electoral politics invite a rethinking of the relationship between voters and their elected representatives in democratically consolidated, yet divided nations.

KEYWORDS:

• Political representation
• political participation
• democratic consolidation
• divided society
• Latvia

Introduction

Unless citizens elect representatives, who share their views, and parties aggregate voters’ diverse interests into policies, governments run the risk of focusing on issues that are inaccurate or irrelevant to their electorate. Reflecting on the repeated electoral success of anti-establishment parties in economically weak and less consolidated democracies in eastern Europe, scholars have long pondered the source of electoral dissatisfaction with key democratic institutions, such as elections and parliament (Abedi 2004; Sikk 2005; Mesežnikov, Gyarfášová, and Bútorová 2013; Pirro 2014; Bågenholm 2013). Limited confidence in institutions of electoral democracy could indicate doubt in the capacity of the democratic regime to transmit citizens’ economic and identity preferences into corresponding policies (Landwehr and Steiner 2017; Bohle and Greskovits 2012). In this article, we assess how Latvian voters perceive the conversion of public interests into public policy and what effect this has on citizen engagement in electoral politics in Latvia’s consolidated, ethnically divided democracy.

The literature on citizen input into democratic processes has evolved rapidly to include assessments of the political landscape in countries affected by the 2008 financial crisis (Caporaso 2018; Usherwood 2019; Stavrakakis et al. 2018). Much of this scholarship points out that political elites’ prioritization of some politically salient issues over other socially relevant ones – for example, the economy’s competitiveness over social welfare – has adversely impacted electoral participation over the past decade (Van Biezen and Poguntke 2014; Hooghe and Quintelier 2014). The Baltic states, Latvia included, were severely affected by the crisis. Yet, our analysis focused on a different politically salient issue that has been dominant throughout the post-Soviet independence period: nation-state building.

Rebuilding democratic nation-states has been high up on political agendas in the Baltic region since the end of the period of Soviet inclusion. Ethnic majorities in Estonia and Latvia, particularly, have long perceived resident Russian speakers to be a liability for the region’s democratic development (Galbreath 2006). Over the pre-accession decade, the ethnic majority dominated political decision-making, facilitated the consolidation of state regimes, and controlled the political process around nation-state building. Though controversial, Estonia and Latvia’s 1990s exclusion of Russian speakers from automatic access to citizenship – and by extension, to electoral rights – ensured political stability, democratic consolidation, and successful European integration (Regelmann 2014; Cianetti 2018; Alijeva 2017). Yet, there is a widespread perception of Russian minorities as a threat to political stability and social harmony in the Baltic states (Pridham 2018). The perceived geopolitical vulnerability of the Baltic states at the EU’s external border has dramatically increased following Russia’s engagement in Ukraine. Changes to nation-state building policies, therefore, appear unlikely in the near future (Lamoreaux and Dyerly 2018).

Politics in Latvia has long been shaped by concerns over nation-state building. As we show below, ethnicity-based exclusion has wielded a considerable effect on democratic participation in Latvian society in general. Ten years ago, it was claimed that the emphasis on ethnonational politics ‘harmed’ the electoral participation of minority groups (Brosig 2010). Today, some 14.1% of Latvia’s long-term residents (including non-citizens [nepilsoņi] (11.1%) and citizens of other states (3%), predominantly Russia) cannot participate in electing Saeima (Latvia’s parliament) deputies or municipal representatives (Centrālā Statistikas Pārvalde 2019). Furthermore, a recent official report suggests that only around 35% of Latvia’s non-citizen, long-term resident population would like to become Latvian citizens and, by extension, participate in the electoral process (Pilsonības un Migrācijas Lietu Pārvalde 2017). If a considerable number of people affected by decisions of political elites do not (wish to) participate in electoral politics, the reasons why some citizens do not go to the polls while others do but are unhappy about the choices on offer, deserves closer attention.

In this context, we focus on voter confidence in democratic institutions within the country. The repeated electoral success of new political parties suggests that incumbent politicians often fail to respond to voter priorities. While we assume that citizen preferences are adequately represented by elected officials, voters from both sides of Latvia’s ethnic spectrum have ‘thrown out the bums’ (to remove lawmakers they dislike) (Pop-Eleches 2010) over successive elections, clearly indicating the prominence of protest votes in recent elections. We identify that the continuous focus on nation-state building, to the exclusion of many other socially relevant and politically salient issues, negatively impacts the electorate’s trust and participation in the democratic process. Literature on electoral politics in Latvia, and in ethnically divided democracies more generally, underestimates the negative effect of ethnopolitical hegemony on voter satisfaction and electoral confidence across the country’s population. Ethnocentric politics, therefore, has not only been a problem for minorities, but it has also negatively affected broader democratic processes and contributed to widespread voter disengagement.

This article is structured as follows: First, we provide both the conceptual and contextual background for our argument. Then, we lay out the types of Latvian voters, based on their attitudes toward democratic institutions and participation. After presenting the data and methodology of this study, we offer details on the statistical analysis underlining that the focus on nation-state building has been the main cause for electorates’ dejection from democratic institutions. Finally, reflecting on quantitative results, we draw attention to their implications for Latvia’s political process over the past decade. In conclusion, we sketch broader implications for political representation in ethnically diverse societies, which, despite the consolidation of electoral democracy, are preoccupied with the political priority of nation-state building.

Background

The institutions of modern, liberal democracies offer multiple formal avenues to translate the electorates’ diverse political, economic, and social preferences into policies. This can happen via the electoral process, the structure of the electoral system, distribution of electoral districts, voting patterns, and the rules of cabinet formation, etc. Yet, a majoritarian democracy encourages elected representatives to emphasize some issues over others, thus prioritizing the representation of one interest group over others. Across the board, scholars highlight policies focusing on nation-state building in ethnically diverse societies uniquely impacting how different ethnic groups understand, relate to, and participate in politics. This section contextualizes our study by reviewing the scholarship on ethnic politics. We particularly note the literature about nation-state building in divided societies, ethnicity-based party systems, levels of majority and minority participation in the democratic process, and trust in institutional democracy.

The dominance of the nation state, in terms of economic, cultural, social, and political relations is acknowledged across the post-Communist region. Yet, over the past three decades, particularly in the Baltic states, elected representatives credibly demonstrated their commitment to liberal democracy. In Latvia, the ethnic majority has framed reforms as a definitive step necessary to consolidate democracy and justify the economic and social cost of transition away from Communism. The securitization of minorities in domestic politics has become the central feature of post-Communist nation-state building (Roe 2006; Galbreath and McEvoy 2012; Jutila 2015). The traumatic experience of the socialist past justified elected elites’ foremost concern with nationalizing policies, at the expense of societal, economic, and political agendas that include minority interests (Agarin and Cordell 2016). Reflecting on Latvia’s experience, Cianetti (2018) points out that, among minorities, nationalizing policies have decisively affected perceptions of the democratic credibility of the contemporary political regime.

The literature on nation-state building in the Baltic region leaves no doubt that concerns over cultural diversity continue to hold sway over domestic politics. Polarized views on national policies around minority accommodation explain electoral preferences, voter turnout, trust in state institutions, and satisfaction with democracy (Poleshchuk and Tsilevich 2004; Kochenov, Poleshchuk., and Dimitrovs 2013). Overall, the challenge of reconciling the competing logics of democratization and nation-state building in such divided societies is assumed to be the major reason why minorities disengage from politics (Agarin 2011; Nakai 2014; Cianetti 2015).

The political salience of ethnic cleavages makes Latvia a unique testing ground concerning the difficulties that political organizations face when aggregating voter preferences into public policy. Latvia’s electoral politics are centered around parties claiming to represent the interests of ethnonational communities. This situation is generally found in ethnically divided societies. Since majority groups tend to close ranks and support the institutions of ‘their’ state (Csergő and Regelmann 2017), minority groups are expected to challenge national political cohesion (see Salehyan 2009; Van Houten 1998). Mylonas (2013, 4) even suggests that minorities are likely to undermine the democratic politics of their countries of residence. Yet, as Cederman, Buhaug, and Ketil Rød (2009, 499) remind us, ‘exclusion along ethnic lines often led to political mobilization by counter-elites, which denounced their under-representation at the centre of power and demanded inclusion.’ As scholarship on ethnically divided societies points out, elected representatives from among minority communities tend to hamper the shift from identity to issue-based politics (Fearon and Laitin 2003), perpetuating ethnopolitical agitation (Lederach 1997) and contributing to ‘ethnic seepage’ (Charnysh, Lucas, and Singh 2015). This is because ethnicity-based elected leaders tend to be mindful of their co-ethnic electors (Horowitz 1985), casting a shadow over their ability to bridge ethnic divisions. As McClelland (2019) skillfully shows, Latvia offers a useful example in this regard: The persistent references to ethnicity and the allegiance of elected political representatives in the Latvian context remain remarkably potent tools for political mobilization of the electorate, particularly among the majority ethnic group.

Constituting around 20% of the entire electorate in 2020, Russian speakers are believed to vote predominantly for the minority-friendly ‘Harmony Center,’ the party that secured around a quarter of the members of parliament (MPs) in recent elections (2011, 2014, and 2018). In the 2011 Saeima elections, the minority-friendly party Harmony Center (Saskaņas Centrs) received more votes than any other party (28.36%) but was excluded from the governing coalition because other parties refused to cooperate with Harmony. The electoral campaign ahead of the Saeima elections in October 2014 again focused on the loyalty of minorities to the State of their citizenship, thereby discrediting parties perceived to align with or represent minority interests at the national level. Similarly, following the 2018 elections Harmony was again excluded from coalition negotiations. Indeed, Harmony is often portrayed as a political force of the ‘fifth column’ and has repeatedly been pushed into the opposition, despite its considerable parliamentary representation (Plakans 2009; Lühiste 2006).

We note that, despite the frequent refusals of ethnic Latvian political elites to cooperate with minority-friendly party representatives, most Russian speaking citizens still turn out to cast their ballot in elections. Over the past three decades, an ever-increasing number of Russian speakers have also contested minority dominated constituencies on ‘majority’ party lists (Muiznieks, Rozenvalds, and Birka 2013).

Thus, ethnic-minority participation in electoral politics is increasing, even though their elected representatives are marginalized by other parties once they assume office. This might appear contradictory. Yet, studies of political participation in democratizing ethnically heterogeneous societies suggest otherwise (see Dahlberg, Edmark, and Lundqvist 2012). As the experience of democratic participation grows, the decision of citizens to vote in elections will not be determined by their ethnic attachments (anticipated representativeness of government) but will reflect procedural constraints (the ability of representatives to promote interests). In nationalizing states, minority citizens, therefore, participate because elections legitimize the democratic representation of popular concerns and are central to inclusive policymaking. Citizens who are members of the majority ethnicity, conversely, are likely to return to the ballot box to ensure the continuity of their incumbent co-ethnics and their policies.

The common expectation about divided societies is that ethnonational state building has a negative effect on minorities’ trust in political institutions. Since Latvia’s political representatives are as ethnically divided as the constituents they represent, the elite-mass congruence in relation to preferences of nation-state building should be higher in the ethnic majority part of the citizenry. The regular and orderly transfer of power should, however, allow all voters to correct policy priorities and re-align policymakers’ preferences with their own. Thus, we compare levels of trust across the entire society revealing a more complex relationship between the minority and majority with institutions of the nation-state. We assess the levels of trust in key democratic political institutions – parliament and elections – rather than trust in the democratic system to avoid mistaking the citizenry’s perception of contemporary (and passing) challenges to the political regime for dissatisfaction with democracy in principle.

Studying dejection – low trust in political institutions coupled with electoral disengagement – across the ethnic divide in Latvia allows us to explain the low level of public engagement with the central institution of democracy, parliamentary elections. Previous research on the region (Rose, Berglund, and Munro 2005; Duvold and Berglund 2014) has demonstrated that uneven relationships between the ethnic majorities and minorities are accepted by both groups because of the democratic consolidation in (what has become) the Latvian national state. Linking electoral participation with trust in democratic institutions additionally allows us to explore the salience of ethnonational state-building for political participation across this divided society. Ultimately, every citizen of Latvia – and potentially every resident, should they decide to naturalize – has the ability to contribute to democratic decision-making via the electoral box (see Chandra 2007). Without citizen participation in regular and meaningful elections, as was the case under Communism, political representation grinds to a halt; the quality of the regime remains stable, yet at a low level.

Participation and confidence in democratic institutions

Having discussed the recent scholarly contributions dealing with the impact of nation-state building on political representation and participation in ethnically divided societies, we can now explore whether these are reflected in the contemporary Latvian situation. Latvia’s policies of nation-state building have affected views on democracy, confidence in democratic institutions, and electoral participation across all segments of Latvia’s society. Since this finding is different from the existing academic and public consensus, we propose a typology of voters showing that, despite common assumptions, ethnicity is not a good predictor of electoral engagement and of trust in political institutions.

Before proceeding with the analyses of the reasons for dejection, we must assess whether confidence in Latvia’s democratic institutions is ethnic group specific. This would be in line with evidence from other democratizing multi-ethnic societies in which ethnic identity issues lose political salience over time despite their importance for both minorities and majorities at the start of national state-building. First, we explore the level of trust between Latvia’s residents and their democratic institutions (parliament and politicians), as well as their assessment of the regime’s democratic credentials to ascertain differences between segments of the local population. Second, we determine residents’ attitudes toward formal/party politics to understand voter turnout and define types of Latvian voters. In doing so, we assess whether there is (relevant) potential for bread-and-butter issues to appease voters across the ethnic divide, rather than continuing with ‘pork barrel’ politics.

The survey data from the fourth wave of the European Value Survey (EVS) containing around 1,500 individual responses from a representative sample across Latvia helps us assess the levels of dejection in Latvia’s society.¹ As we can glean from the most recent data (2008) (Table 1), the level of support for democracy is fairly good across Latvian society but compared to people in the ethnic majority, minorities show a statistically significantly lower level of satisfaction with the democratic system.

Table 1. Evaluation of the democratic system in Latvia

	Very good	Fairly good	Fairly bad	Very bad	TOTAL
Latvians	131 (15.4%)	557 (65.5%)	136 (16.0%)	26 (3.1%)	850 (100.0%)
Russian speakers	64 (13.6%)	280 (59.6%)	100 (21.3%)	26 (5.5%)	470 (100.0%)

Question asked: ‘For each one, would you say it is a very good, fairly good, fairly bad, or very bad way of governing this country? Having a democratic political system (Q66).’ Source: European Values Study (2016).

We find a similar trend across other available datasets, like New Baltic Barometer, European Social Survey, and World Values Survey. To gain a more granular view, we turn to group-specific differences for levels of trust/confidence in sets of political institutions associated with democracies. The European Social Survey (ESS) round four (held in 2009 in Latvia) offers the most comprehensive set of information. The EES asked Latvia’s residents to report their confidence in democratic institutions (politicians, parliament) and can be broken down by language groups (see Table 2). It is interesting, however, that the difference between Latvian and Russian speaking segments of the sample is not statistically significant either for politicians or for parliament. In part, this is due to the overall low levels of confidence in these institutions by representatives of both groups (all show nearly ‘no confidence at all’).

Table 2. Trust in and satisfaction with the democratic system and institutions

	Latvians	Russian speakers	T-test for mean score indifference (p value)
Satisfaction with Democracy (N = 1877)	3.44 (.062)	2.67 (.096)	.000
Confidence in National Parliament (N = 1945)	1.99 (.055)	1.88 (.096)	.275
Confidence in Politicians (N = 1942)	1.65 (.053)	1.79 (.089)	.175

The table presents average scores on a scale from 0 to 10; zero meaning not confident/satisfied at all, and ten meaning complete confidence/satisfaction. Standard Error is noted in brackets. Questions asked: ‘How satisfied are you with the way democracy works in Latvia? (B27)’ and ‘How much do you personally trust each of these institutions: Latvia’s parliament? (B4)’ and ‘Politicians? (B7).’ Source: European Social Survey (2008).

If anything, these results show that Russian speakers in Latvia have a lower level of trust in the democratic system, yet there is no intergroup difference in trust in democratic institutions (parliament) or actors (politicians). In fact, it appears that Russian speakers have no trust in the political institutions of the state established to cater to the interests of the ethnic majority, Latvians (see also, Duina and Miani 2015). Equally, it is perhaps unsurprising that Russian speakers distrust political actors, if they are ethnic Latvians, given the lengths to which policymakers have gone to craft nationalizing policies that affect minority residents (Zepa and Šūpule 2006; Csergő and Regelmann 2017; Braghiroli and Petsinis 2019). Yet, it is astonishing that Latvian speakers seem to share similar levels of skepticism about ‘their’ institutions and ‘their’ representatives.

The crucial distinction here is not in both groups’ lack of confidence in the institutions and actors that underpin democratic politics, but in the different understandings of democracy in both groups. In their comparative study of democratic consolidation in the post-Soviet area, Mishler and Rose (1999) suggest that post-Communist citizens view political regimes that they experience after Communism as an ‘ideal-typical democracy’ because they have known no other example of democracy. Latvia’s residents have no (personal) experience of any other democracy other than the post-Soviet one. Therefore, Latvia’s Russian speakers are inclined to believe that democratic regimes are similar to the system of ethnic ascendancy that have been in place since the demise of the Soviet Union (Smooha and Järve 2005). Using the same logic, Latvian speakers have a more positive view of the current regime as a representative democracy ‘for people like ourselves.’ Their experience is that Latvia’s democratic instruments work and cater to the majority ethnic group, and the political regime and the system of institutions upholding it are not designed to account for the interests of ethnic ‘others.’

Thus, this mirrors the views of the Russian-speaking minority in Latvia. They demonstrate the same extremely low level of confidence in parliament and political elites as their Latvian peers. Still, they have significantly lower appreciation for democracy because the only democratic system they are familiar with is the one that serves the majority group. Russian speakers can be certain that the system is working, but not designed to do so to their advantage. We found evidence to support this by removing the non-citizens from the pool of Russian speakers in this analysis (265 out of the overall sample of 1,501). As non-participants in the political process, we expect these respondents to be more skeptical about both institutions and elected representatives, regardless of this group’s assessment of Latvia’s democracy. We find far lower levels of statistical significance of satisfaction with democracy in principle, but a more robust convergence of distrust in parliament and politicians in particular. So, overall, it is fair to say that Latvians seem to acknowledge that democracy works well for Latvians; Russians too, agree that democracy works well for Latvians.

Since we focus on the effects of the continuous emphasis of nation-state building on political dejection in Latvia’s society and since ethnic communities show no difference in trust in Latvia’s parliament or in its elected representatives, we need to establish how levels of confidence translate into voting intentions and whether these patterns are ethnically distinct. Using the EVS, we included both Latvian and Russian speakers into the matrix of four categories, depending on their trust in parliament (or not) and their intention to vote (or not) (Table 3). We omit responses from non-citizens because they are ‘not eligible’ to participate in elections.

Table 3. Types of Voters (N = 1182)

		How much confidence in Parliament? (Q63G)
		High (great deal, quite a lot)	Low (not very much, none at all)
Vote Intention? (Q75)	Yes	Confident Voter 187 (15.8%) Latvian 151 (16.85%) Russian speaker 36 (12.59%)	Objecting Voter 689 (58.3%) Latvian 509 (56.81%) Russian speaker 180 (62.94%)
	No	Complacent Abstainer 51 (4.3%) Latvian 40 (4.46%) Russian speaker 11 (3.85%)	Dejected Abstainer 255 (21.6%) Latvian 196 (21.88%) Russian speaker 59 (20.63%)

Questions asked: ‘How much confidence do you have in them: is it a great deal, quite a lot, not very much or none at all? Parliament (Q63G)’ and ‘If there were to be a general election tomorrow, would you vote? (Q75).’ Source: European Values Study (2016).

Along with the two dichotomously ranked categories – trust in parliament and intention to vote – we can plot clearly distinct categories of political engagement among Latvia’s residents. We label them in the following way – confident with the Saeima and intent to vote (Confident Voters, 15.8%); not confident in the work of the Saeima, but willing to make their uneasiness heard at the ballot box (Objecting Voters, 58.3%); those who have neither confidence in the parliament nor plan to participate in elections (Dejected Abstainer, 21.6%); and finally those who are confident in the parliament but do not intend to exercise their right to vote (Complacent, 4.3%).

It is worth noting that over half of all respondents (58.3%) fall into the Objecting Voters category, that is, they would vote despite low-to-no trust in parliament: 56.81% of Latvians fall into this group, slightly fewer than Russian speakers (62.94%) (See Table 3). Statistically, these are not significant differences. These citizens might exercise their right to oust parliamentarians from the Saeima as much as to bring new ones in; arguably, these citizens constitute a ready-to-use support base for new political parties which have regularly emerged in Latvia (Auers 2003; Galbreath and Auers 2009). Importantly, the sheer size of those who intend to vote despite showing limited confidence in the parliament suggests that these voters use the opportunity given by the elections to simply ‘throw out the bums’ (Pop-Eleches 2010).

In short, notwithstanding the challenges of reconciling nation-state building with ethnic divisions in society, 74.1% of Latvia’s voters – across ethnic communities – clearly trust democratic elections to eject inadequate representatives from office, rather than as an avenue to translate their preferences into policy. Citizens do not challenge the ethnonational status quo and are voting against, rather than for, candidates. Crucially, this indicates that the usual focus on how democratic political structures (and strictures) of the nation-state affect minority political (dis-)engagement misses the point. There is a clear consensus across both ethnic groups in Latvia that while elections make sense, voters’ concerns are not represented satisfactorily by the elected parliamentarians.

The Dejected group in our sample includes 21.88% of Latvian respondents, nearly the same proportion as Russian speakers, 20.63%. It seems ironic that the policies of the post-Soviet state have aimed to integrate Russian speakers into mainstream Latvian culture and society, but these same efforts have had a devastating effect on the Latvian majority’s appreciation for electoral politics in their ‘own’ state. The scholarship identifying political disengagement as a feature of the Russian speaking part of the Latvian citizenry (for example, Cheskin 2012; Ehin 2007) appears to have overlooked the fact that minority identity does not explain dejection. The suggestion of a rift between ethnic groups in Latvia is, at the very least, overstated. Political dejection affects all parts of Latvia’s society. Since there is no statistical difference in participation or levels of confidence in democratic institutions between ethnic groups, more attention is warranted for the group of nonvoters. Identifying the causes of political dejection would bring us closer to understanding what voters view as challenges to better quality policymaking in Latvia. The ethnic dimension per se does not explain respondents’ withdrawal from the democratic political process; in the following section we explore what does.

Measurement and results

As we have established above, close to three quarters of Latvia’s voters expressed their intention to participate in elections to ensure an orderly exchange of political representatives in national office. At the very least, this means that democracy is the only game in town and EU accession was not just a temporary departure from Communism (Vachudova and Hooghe 2009). Almost 90% of voters declare low-to-no confidence in parliament and to identify the cause of their dejection, we contrast the two types of Latvian voters that fall into this group: those who do not plan to participate in the parliamentary elections and those who do. We have identified these groups as Objecting Voters and Dejected Abstainers, and we are satisfied that the differences between Latvians and Russian speakers in these groups are not statistically significant. We will explore why low-to-no trust in parliament is so prevalent and identify the underlying problems they see with representation in Latvia’s democracy.

Latvia’s citizens have only experienced democracy in its current form. The process of democratic consolidation in Latvia, as across the post-Communist region, involved several interlocking institutional transformations, including state-building, democratization, development of a market economy, and nation-state building (Offe 1991). Though ultimately successful, the democratic consolidation was far from smooth. The wide-ranging de-regulation of the market, the rescinding of state responsibility for citizens’ welfare, and commitment to nation-state building in Latvia, as well as in Estonia, have consolidated ‘two societies’ – the haves and have-nots (Taagepera 2006, 79). Maintaining stable liberalization in political, economic, and societal spheres was largely possible thanks to ethnicity-based inclusion of the citizenry (Agarin 2013); this added the key fourth element to transition: ethnocentric politics. Indeed, as Kuzio argues (2001), the fusion of institutional processes of state-building, market liberalization, and social change with state-sponsored nation-building merits a particular focus as a key determinant of political support.

As outlined above, Latvia’s political representatives have continuously prioritized nation-state building over other socially relevant and politically salient issues. Consequently, the electorate’s confidence in the ability of their representatives to translate popular concerns into policy has plummeted. In the past, citizens’ endorsement of democratic principles (Duvold and Berglund 2014), a market economy (Cianetti 2018), individual liberties (Cianetti and Nakai 2017), and ethnonational state-building (Ījabs 2016; Kaprāns and Mieriņa 2019) have been used to predict political participation individually, but not in combination with one another. In the analysis that follows, we draw these discrete observations together. Since at the ballot box and in surveys, citizens express trust in democratic political institutions and evaluate opportunities to participate in elections simultaneously with considering the effects of political, economic, and societal circumstances simultaneously, the researchers analyzing their responses should also do so. We consider whether citizens trust democratic political institutions and assess elections as opportunities to promote political, economic, and social reforms.

The impact of the four transitions – state-building, market liberalization, social change, and nation-building – has been cumulative and affected Latvia’s entire society. By reflecting on responses to three statements available in EVS on each of the elements of the ‘quadruple transition,’ we seek a more specific assessment of the impact of these factors on political dejection (see the Appendix 1 for the operationalization and the results of the factor analysis). The EVS contains multiple indicators about citizen evaluations of political, economic, societal, and cultural change (for example, respondents’ degree of agreement with the following statements: ‘Democracies aren’t good at maintaining order’ and ‘Democracy may have problems, but it’s better than any other form of government’). The technique of factor analysis integrates these overlapping attitudes into attitudinal dimensions, allowing us to assess respondents’ individual opinions about democratic principles, economic policy, civil liberties, and nation-state building. These attitudes about the four transitions are essentially background determinants of political behavior. They inform the reasons why voters go (or do not go) to the ballot box. Additionally, to explain the likelihood of citizen dejection from politics, we draw on standard demographic variables (age, gender, education, and income) available in the EVS, as well as ethnic background (whether Latvian or Russian speakers).

In the following, we use multiple logistic regression to test the effects of the ‘quadruple transition’ on voter turnout in the group that has low-to-no trust in parliament and to identify factors that significantly determine the likelihood of dejection (Table 4). We calculate the odds ratio, whereby reference above 1.0 reflects a positive impact and values below 1.0 show a negative impact on the dependent variable. We use the standardized odds ratio to report the impact of these variables, and then we compare them to state that a larger (or smaller) odds ratio has a larger (or smaller) impact over other parameters.²

Table 4. Explanatory factors for voter objection and dejection

sample group	All				Latvian				Russian						All
model	1				2				3						4
categories	Objection		Dejected		Objection		Dejected		Objection		Dejected			Objection			Dejected
Age (10 years)	1.339	**	.791		1.389	**	.748	*	1.343		1.123			1.305		**	.791
Education (final year)	1.042		.798		.892		.648	**	1.525	*	1.511			1.030			.780	*
Income (dodecile scale)	1.116		1.060		1.175		1.076		.986		1.011			1.108			1.109
Gender (female)	1.057		.885		1.091		.892		.948		.842			1.049			.896
Russian speakers	.924		.607	***										.912			.606	***
Opposition to Ethnonational state building	1.424	**	1.812	***	1.220		1.530	**	2.000	***	2.695	***
Nostalgia for state-owned economy	1.258	**	1.174		1.281	*	1.178		1.259		1.310
Anger against neoliberal individualism	1.036		1.322		1.113		1.454	**	.818		.930
Skeptical of democratic principles	.986		.979		.919		.937		1.203		1.160
individual component consisting of four factors above
Respect for Latvian political institutions is not important														1.053			1.521	***
Having Latvian ancestry is not important														1.338		**	.898
Ability to speak Latvian is not important														1.140			1.607	***
State’s responsibility to provide														1.116			1.248
State controls firms more effectively														1.089			.987
Government ownership of business														1.170			1.051
Right to refuse job when unemployed														1.113			1.350	**
Competition is harmful for people														.862			1.039
Incentives for individual effort														1.053			.975
Democracy is not the best political system														.899			.900
Democracy is not indecisive														1.013			1.035
Democracy can maintain order														.893			.903
N	707				534				173						707

Dependent variable: Objecting Voter (1), Dejected Abstainer (2), or others (0) in our analysis: Logistic Regression Analysis, with standardized odds ratio and robust standard error are reported.

Note on significance levels: *** p < .01, ** p < .05, * p <,10.

In our baseline model, we include the cumulated individual-level effects of the ‘quadruple transition,’ in addition to standard demographics, to identify the reasons for political disengagement across Latvia (Model 1). The baseline model is needed to evaluate the validity of our argument for each ethnic group. Common or different effects of each parameter on the likelihood of respondent dejection by ethnic groups give us important information for interpreting the results. Model 2 focuses only on Latvian respondents, while Model 3 focuses on Russian speakers. If we find common effects between Models 2 and 3, it will tell us that some phenomena occur cross-ethnically. Model 4 offers the robustness check of our argument and guides a more nuanced interpretation of the impact that individual-level effects of ‘quadruple transition’ have on the likelihood of dejection in society.

Results

Our primary interest is in the reasons for voter dejection, but as we noted earlier, a significant proportion of Latvia’s voters fall into the Objecting Voters category. While ‘Dejected Abstainers’ and ‘Objecting Voters’ represent – at least in our working definition – two faces of disenchantment with democratic institutions, we run the statistical analysis of both to assess the probability with which respondents are likely to fall into either category. We have run a multinomial logistic regression to contrast the reasons for disengagement among Objecting Voters and Dejected Abstainers.

While demographics such as younger age and lower levels of income have a major impact on the probability of electoral abstention, being a Russian speaker, in and of itself, does not have a demonstrable effect on withdrawal from political participation. Rather, identifying as a Russian speaker increases the likelihood of political participation (decreasing the probability of nonparticipation, as we discuss below).

When we include opinions about the effects of transition, identity as a Russian speaker plays a role in explaining levels of dejection (Model 1). Opposition to societal competition and to ethnonational state-building, however, have overriding effects on the likelihood of political disengagement. Among the respondents’ preferences that play a role, neither skepticism about democratic principles, nor disagreement with neoliberal individualism, nor concerns about the de-regulated economy are statistically significant for one’s likelihood of alienation from politics. As we see in Model 1, however, when we include the attitudes about the importance of ethnonational policies for state-building alongside other factors, the probability of dejection from the democratic process increases perceptibly. Thus, neither the governments’ economic policies nor skepticism about democracy contribute as much to citizens’ withdrawal from electoral participation and low-to-no trust in parliament as does opposition to nation-state building.

These trends are pronounced equally among the Latvian and the Russian speaking segments of the sample (Models 2 and 3). Opposition to nation-state building explains the likelihood for political dejection across both groups, as well as the likelihood for electoral participation in the minority part of the sample (our category of Objecting Voter). In other words, if someone opposes ethnonational state-building, they are very likely to have low-to-no trust in the parliament, regardless of whether they vote or not.

Exploring the data from Model 1, we plotted the correlation between respondents’ views on ethnocentric nation-state building and the likelihood of their dejection from politics (Figure 1): Those identifying as Latvians are more likely to withdraw from political participation than Russian speaking respondents. Moreover, the steeper slope in the figure indicates that individuals more critical of nationalizing policies are more likely to abstain from electoral politics. In inter-ethnic comparison, this trend appears stronger among the Latvian, rather than the Russian speaking parts of the sample.

Figure 1. Opposition to ethnonational state building in dejected abstainers and objecting voters, by ethnic group

Note: The solid lines indicate the fit line for Latvian and the dot lines for Russian speakers.

Most importantly, the perception that ‘Latvian state-building’ is ethnocentric (enshrined in national institutions, mechanisms of representation, and government formation) alienates not only Russian speakers but also those who self-identify as Latvian speakers (Models 2 and 3, also Figure 1). Not only do we witness the same trend across the entire population, but we also see that Latvian speakers who are young, lower educated, and opposed to competition-based society tend to be more dejected from contemporary politics. If we limit the analysis to Russian speakers, the attitude toward Latvia’s nationalizing policies features as a somewhat important category to explain participation in contemporary politics. The more members of this group oppose ethnonational state-building, the more likely they are to fall into the group of Dejected, rather than Objecting, voters. This reflects Brosig’s conclusion (2010) that the ethnonational consolidation of Baltic politics incentivized Russian speakers to view elections as one of the democratic means to ensure their acquiescence to the majority’s nation-state building and to discourage their participation.

These results should be surprising and alarming for the nation-state builders. It seems that the ethnic background of respondents does matter for the levels of dejection, but it is the Latvians who are more adversely affected. We do find that Russian speakers tend to object to ethnonational state-building, but they do so less consistently than their Latvian counterparts. Holding other factors constant, we find that the probability of dejection from politics is lower for the Russian speakers than for Latvians. The opposite can be said about the political participation of those Objecting Voters who are Russian speakers. Their attitudes about nation-state building matter only slightly (Model 3). If Latvians oppose the nationalizing policies of the state, however, they seem to withdraw from political participation altogether (Model 2).

What distinguishes the dejected from the objecting citizens in their attitudes toward the effects of Latvia’s democratic transition? We have checked the robustness of our analysis by inputting into Model 4 the individual parameters used to construct the four factors: objection to democratic principles, objection to economic liberalization, appreciation for individual liberties, and endorsement of nation-state building (Model 4). The results are similar to the ones discussed above, yet they offer further nuances. Across the sample, two of the three items indicating objection to ethnonational state-building weigh heaviest when explaining the levels of political dissatisfaction. Those who believe that a ‘true Latvian’ does not need to respect Latvian political institutions or be fluent in the Latvian language are more likely to be dejected. These items increase the likelihood of withdrawing from political participation. It is, therefore, evident that the political preoccupation with building an ethnonational state for Latvians (with an emphasis on loyalty and language fluency) is the immediate cause for citizen dejection from politics.

The take-home message from this appears to be that the Russian speakers are not the only ones alienated by the current political climate in contemporary Latvia. The exclusive nature of the formal political process has alienated members of the Latvian majority, especially those critical of contemporary Latvian ethnonational state-building codified in the national narrative. The more emphasis there is on a national language and loyalty to the state, the more dejected a person will be from the contemporary political process. At the time of the survey, these individuals constituted a significant group of around 21% of the (voting) population. Views of this electorate are not a mere statistical happenstance and should be addressed to regain trust in democratic representation in Latvia.

One of the principal criticisms we anticipate is that by using the most recent survey data available (the EVS and ESS data sets), we rely on evidence that predates some of the tectonic changes in the social and political landscapes of the country. The period that followed was marked by deep social and economic ruptures from the global financial crisis; the crisis is still acutely felt in Latvia today.³ And though our data might appear somewhat dated, we do have secondhand evidence from the general elections held in 2010, 2011, 2014, and 2018 that give credence to the dynamics we sketched based on quantitative data. We explore this evidence in the following section of this article.

The persistence of dejection: evidence from the Saeima elections

If a ‘nationalizing state’ is one in which minorities are systematically excluded from politics in their host countries, then Latvia easily qualifies (Brubaker 1996). For such states, the relationship between the dominant culture and state institutions is sacrosanct. This explains why Latvia’s efforts of integrating ethnic minorities into the political community did not seek accommodation of their concerns and ostracized them from active participation in electoral politics, as voters and as representatives. Leaving ethnic minorities outside the political process allowed political elites to form alliances and coalesce rhetoric around nation-state building. But the exclusive nature of the formal political process has alienated members of the Latvian majority, especially those critical of the ethnonational state-building codified in the narrative of the statehood.

In this section we offer further evidence that both the concern with ethnonational state-building and cross-ethnic dejection have been consistent features of Latvian politics. Our discussion illustrates the effects of ethnocentric politics on voter preferences at the ballot box and its effects on political volatility. Here, we witness voter defection from incumbent parties at a time when overall electoral turnout has declined and minority voters have consolidated around the opposition Harmony party. The recent Saeima elections demonstrate that Latvian voters are unhappy about the incumbents and vote for newcomers when they have a chance, while Russian speaking voters mostly keep faith in Harmony. This does not mean to prove that Latvian speaking voters are oblivious to their representatives’ preoccupation with ethnonational issues. Rather, despite the growing dejection from politics, they appear reluctant to question and act upon the effects of nation-state building and ethnonational politics on society as a whole.

We also see that following all these elections, the highly fragmented representation of the ethnic Latvian vote repeatedly built a ‘cordon sanitaire’ around the minority-friendly Harmony in an attempt to form (ultimately unstable) government coalitions committed to ethnic politics. The resultant policy process focused on ethnopolitical issues and has significantly undermined the ability of political elites to address other socially and politically relevant issues. This is consistent with the evidence that the transmission of popular interests into party programs and government policies has been disrupted by ethnic majority voters and their representatives alike. Both prefer not to challenge the ethnopolitical status quo in elections and, as we have witnessed over electoral cycles, are complacent about the wider effects of nationalizing politics on Latvia’s society.

Elections are key moments for democracies. They allow citizens, in regular and orderly fashion, to affirm or correct policy preferences made by their elected representatives; in doing so, voters concretize or shift government priorities. Over the past decade, we notice several electoral patterns in Latvia’s electoral politics. First, voter turnout has dropped continuously over the entire period. Second, the parties of the previous government were repeatedly defeated, while the opposition, a minority-friendly party became ever more consolidated and successful. Third, since 2008, we have seen new parties contest seats and enter the Saeima. In the earlier part of this article, we established that political dejection was widespread in 2009, when policies of nation-state building had been prioritized over other social and political concerns. Since the dominance of ethnic issues in politics persists, this section reflects on the long-term consequences that the focus on ethnonational issues has had on electoral outcomes, political elites’ fortunes, and interest-group representation in Latvia.

In democratizing contexts, the outcomes of general elections after ‘two different ideological camps have had a significant shot at governing’ (Pop-Eleches 2010, 233) are an indication of whether (or not) citizens believe electoral democracy ensures representation of their interests. Pop-Eleches observes that by the 1998 Saeima election the ‘menu of untried mainstream alternatives to the incumbent parties/coalitions’ (2010, 236) was considerably reduced in Latvia; and that the 1998 and subsequent 2002 and 2006 parliamentary elections have seen the successful performance of ‘new/centrist populist parties,’ for example, the People’s Party (Tautas partija). Similarly, Hanley and Sikk (2016, 531) suggest that anti-political resentment might have crystalized following a repeated vote for what materialized to be unsuccessful alternatives – first in opposition to Communism, then to those in government. While Hanley and Sikk surveyed the breakthrough of eastern Europe’s ‘anti-establishment reform parties,’ they only mention the Zalters’ Reform Party in 2011 (Hanley and Sikk 2016). Bågenholm (2013) notes the unique performance of the New Era Party as a successful ‘anti-corruption party’ over three electoral rounds (up until the 2010 elections).

The studies by Pop-Eleches, Bågenholm, Hanley, and Sikk use cross-country comparisons to identify sources of voter disillusionment with the established parties. Despite the different types of post-Communist political organizations at the center of each study, all these studies identify the population’s concern over high levels of political corruption, inconclusive parliamentary decision-making, and the overall immobility of the party system as factors facilitating voters’ choice of new parties. Pop-Eleches and other authors suggest that voters’ alternation between supporting governing parties and the main opposition to new parties over rounds of post-Communist elections is contributing to the relative instability of the political party system. Critically, this has also prevented the institutionalization of voter representation, essential for the effective conversion of citizens’ concerns into public policy. Pop-Eleches (2010, 257) argues that by endorsing the ‘unorthodox parties,’ voters punish the established political organizations without committing to an ideologically defined political alternative. Our brief overview of electoral outcomes below indicates that Latvia’s political parties repeatedly failed to consider that the dejection of Latvia’s voters on both sides of the ethnonational spectrum was caused by an emphasis on ethnic politics, over concerns relevant across both communities.

Perhaps the most pertinent lesson from the consecutive elections in Latvia speaks to our conclusion based on the data from 2008–2009. There has been a systematic decrease in electoral engagement. Voter turnout in Saeima elections was at 61% in 2006, 62% in 2010, but has since slid gradually to 59.5% in 2011, 58.9% in 2014, and 54.6% in 2018. Our data suggests that, had there been an election in 2009, only 25.9% of the electorate planned to abstain (12.1% less than those who did so on voting day in 2010). The low turnout in all subsequent elections after 2010 supports our claim that abstention has remained a widespread, meaningful type of voter behavior.

The actual turnout on the election day is, of course, influenced by a range of factors, including the perceived importance of representation, access to voting stations (especially relevant for Latvian voters abroad), and even weather conditions (see Kostadinova 2003). Voter perceptions of the effectiveness of participation for achieving desired policy outcomes, however, has long been acknowledged as the most important factor explaining turnout (Balch 1974; Wolfinger and Rosenstone 1980). In our analysis, we have established that cross-ethnic skepticism toward the continuous emphasis on nation-state building has contributed to disengagement from electoral politics. With a high number of political elites consolidating their reputation as representatives of one or other ethnic group and repeatedly running for seats in parliament, persistent levels of dejection and decreasing turnout in elections are expected and logical consequences.

With only one episode of retrenchment in the 2010 Saeima election, we witnessed the consolidation of political dejection in the Latvian context. In and of itself, the 2010 elections might not be an outlier after all (Kott 2017).⁴ Then, the National Alliance collected signatures to force a referendum on the transition of publicly funded schools to Latvian-only education, the issue which galvanized the electoral campaign. The resulting ethnopolitical polarization led to the 2012 referendum on the status of the Russian language in Latvia – the most divisive political event since the country’s EU accession (see Ījabs 2016). Thus, by turning up to vote in 2010 in higher numbers than usual, Latvian voters responded to mobilization by their ethnopolitical representatives, in line with past research indicating that the salience of ethnicity spikes around pivotal political events, such as elections or referenda (Higashijima and Nakai 2016). Despite the modest rise of participation in the 2010 elections, since then there has been a persistent overall decline in political participation across all ethnic groups.

As could be anticipated from the 2009 data, established parties and electoral alliances gradually lost voters’ support. The incumbent members of the ethnic majority coalition in government have repeatedly lost seats in the successive 2011, 2014, and 2018 elections. The two-party coalition established after the 2006 elections between the economically liberal, yet socially conservative, Unity Party (Vienotība, later New Unity, Jaunā Vienotība) and the rurally based conservative Green and Farmers Union (Zaļo un Zemnieku savienība) lost seats; decimated in the 2011 elections, they went from 55 to 33 seat. After the defeat, they again sought a coalition with the conservative National Alliance (Nacionālā apvienība). The 2014 elections took place in the context of the conflict between Russia and Ukraine, Latvia’s successful membership in the Eurozone, and a wave of interethnic Latvian civic patriotism.⁵ Collectively, these events generated the expectation that ‘national’ parties would be relatively more successful. Yet, all three parties saw their share of the vote decline in the subsequent 2018 elections, when they secured a paltry 32 seats between them. This change in voter orientation came despite the relative economic stability during the previous parliamentary period; yet the widespread perceptions of high and rising political corruption (Kuris 2012; Hanley and Sikk 2016) can be clearly linked to the persistence of distrust of the governing parties.

Since the emphasis on nation-state building had not changed since the 2010 elections, the parties on the other side of the main ethnopolitical cleavage retained their electoral appeal. Across the elections of 2011, 2014, and 2018, the most vocal representation of Russian speakers, Harmony, was wholly excluded from government formation talks, despite being the largest party in the Saeima. As we can observe in retrospect, Harmony’s non-inclusion into government coalitions might have encouraged ethnic minority voters to continuously support the party in elections, while saving the party from having to make policy choices unpalatable to their supporters. Consistently criticizing the references to ethnonational state-building in Latvian politics and its societal, economic, and political costs, Harmony has repeatedly provided elected representatives from the majority with good reason to avoid coalition talks (McClelland 2019). As the only main party with no experience in the national government, Harmony has a vocal stance on social and political issues associated with the center left and it consistently appeals to the ethnically Latvian voter. Clearly, Harmony is the main electoral home for those Latvian and non-Latvian voters we have classified as Objecting.

Finally, since the data we have used was collected, new parties have contested seats and entered the Saeima in every subsequent election. The levels of citizen dejection from their (past) political representatives, patterns of increased protest voting, and the success of the ‘genuinely new parties’ (Sikk 2005) have become ever more pronounced as Latvians went to the polls over the entire set of ‘third generation elections.’ In all elections since 1998, Latvian voters have faced an impossible choice between the incumbents and the opposition parties, such as the Latvian Russian Union (Latvijas Krievu Savienība, formerly, For Human Rights in a United Latvia, Par Cilvēka Tiesībām Vienotā Latvijā) or even the minority-friendly Harmony. The ascent of new parties was predictable since Latvia’s voters discarded left of center programmatic options and in every election since 1990 remained consistently reluctant to embrace social democratic alternatives (such as the Latvian Social Democratic Workers’ Party, Latvijas Sociāldemokrātiskā Strādnieku Partija, or the Progressives, Progresīvie). Coming from outside of politics as the representatives of the vaguely defined ‘Latvian people,’ new parties built their constituencies from the ground up, attracted voters frustrated with previously elected politicians but did not seek to challenge the ethnocentric cordon sanitaire around Harmony (Petsinis 2020).

The 2011 elections saw the ‘meteoric rise’ of the Zalters’ Reform Party (Reformu partija), which came second after Harmony with 22 seats; its electoral breakthrough was coordinated by the former president Valdis Zatlers, who had little experience in electoral politics but had a considerable reputation for impartiality and the supposed ability to represent all Latvians. In the 2014 elections, two new parties – For Latvia from the Heart (No Sirds Latvijai) and Latvia Regional Alliance (Latvijas Reģionu Apvienība) – enjoyed considerable electoral success, mainly due to their leaders’ popularity. Prior to the success of their parties, both leaders – the former auditor general Inguna Sudraba and the former professional basketball player Mārtiņš Bondars – had little exposure to ethnopolitical rhetoric or electoral run-offs.

The outcomes of the 2018 Saeima elections further illustrated the impact of dejection on a system with limited ideological alternatives and a consolidated ethnonational state-building. KPV LV (Who Owns the State? Kam Pieder Valsts) of former actor Artuss Kaimiņš received 14.1% of the popular vote after being created just two years earlier, in May 2016. KPV LV has been branded as ‘Latvia’s first populist party’ (LSM 2018), gaining critical momentum in the months preceding the election, which, in part, reflected Kaimiņš’s style of political rhetoric that scored him points with voters determined more to oust the old political class than to bring in outsiders with fresh political convictions. Likewise, the New Conservative Party (Jaunā Konservatīvā Partija) was formed in 2014 as the small personal party of a former minister of justice, Jānis Bordāns. It became rapidly popular after welcoming Juta Strīķe, a well-known former member of the Bureau for Prevention and Combating of Corruption (KNAB) and building upon the party’s previous success in the 2017 municipal elections. Finally, Development/For! (A/P, Attīstībai/Par!) included several ex-ministers and highly visible officials and was formed in 2018 from several splintered liberal parties and movements. Pivoting toward a social-liberal agenda and putting forward Artis Pabriks (a former member of Unity who served as minister of defense, as a member of the European Parliament) A/P! attained success whilst proposing to return to the liberal agenda of the early 2000s. All three parties opted for a ‘clean and new’ image, and, as in the case of A/P!, appeased voters with low-to-no trust in parliament, the Objecting and Dejected in our analysis.

Given the effects of the politics observed above – the repeated dismantlement of governing party coalitions, stable voter retention for the opposition party Harmony, and the electoral success of newcomer parties – it seems that the size of Dejected Abstainers and Objecting Voters, as established by our 2008 data, continues to matter for electoral metrics and feeds the pool of critical swing voters. We have found in the previous section of this article that all factors considered, ethnicity is not a significant predictor of the level of political alienation in Latvia, but ethnocentric politics are.

As our review of Saeima elections shows, the persistence of ethnonational politics has reduced opportunities for inter-ethnic cooperation on issues of social and political relevance for all citizens. More significantly, in prioritizing nation-state building, the elected representatives have gradually decimated viable ideological competition and cutoff intra-ethnic discussions of social and political alternatives to nation-state building. Our analysis does not offer sufficient information to claim that the attitudes of the electorate toward the ethnonational policies of the Latvian state explain the high levels of dejection from politics in Latvia (see Agarin 2013). What we find, however, is that the most distinct factor defining the intention of voters to eject incumbents from their posts seems to lie in the representatives’ over-emphasis on descriptive representation, ‘being rather than doing’ (Mansbridge 1999) and their attention to majority group-specific interests in nation-state building.

Conclusion

Latvia is a consolidated democracy where access to electoral representation is guaranteed to all citizens. Close to three quarters of the citizens believe that elections can effectively ensure the representation of their interests. Even if as many as 16.2% of Latvia’s resident population have no right to participate in any elections, almost 60% of Latvia’s Russian speakers have domestic citizenship and, thus, can elect candidates and be elected to power (Centrālā Statistikas Pārvalde 2018).⁶ Close to 90% of the citizenry, however, do not trust the parliament to effectively represent their interests. Our data suggests that distrust and disengagement are closely linked with opposition to ethnonational state-building within both Russian speaking and Latvian segments of society and that there is no statistical difference in majority and minority levels of confidence in the democratic institutions of the parliament and elections.

Large swaths of the Latvian population showed high levels of opposition to contemporary state-building projects that prioritized ethnic ascendancy over other issues. We have identified that the disengagement of citizens from Latvia’s electoral process does not reflect the ethnic dimension of politics and we were able to pin down opposition to ethnonational state-building as the main source of dejection. This suggests that ethnonational politics have hampered cross-ethnic representation and the implementation of policy solutions that are relevant for the entire society and as such pose a long-term challenge to the consolidation of positive relationships between the voters and their representatives.

As discussed in the last section of this article, the repeated electoral success of new parties is the result of their effective mobilization of voters frustrated by the dominance of ethnopolitical issues over socially pressing concerns that cut across the ethnic divide. Most new parties, however, failed to establish themselves as sustainable political representatives and were held accountable for their lack of progress on socioeconomic and political policies. As such, the repeated electoral breakthrough of new parties further contributed to the erosion of the link between citizens and representatives. This link is essential for the stability of consolidated democracies.

Thus, the erosion of trust between the voters and their representatives was not due to the institutional design of Latvia as a nation-state of (ethnic) Latvians, but rather the result of a political process focused on ethnonational issues. Prior to their election, politicians are often seen as aligned with their voters who oppose the existing preoccupation with ethnopolitics. Once elected, however, representatives are perceived by voters as partisans for the ethnopolitical regime they were mandated to reform.

In our analysis, we have offered insights into the general impact of transition and ethnonational politics on voter participation and electoral representation in ethnically divided societies. Two decades after Linz and Stepan (1996) identified the complex demography and historical experiences of the post-Communist region as challenges for democratization, we point to a contemporary challenge to Latvia’s democratic stability: the weak link between the voters and their representatives. This fragility is attributable to the poor quality of representation in an ethnically divided democracy. That the interests of citizens were not well served during Communist misrule is widely acknowledged; but citizens remain unsure whether their interests are being acted upon and represented in Latvia’s contemporary, post-Communist democracy. Regardless of citizens’ skepticism, those studying post-Communist politics often tend to interpret citizen participation in elections and the regular turnover of ruling elites as evidence of a democracy’s functionality (Rybář and Deegan-Krause 2008; Powell and Tucker 2014). Scholars of electoral politics agree that mechanisms are in place for public opinion to materialize into policy (Levitz and Pop-Eleches 2010) and political outputs, as much as actors’ commitments to safeguard liberal institutions ensure democracy’s stability (Bohle and Greskovits 2012). Since the political institutions and party organizations promoted by Communism have been thoroughly discredited, even low electoral turnout, low satisfaction with democracy, and poor performing elected representatives are rarely believed to challenge democracy (Mishler and Rose 2001; Norris 2011).

Our article suggests otherwise. Limited confidence in democratic institutions and low levels of political participation are evidence of voters’ dejection from politics, which rewards the focus on ethnonational issues. Political representatives’ attention to and promotion of nation-state building policies over other issues is often said to reduce democratic elections to an ethnic head count (Chandra 2007). Yet, as we witness in Latvia, continuous appeal to ethnonationalism for electoral purposes has made all segments of the citizenry skeptical about the ability of elected representatives to attend to citizens’ other concerns. If anything, unstable Latvian electoral politics over the past decades serves as a potent reminder that by prioritizing nation-state building in an ethnically divided society, democratically elected representatives can run the risk of alienating voters and squandering their trust. All elected politicians can occasionally disregard some interests of their voters, but if the exclusion of some residents’ interests is repeated, there are serious consequences for voters’ support of their representatives. If neglect of socially salient and politically relevant issues is justified with reference to identity categories, even the identity-based links between voters and their representatives will erode over time.

Our article invites future scholarship to assess the long-term effects of politics focused on one policy area (over many other policy-relevant interests) on citizen disengagement from political participation in divided societies and, eventually, on the erosion of citizens’ trust in their representatives. Our article makes a strong contribution to the study of ethnically divided democracies, by showing how ethnocentric nation-state building affects not only the minority’s engagement with democratic institutions (as it is broadly expected) but also the majority’s. Without a doubt, evidence from Latvia presented here provides context for understanding citizen dejection in other consolidated ethnically divided democracies where representatives derive their legitimacy from their appeal to identity, rather than from their performance in office or from policy outputs.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Timofey Agarin

Timofey Agarin is Reader in Comparative Politics and Ethnic Conflict at Queen’s University Belfast, where he is also the Director of the Centre for the Study of Ethnic Conflict. His research interests include ethnic politics and their impact on democratic and societal transition, including majority-minority relations, nondiscrimination, migration, and civil society, with a particular focus on post-Communist states in central-eastern Europe. Email: [email protected]

Ryo Nakai

Ryo Nakai is Associate Professor in the Department of Policy Studies at the University of Kitakyushu, Japan. He researches party systems, elections, and ethnopolitics in newly democratized countries, through qualitative and quantitative research on especially Latvia and Estonia. He has investigated how elections and party competitions affect governments’ policies and peoples’ sense of belongings toward nations. Email: [email protected]

Notes

1. Fieldwork was done from 1 June 2008 to 31 October 2008. Over half of the data was collected by 16 August 2008. The exact file version is ZA4800_v4-0-0. The data was retrieved from the GESIS: ZACAT data repository.

2. Technically speaking, it represents the odds of dependent categories’ occurrence as a result of independent variables’ showing 1 standard deviation (z-value).

3. We are grateful to Daunis Auers for pointing this out to us. The 2008 global financial crisis made Latvia the worst hit country in the region. Latvian GDP contracted by around 10% in the last quarter of 2008 and by a further 18% over 2009, driving up unemployment from 7% in December 2008 to 22.8% in December 2009. The analysis presented is based on the data collected by the EVS between 2 April 2009 and 8 September 2009, with over half of the data collected by the end of May 2009.

4. We would like to thank Matthew Kott for drawing our attention to his paper where he extends a similar argument.

5. The construction and the opening of the new National Library building in Riga, the successful performance of Latvia’s team in the 2014 Winter Olympics, concerns about the regional impact of the conflict in Ukraine and the importance of the topic in the European Parliament elections, as well as the preparations for Latvia’s six-month rotating Presidency of the European Union in the first half of 2015 have all been separate events. Yet, all of these gave a sense of common purpose to all ethnic communities, were imbued with high symbolism, and generated a sentiment of interethnic accord and of joint belonging to ‘European Latvia’ (see also, Kaprāns and Mieriņa 2019).

6. This figure had risen to 64% by 2018, according to the Latvian Population and Statistic Bureau.

Appendix 1.

Results of the factor analysis

To test the effects of the quadruple transition – state-building, democratization, market economy, and nation-building (Kuzio 2001), we have computed four factors after reversing the original answer scale (Antipathy to Latvia nationalizing policies; Nostalgia for the state-owned economy; Anger against neoliberal individualism; and Skepticism of democratic principles – we dubbed these state-building, democratization, development of market economy, nation-building in our analysis) for our regression analysis from the following items.

Table

Antipathy to Latvia Nationalizing Policies	N = 1445
EigenValues: Factor 1 = 1.72, (Factor 2 = 0.74 – omitted)
To respect Latvia’s political institutions and laws is not important for being considered truly Latvian (1–4)	0.703
To have Latvia’s ancestry is not important for being considered truly Latvian (1–4)	0.748
To be able to speak Latvian is not important for being considered truly Latvian (1–4)	0.819
Nostalgia for State-owned Economy	N = 1333
EigenValues: Factor 1 = 1.55, (Factor 2 = 0.81 – omitted)
The state should take more responsibility to ensure that everyone is provided for (0–10)	0.627
The state should control firms more effectively (0–10)	0.753
Government ownership of business and industry should be increased (0–10)	0.768
Anger against Neoliberal Individualism	N = 1431
EigenValues: Factor 1 = 1.34, (Factor 2 = 0.96 – omitted)
People who are unemployed should have the right to refuse a job they do not want (0–10)	0.788
Competition is harmful, it brings out the worst in people (0–10)	0.510
There should be greater incentives for individual effort (0–10)	0.675
Skeptical of Democratic Principles	N = 1179
EigenValues: Factor 1 = 1.58, (Factor 22 = .094 – omitted)
Democracy may have problems, but it is better than any other form of government (agree 1–4 disagree)	0.436
Democracies are indecisive and have too much squabbling (agree 1–4 disagree)	−0.804
Democracies are not good at maintaining order (agree 1–4 disagree)	−0.863

The statistical software (Stata 14) automatically drops factors which have an eigenvalue less than 1 (meaning that it has no substantive explanatory power as a factor).