CC BY
23
REGIONAL TRENDS IN ELECTORAL SUPPORT FOR LATVIAN PARTIES: THE NEIGHBOURHOOD EFFECT
L. S. Zhirnova
MGIMO-University
76 Vernadskogo St., Moscow, 119454, Russia
Received 22.06.2021 doi: 10.5922/2079-8555-2022-1© Zhirnova, L. S., 2022
The article analyses the neighbourhood effect in the voting behaviour of the Latvians at the four recent parliamentary elections, the ethnic and national leaning of parties considered. The study expands a set of electoral geography tools by adding modern techniques of spatial analysis as well as by increasing the knowledge on the position of the Russian speakers within Latvia's political party landscape. The research aims to evaluate the role of the neighbourhood effect at Latvian elections and identify stable spatial voting clusters. The degree of spatial autocorrelation and changes in it were analysed for each parliamentary party and the non-parliamentary but still influential Latvian Russian Union (LRU). Statistically significant spatial clusters of high and low support were identified and compared; their steadiness over the study period was examined. The structure of these clusters is generally the same for the 'Russian' parties (Harmony and the LRU), whilst the 'Latvian' parties are characterized by greater spatial diversity. The analysis shows that regions bordering on Russia have clear spatial clusters where election results correspond to the parties' attitudes towards Russian speakers and the Russian Federation. The 'Russian'parties and those more or less favourably disposed to Russian speakers (For a Good Latvia, For Latvia from the Heart) have clusters of high support in the area and the 'Latvian' parties of low. This pattern, however, may be due to the high proportion of the non-Latvian population in Latgale (a region with strong historical connections with Russia) and the character of the development of the border area, rather than to the proximity to the Russian border.
Keywords:
spatial analysis, electoral geography, Latvia, Latgale, Russians parties, ethnolinguistic cleavage, election
The ethnolinguistic cleavage in Latvian electoral politics has repeatedly drawn the attention of researchers from various countries. At the same time, modern spatial analysis methods offer a new approach to this issue based on relatively large amounts of data and visually compelling cartographic representation.
To cite this article: Zhirnova, L. S. 2022, Regional trends in electoral support for Latvian parties: the neighbourhood effect, Balt. Reg., Vol. 14, no. 1, p. 138-158. doi: 10.5922/2079-8555-2022-1-9.
BALTIC REGION ► 2022 ► Vol.14 ► №1
The article is topical as it enlarges the set of tools of electoral geography. Moreover, it focuses on the political landscape of a foreign country with a considerable Russian-speaking population. Thus, the issue is highly relevant for Russian foreign policy prioritizing the support for the compatriots abroad. The study analyses the neighbourhood effect in the last four parliamentary elections employing the methods of spatial analysis that are not so widespread in Russian political science.
The research aims at identifying the spatial structure of the ethnolinguistic cleavage in Latvian elections. To achieve this, the study evaluates the neighbourhood effect in voting for Latvian parties, highlights neighbourhood clusters for each party and tests their sustainability, as well as defines to which extent these clusters reflect the ethnolinguistic cleavage in voting.
A reasonable hypothesis is that the election performance of the so-called "Russian" parties, i. e. Harmony and the Latvian Russian Union, will form sustainable clusters of high support in Latgale and clusters of low support in other regions of Latvia. However, it is interesting to evaluate the success of the attempts of certain "Latvian" parties to overcome the ethnolinguistic fault lines of the political space and gather votes in the regional clusters usually secured by the "Russian" political forces.
Research methods
To evaluate election results, the study employs spatial statistical analysis methods allowing a more thorough cartographic representation of the spatial structure of socio-political processes [1, p. 9] and, in this case, to deepen the understanding of spatial aspects of the ethnolinguistic cleavage in Latvian politics. Back in the 1970s, P. Taylor and R. Johnston [2, p. 265] analysed the neighbourhood effect in electoral behaviour noting that it can play a decisive role in the voting outcome. Modern development of geographic information systems allows testing these suppositions on large amounts of data.
Moran's I index of spatial autocorrelation showing to what extent the results in a region correlate with the results in its neighbouring regions has been rom-puted for each party. The study has employed the method of local indicators of spatial association (LISA) to define statistically significant clusters where high or low electoral support of a party in a region correlates with that in its neighbours [3, p. 94—95]. LISA maps also show "mistakes" where the predictions based on the neighbourhood effect do not coincide with reality. These cases are also relevant for analysis [4, p. 161 — 163, 166 — 168].
The division into five electoral districts (Riga, Vidzeme, Zemgale, Kurzeme and Latgale) does not suit the purposes of the spatial analysis, so the study considers the municipal level with 119 elements. The electoral statistics have been attached to the cartographic base for further analysis using geographic information systems OGIS and GeoDa. We have produced a cartographic base with 119 municipalities, as there was no suitable one available in the open access.
Consequences of the ethnolinguistic cleavage
For more than three decades since the restoration of independence the main feature of the social structure and political landscape in Latvia has been the acute ethnolinguistic cleavage between the Latvian majority and the Russian-speaking minority Although more than one third of the population is Russian-speaking, the parties that represent this group are kept off from forming the government even when they get more votes than all the other political forces.
It is worth noting that not all the Russian-speaking are eligible for political participation as many of them remain aliens (nepilsoni), which is a special status of permanent residents who do not have access to a wide range of rights including the basic political right to vote and be elected. In the early 1990s, the ruling elite of independent Latvia decided to restore the Constitution of 1922 and provide citizenship automatically only for those who had been the citizens of the first Latvian Republic before June 17, 1940, and their descendants. Others, about a third of the permanent population, became aliens. Originally, the status was deemed temporary, however, almost thirty years later every tenth Latvian is an alien2.
As a result, the Russian-speaking account for around 36 %3 of the total population and only for around 27 % of the citizens of Latvia. Consequently, the electorate of parties relying on this ethnolinguistic minority is artificially restricted. At the same time, although these aliens can neither run for electoral office nor vote at any level (in contrast with the non-citizens of Estonia who can vote in local elections), the issue of aliens remains a crucial fault line for Latvian politics.
However, the "red lines" of the Latvian elite that avoids cooperation with "Russian" parties (at least at the national level) wields more influence on the political structure. As a result, after every parliamentary election the will of around a quarter of voters, who repeatedly secure the first place for the Harmony party, is ignored. The situation is aggravated by the convention established over the past twenty years, according to which the ruling majority does not let the opposition considerably influence law-making and policy-making [5, p. 120].
Such a situation has a whole range of negative consequences. Firstly, the marginalized status prevents Russian-speaking political powers from consolidating. It is worth noting that the ambitions of their leaders also play an important role.
1 While we refer to the contradictions between two ethnolinguistic communities, the Russian-speaking and the Latvian-speaking, the use of the latter term in the case of Latvia appears to be excessive as the Latvian-speaking community is comprised almost exceptionally of Latvians, whereas the Russian-speaking are much more ethnically diverse.
2 Iedzivotaju skaits un ipatsvars pec tautibas un valstiskas piederibas gada sakuma // Centrala statistikas parvalde. URL: https://stat.gov.lv/lv/statistikas-temas/iedzivotaji/iedzi-votaju-skaits/tabulas/ire060-iedzivotaju-skaits-un-ipatsvars-pec (accessed 30.09.2021) (in Lat.).
3 60,8 % Latvijas iedzivotaju dzimta valoda ir latviesu // Centrala statistikas parvalde [Электронный ресурс] URL: https://www.csb.gov.lv/lv/statistika/statistikas-temas/ied-zivotaji/meklet-tema/2747-608-latvijas-iedzivotaju-dzimta-valoda-ir-latviesu (accessed 30.09.2021) (in Lat.).
As Ikstens noted [6, p. 51], the situation is aggravated by the fact that these parties will not get a chance to become part of the government until the value gap between Latvians and the Russian-speaking is bridged. Fifteen years later, no significant progress in this regard can be seen.
Secondly, the fact that such a considerable part of the electorate is alienated undermines the stance of Latvian mainstream parties that are forced to resort to serious compromises and form unstable coalitions [7, c. 87]. Over the thirty years of independence, Latvia saw nine parliamentary elections and more than twenty governments. Still, no trend towards stabilization can be seen. After the 2018 elections, the government was formed by five diverse parties united first and foremost by the wish to keep away from the government the unwanted winner of the elections — the Harmony party that relies on the Russian-speaking electorate. It is hardly surprising that two and a half years later half of the cabinet ministers have changed, the coalition expelled one of its partners and was gravely weakened by internal differences. The efficiency of such a government is considerably limited, whereas in times of the pandemic the country needs clear and consistent management.
Thirdly, the trust of people in the political system is gradually eroded by the persistent disregard for the will of around a quarter of voters, when the winner of the elections is repeatedly kept away from the government. Assembling ideological rivals inside of one government, thus preventing them from fulfilling their election programme, has the same effect. As Vorotnikov points out [8, p. 85], the discontent with ruling politicians fosters political apathy that manifests itself in the gradual decrease of the turnout: from 91.2 % in 1993 to 63.12 % in 2010 and 54.6 % in 2018 (Fig. 1).
100
0
1990 1995 2000 2005 2010 2015 2020
Fig. 1. Parliamentary election turnout in Latvia, % 4
4 Saeimas vélésanas / Centrálá vélésanu komisija URL: https://www.cvk.lv/lv/velesanas/ saeimas-velesanas (Accessed 30.09.2021) (in Lat.).
Distrust of the political system fuels new, primarily populist, party projects and aggravates the fragmentation of the political field. However, as Auers states [9], there are other important factors including low requirements for the number of party members and relatively late introduction of state financing for parties in comparison with, for instance, neighbouring Estonia where the extent of institutionalization of the party system is much higher and the field for populist projects is much more limited.
By the classification of Sartori [10, p. 111 — 112], the Latvian party system is a moderately pluralistic one with increasing fragmentation. Suffice it to say that at the moment there are seven parties with four of them forming the government. Moreover, 2021 saw the emergence of a few more party projects that will lay claim to the seats in the Saeima after the elections in 2022.
A high level of fragmentation is reflected in the indexes of the effective number of parties by Laakso and Taagepera (8,4) and Golosov (5,8) (by the results of the 2018 elections) [11, p. 188]. The Golosov index is lower because it allocates less weight to the parties getting considerably fewer votes than the winner of the elections, which is highly topical for Latvia where "Harmony" wins the election by a large margin.
Research overview
Electoral studies are rather popular in political science due to their high practical importance in explaining voting behaviour and predicting the outcome of the elections in future [12, c. 187]. Spatial factors of electoral behaviour have also attracted the attention of researchers. In the early 20th century, one of the founders of political geography A. Siegfried [13] studied the effect of geographic variables on the voting outcome along with economic and socio-cultural factors. The ideas of Lipset and Rokkan [14] about the effect of social group conflicts on the political system are key to electoral studies. The researchers distinguished three types of such cleavages: among classes, among religious groups and between the centre and the periphery.
In their book "The Geography of Elections" Taylor and Johnston [2] set out the theory of social cleavages reflecting the territorial structure of society. The authors pointed out that the neighbourhood effect can significantly influence the voting outcome, however, they could not give a precise measurement of this effect by the instruments available at that time. As Johnston and Pattie wrote later [15, p. 396], it would be an overstatement to say that local context defines the election outcome, but parties can considerably benefit from taking it into account.
Russian researchers have also studied spatial patterns in electoral behaviour. For instance, Turovskii examined different levels of support of left and right parties in urban and rural electoral districts [16]. Akhremenko [17] considered the potential of spatial electoral analysis as a method of political geography. The project of Sidorenko "Electoral Geography 2.0" is worth mentioning as it studies spatial effects in voting in Russia and other countries [18].
Latvian elections and the country's party system regularly attract the attention of Latvian and Russian researchers, with the ethnolinguistic cleavage usually becoming the focus of studies. The novelty of this article lies in using the methods of spatial analysis for studying this issue based on relatively large amounts of data. These methods allow to identify sustainable spatial patterns in voting and enlarge the existing knowledge of electoral behaviour of Latvians in general and of Russian-speaking Latvians in particular.
The abovementioned Ikstens and Balcere [19, p. 258] highlight the fact that in the Latvian socio-political discourse the traditional division into left and right parties usually differs from that in the Western countries as it is defined primarily by the division into so-called "Russian" and "Latvian" parties. "Russian" parties are defined as left or centre-left, whereas "Latvian" ones as right or centre-right.
The same point is made by Vorotnikov [8, p. 85]. He writes that centre-left parties that have stayed in the opposition ever since the independence in Latvia are primarily associated with the pro-Russian orientation (meaning Russia, not the Russian minority), rather than with the alternative socio-economic programme.
It is worth mentioning that it is not only the ethnolinguistic cleavage that defines the division of parties by this principle. Party elites are eager to employ ethnic mottos in the political struggle. As Zepa and Supule show [20, p. 36], the active usage of such rhetoric by politicians remains one of the main catalysts of ethnic tensions in society. The decades that have passed have not changed much in this regard.
Moreover, as the research by Nakai proves [21, p. 214], it is exactly in the run-up to the elections when the nationalist sentiment grows both among the representatives of the ethnic minorities and majority, so the ethnic cleavage is aggravated. As Nakai notes in another article written together with Higashiji-ma [22], as the political system develops, ethnic parties further enhance ethnic identification not only of their voters, but also of other groups that feel threatened by them. Therefore, the researchers believe there is a need for legislature limiting the ability of parties to appeal to concrete ethnic interests, otherwise, ethnic antagonism will only increase and may lead to violent conflicts.
Bennich-Bjorkman and Johansson [23] also explain the persistence of the intense ethnic stand-off in the political system by the inner logic of party interaction. They note that outside politics Latvians and non-Latvians coexist much more peacefully and have many horizontal ties. An opposite example is Estonia where ethnolinguistic communities are more detached, whereas politics are less and less driven by ethnic motives.
Bloom [24, p. 175 — 176] makes an interesting point proving that the attempts by Latvian nationalists to blame the Russian-speaking for the economic crises of 2008 — 2009 failed. The parties dwelling upon the economic agenda without any ethnic claims secured more votes in the subsequent elections.
In the early 2010s, Rozenvalds [25, p. 160] captured a trend towards the "de-sealing" of the Latvian ruling elite and anticipated more access to state power for ethnic minorities, especially taking into account certain distancing of the parties relying on minority votes from Moscow. However, as Ijabs states
[26, p. 308 — 309], after the referendum on the status of the Russian language in 2012, it became clear that these expectations were over-optimistic, as ethno-linguistic fault lines became more evident, and the dominance of the Latvian language and culture was enshrined in the constitution. As geopolitical tensions between Russia and the West escalated over the previous electoral cycles, Latvian political powers continued proving the essential character of the "red lines" against letting Harmony participate in the government, so any "de-sealing" is rather unlikely in the foreseeable future.
As Commercio [27] notes, the Russian-speaking who want to join the Latvian ruling elite can do so only under the conditions defined by the ethnically Latvian establishment. This means that not only do they have to be citizens and master the Latvian language, but they also have to deeply integrate into the Latvian (ethnic) community. For many representatives of the Russian-speaking minority, it is easier to emigrate with a view to finding better labour market conditions. As Ivlev [28] proves, they are much more prone to leaving the country than ethnic Latvians.
Taking into account the previous experience, it is hard to expect that any international institutions will pressure Latvia to ensure the rights of the Russian-speaking population. As Duina and Miani [29] state, Latvia has managed to become a member of the European Union despite the fact that the country has not fully implemented the European legal requirements concerning minority rights protection, even afterwards Brussels did not appear to be particularly eager to pressure the Latvian authorities into fulfilling these obligations.
As consequently "Russian" political forces are kept off the power, and the institution of aliens is preserved, some researchers and Russian-speaking human rights activists believe that in cases of both Latvia and Estonia one can speak of an ethnocracy [30]. However, a more plausible explanation is the late emergence of a nation-state, as this concept of state-building is becoming a thing of the past [31, p. 194]. This view is shared by Oskolkov. He notes that Estonia has to some extent succeeded in moving forward to overcoming the ethnolinguistic cleavage [32, p. 13], in contrast with Latvia where ethnic contradictions remain the major driving force in politics.
Solopenko [33, p. 30] highlights another characteristic of the Latvian political landscape: the ethnolinguistic factor in voting overlaps with the territorial, as the Russian-speaking are settled unevenly around the country and concentrate in large cities whereas Latvians live both in cities and in rural areas (Fig. 1). The exception is rural areas in Latgale next to the borders with Russia and Belarus, where the proportion of the non-Latvian population is traditionally high.
At the same time, as Nemeth and Dovenyi state [34, p. 798], the National Alliance party promoting the idea of a "Latvian Latvia" is more popular in ethnically heterogeneous cities with a high proportion of non-Latvian population rather than in ethnically homogenous municipalities with primarily Latvian population. Ethnic Latvians are inclined to support the ethno-nationalist project of the National Alliance instead of more moderate parties in cities. Rural areas witness less ethnic polarization and less support of the idea of a "Latvian Latvia".
■ <1*0) d : 1.690]
□ 1%-10%C9> [1.690:3]
□ 10%-50% (49) [3:9]
□ 50% - 93% (49) [9:28)
■ 90% - 99% (10) [2S : 49.
■ !>99%(1) [49.860 : 54]
Fig. 2. Percentile map of the share of the Russian population in Latvia, 2020 5
Analysing the electoral geography of Latvia, Lithuania, Estonia, Russia and Ukraine, Meleshevich [35, p. 119] pointed out the regions with a persistent trend towards voting differing from national results. In Latvia, such a region is Latgale, the region next to the border with Russia and Belarus with the prevalence of the Russian-speaking population. Naturally, parties advocating for the rights of the Russian-speaking received most votes here.
At the same time, Janis and Juris Paiders [36] have not managed to find any considerable influence of the closeness of the Russian and the Belorussian border at the 2010 parliamentary elections in Latgale. According to them, the second major factor of the voting behaviour was personal, as certain bright candidates managed to gather substantial support. Another study of these researchers [37] proves that in rural areas the ethnolinguistic composition of the electorate has more influence on voting than in major cities.
Electoral cycles
Over the last four electoral cycles, the number of parties getting into the parliament has grown. After the 2010 and 2011 elections, five parties gained the seats in the Saeima, in 2014 — six parties, in 2018 — seven parties. The number of parties running for the parliament has also increased from 13 in 2010, 2011 and 2014 to 16 in 2018.
The 2010 elections of the 10th Saeima were preceded by the consolidation of the conservative right triggered by the growing popularity of Harmony Centre supported by the Russian-speaking electorate. Three political powers (The New
5 Iedzivotaju skaits un ipatsvars pec tautibas un valstiskas piederibas gada sakuma // Centrala statistikas parvalde URL: https://stat.gov.lv/lv/statistikas-temas/iedzivotaji/ied-zivotaju-skaits/tabulas/ire060-iedzivotaju-skaits-un-ipatsvars-pec (accessed 30.09.2021) (in Lat.).
Era, The Civic Union and The Society for Political Change) created the Unity party. Latvian nationalists also consolidated: three months before the voting, parties "All for Latvia" and For Fatherland and Freedom/LNNK formed an election union called National Alliance. Latvia's First Party/Latvian Way also merged with the veteran of Latvian politics, the People's party, and several regional forces.
The election campaign was defined by the economic crisis and the discussion around the 2009 Riga riot. Unity won the election (31.22 % of votes) and became the main party of power in Latvia for many years. It still remains in power. However, after 2010, Unity has never managed to secure the majority of votes. In the subsequent elections it lost its leading positions to Harmony that was not allowed to get into the government.
The 10th Saeima did not survive even one year. The economy was stalling, the discontent in the society was growing. President Valdis Zatlers had repeatedly advocated for the dissolution of the parliament as far back as 2009. When, in 2011, a freshly elected parliament declined the request of the Public Prosecution office to allow the raid at the house of A. Slesers, the president proclaimed that the Saeima had lost people's trust and initiated the referendum on the dissolution of the Parliament. The initiative got overwhelming support from the electorate with 94.5 % voting for the dissolution.
A new electoral campaign passed under the motto of "fighting the oligarchs", including Slesers. As a result, his party running as Slesers' LPP/LC Reform Party did not make it over the 5 % barrier. Harmony Centre won the elections with 28.37 %, Zatlers' Reform Party came second (20.82 %) due to the popularity of the president who dissolved the Parliament. The prime minister's party Unity faced a sharp decrease in support and came the third (18.83 %). That did not prevent him from getting the post of prime minister again and forming the government leaving the winner of the election behind.
As the 2011 elections were snap, the 11th Saeima worked for three years instead of four. Over this period, the MPs managed to allow dual citizenship and added a preamble to the Constitution (Satvesme) stating that the Latvian state aims to guarantee the existence and development of the Latvian nation, its language and culture. Adding the Preamble was preceded by the failure of the referendum on making Russian the second state language, after which the Latvian ruling elite almost completely stopped taking into account the Russian-speaking minority. When the Constitutional Court considers the cases brought up by the Russian activists, it cites the preamble to Satversme to substantiate the claims that forcing out the Russian language from every level of education in Latvia is constitutional.
A grave tragedy overshadowed the tenure of the 11th Saeima. The collapse of the Maxima trade centre in a Riga district of Zolitude on November 21, 2013, took the lives of 54 people. Prime minister Valdis Dombrovskis (Unity) leading the government since March 2009 assumed political responsibility for the accident and left the post. He was replaced by Laimdota Straujuma from the same party, who retained the post after the elections of 2014.
The election to the 12th Saeima on October 4, 2014, was also won by Harmony, although it got fewer votes than in 2011 (23 %). Unity lost by a small margin and took second place with 21.87 %. It did not have any trouble forming the government with the Union of Greens and Farmers and the National Alliance. However, in 2016 it lost the prime minister's post to the Union of Greens and Farmers as a result of internal intrigues.
The education reform of 2018 marked the work of the 12th Saeima. It put an end to bilingual school education and Russian-language higher education programmes, including private. Moreover, the cabinet of ministers made Latvian the main language of communication in kindergartens regardless of parents' wishes.
At the elections of the 13th Saeima on October 6, 2018, Harmony won again, with an even lower result than before (19.8 %). A new party KPV LV ("Who owns the state?") led by actor Artuss Kaimins was in second place. The parliament turned out to be highly fragmented with seven parties. Given the resolution to keep Harmony away from the government, forming the ruling coalition was extremely hard and took unprecedented 109 days. After several futile attempts by other parties, the coalition was headed by the ex-member of the European Parliament Arturs Krisjanis Karins representing the Unity (that ran at that elections under the brand New Unity consolidating with several regional partners). It is especially striking as New Unity got only 6.69 % of votes, less than any other party that overcame the 5 % barriers. The coalition united five parties that were fierce rivals in the run-up to the election, which guaranteed instability in its work.
Finally, after numerous disputes and scandals in early June 2021, on the very eve of the municipal elections, KPV LV was expelled from the government. By that time, it had lost most MPs and popularity and had become the weakest link in the coalition. The former partners agreed to violate the coalition agreement and redistributed its ministerial posts.
For the Russian-speaking population, the key decision of the 13th Saeima is the law on automatic citizenship for the children of aliens, so no new aliens have appeared since January 1, 2020. However, the main topic of this tenure is the pandemic of COVID-19. At first, the pandemic did not hit Latvia hard, but by the end of 2020, the country had felt all the pressure of the virus, people once again faced harsh restrictions. The situation started to improve only closer to summer 2021. The actions of the government cannot be considered efficient taking into account inconsistent restrictions and the scandal with declining 700,000 dozes of the Pfizer vaccine in December 2020 that brought about the shortage of vaccines. The distrust of the government undermined the trust in the vaccination campaign.
Neighbourhood effect in voting
Let us consider spatial patterns in voting for Latvian parties over the last four election cycles. The neighbourhood has been calculated based on the method of k-nearest neighbours with five neighbours. Five electoral districts do not provide enough instances for analysis, so the study considers results in 119 municipalities.
Let us start with the so-called "Russian" parties: Harmony and the Latvian Russian Union.
Harmony (earlier Harmony Centre) was established in April 2010 by the merger of three political powers: Social Democratic Party of Egils Rutkovskis, New Centre of Sergei Dolgopolov and the National Harmony Party, whose leader Janis Urbanovics was the only member of every Saeima since 1995. Although the leadership of the party has claimed that it relies both on Russians and Latvians, it has been supported primarily by Russian-speaking voters. Over the last decade, it was Harmony that accumulated the majority of Russian votes [33, p. 22]. The initial popularity of the party was based on several factors ranging from the personal popularity of its leader Nil Ushakov to the successful promotion through the First Baltic Channel (and adverse publicity for its rivals). However, the hopes that the party could come to power were crucial for its success. In Riga, these hopes come true whereas at the national level the party remained "the eternal opposition".
Since 2011, the party always gained more votes than other parties. Nevertheless, its results have gradually decreased as some voters are not enthusiastic about voting for "the eternal opposition", others are dissatisfied by the lack of effort to protect the rights of the Russian-speaking population and by certain statements of the party leadership against Russia. However, in the Latvian information space, Harmony has a stable reputation as the «arm of the Kremlin». Therefore, at the national level, any cooperation with it is practically impossible for the Latvian mainstream parties. It is worth noting that at the municipal level, where economic issues are more topical than geopolitics, such cooperation is sometimes possible, for example, in 2013, the representatives of Harmony Centre joined the ruling coalition in Adazi duma with the representatives of the Union of Greens and Farmers, Unity and even National Alliance6.
For ten years, Harmony held power in Riga, effectively placing the administration of the capital, where even officially, one third of the country's population resides, in opposition with the central government. However, since Nil Ushakov was removed from the office of mayor and went on to work in the European Parliament, the party has remained in a deep crisis aggravated by the failure at the snap elections in Riga.
The analysis shows moderate positive spatial autocorrelation of the voting for Harmony at the parliamentary elections (Moran's I in 2010 was 0,578, in 2011 — 0,6, in 2014 — 0,586, in 2018 — 0,584).
Local indicators of spatial association (LISA) show two neighbourhood clusters found in all the four electoral cycles under examination (Fig. 3). One of them is the cluster of high support in Latgale, a region with a high proportion of Russian-speakers. The other is the cluster of low support in Kurzeme, a region with very few Russians. The monolithic structure of this cluster is disrupted by Liepaja and Ventspils, large cities with a significant share of Russian-speakers.
6 Ruska R. Kam pieder atslegas / Latvijas Avize, news.lv. URL: https://news.lv/Latvijas Avize/2013/08/09/Kam-pieder-atslegas (accessed 30.09.2021) (in Lat).
Three more clusters are seen only in some electoral cycles. Two of them are the clusters of low support in Vidzeme where Russian-speakers are also not numerous, the other one is that of high support around the capital.
Fig. 3. LISA maps of the voting results of Harmony at 2010, 2011, 2014 and 2018 elections (left to right, top to bottom)
Note: blue are the clusters of low variable value, red are those of high value, light blue and pink are the regions where the logic of neighbourhood does not apply.
Another "Russian" party, the Latvian Russian Union has not succeeded in securing seats in the parliament over the last four elections although it remains a significant player in Latvian politics. The Latvian Russian Union (before 2014 — For Human Rights in a United Latvia) is one of the oldest Latvian parties. It had its representatives in all the Saeimas from 1993 to 2010, then it lost parliamentary representation. From 2009 to 2020, it could not secure any seats in the Riga City Council. Still, the party leader Tatjana Zdanok successfully ran for the European Parliament four times ensuring international cooperation and high representation for her party. As voters withdrew their support from Harmony, the Latvian Russian Union managed to improve its performance in the 2018 parliamentary elections although it ended up with just 3.2 % of votes and did not pass to the Saeima. Lately, the party has been on the rise inspired by the return to the Riga City Council (6.52 % of votes and 4 members of the council) and the increase in popularity due to a clear and consistent stance on protecting the rights of the Russian-speaking population.
The analysis shows moderate positive spatial autocorrelation in voting for the Latvian Russian Union over the last four electoral cycles with a declining trend (Moran's I in 2010 was 0.446, in 2011 — 0.406, in 2014 — 0.395, in 2018 — 0.362).
The LISA clusters of this party remind those of Harmony and coincide with the proportion of Russian-speakers (Fig. 4). Latgale cluster of high support and Kurzeme cluster of low support (with occasional exceptions of Ventspils and Liepaja) persist over the entire studied period. However, in contrast to Harmony, the cluster of low support in Vidzeme appears in every election cycle, whereas the cluster with Jurmala and several Zemgale municipalities was seen only on the map of 2011.
*
* * iu * % *
Fig. 4. LISA maps of the voting results of Latvian Russian Union at 2010, 2011, 2014 and 2018 elections (left to right, top to bottom)
Note: blue are the clusters of low variable value, red are those of high value, light blue and pink are the regions where the logic of neighbourhood does not apply.
Let us examine the results of "Latvian" parties that participated in all four elections.
As mentioned earlier, Unity (since 2018 running as New Unity with five regional partners) has been the party of the prime minister since its establishment except for the period from 2016 to 2018, when it gave up the leadership in the cabinet to the Union of Greens and Farmers as a result of internal intrigues. In summer 2018, several months before the election, the party rating was around 3 %, nevertheless, it managed to consolidate, find partners and pass to the 13th Saeima, albeit with the smallest number of MPs. However, it was New Unity that eventually succeeded in forming a government, although from the very beginning it was weakened by a diverse coalition and a small fraction of the prime minister's party.
Despite all the difficulties associated with the pandemic and internal disputes that even resulted in expelling one of the coalition partners, the current government of Krisjanis Karins is unlikely to collapse as none of its members wants to take on responsibility for the fall of the cabinet. Taking into account the victory
at the elections to the European Parliament in 2019 (26.24 %) and successful performance at the early elections to the Riga City Council in 2020 (the third place and 15.24 %), New Unity's internal crisis seems to have peaked.
Although Unity has the reputation of the party of bureaucracy, whereas the brand of the main Latvian nationalists is upheld by the National Alliance, over the whole decade in power, Unity has implemented the policy of limiting the rights of national minorities. It was the representatives of this party that forged and put into effect the most severe reforms in this area, including the education one.
The analysis shows a moderate positive spatial correlation of voting for Unity in 2010 and 2011 and weak correlation in 2014 and 2018 (Moran's I in 2010 was 0.056, in 2011 — 0.622, in 2014 — 0.279, in 2018 — 0.173).
LISA maps identify only one cluster that can be seen throughout the period. That is a cluster of low support in Latgale where most non-Latvians live (Fig. 5). However, in 2010 and especially in 2011, this cluster encompassed most of Latgale, while in 2018 there were only three municipalities in the northeast.
A cluster of high support in Vidzeme can be seen throughout the first three elections but not the last one.
f% * *
Fig. 5. LISA maps of the voting results of Unity at 2010, 2011, 2014 and 2018 elections (left to right, top to bottom)
Note: blue are the clusters of low variable value, red are those of high value, light blue and pink are the regions where the logic of neighbourhood does not apply.
Another significant power in the Latvian political arena is the Union of Greens and Farmers established in 2002 by the Latvian Farmers Union and the Green Party. The same year the Union successfully ran for the parliament. Since then,
not only has it enjoyed representation in every parliament but also participated in ruling coalitions except for two — from 2011 to 2014 and at the present moment. The party's positions at the 2011 election were undermined by the anti-oligarchical campaign, as a cooperation partner of the Union, For Latvia and Ventspils, was headed by Aivars Lembergs, one of the three major oligarchs in the country. After 2014, the party returned to the government and even led it from 2016 to 2018. In 2018, the Union lost half of its support, gained only 9.91 % compared to 19.5 % in the previous elections and found itself in the opposition. Nevertheless, the party remained strong at the local level where one way or another it controls around a third of all municipalities.
The party relies on a conservative rural Latvian electorate and subsequently leans toward the slogans of protecting the Latvian nation. At the same time, it has established fruitful cooperation with Harmony in the opposition.
The analysis shows moderate positive spatial correlation (in 2010, Moran's I was 0.534, in 2011 — 0.531, in 2014 — 0.614, in 2018 — 0.332).
Only one cluster can be identified on all four LISA maps. That is the cluster of low support in Riga and the Riga Region (Fig. 6). This reflects the "Riga curse" that has hung over the party since its establishment. The Union has nothing to offer to secure the support from the voters in the capital. The maps of the first three cycles show clusters of high support in Kurzeme, including Ventspils, and a cluster of low support in Latgale.
Fig. 6. LISA maps of the voting results of the Union of Greens and Farmers at 2010, 2011, 2014 and 2018 elections (left to right, top to bottom)
Note: blue are the clusters of low variable value, red are those of high value, light blue and pink are the regions where the logic of neighbourhood does not apply.
The National Alliance "All For Latvia!" — "For Fatherland and Freedom/ LNNK" is the main stronghold of Latvian nationalists, its representatives publicly postulate the aim of creating a "Latvian Latvia". The political force was the part of the government in three out of four studied electoral cycles except for the first one. In 2010, 2011 and 2014, it increased its results at the parliamentary elections up to 16.6 % in 2014, it showed a lower result (11 %) only in 2018. However, just a year later, in the elections to the European Parliament, the party succeeded in proving that it enjoys stable support and sent not one but two representatives to Brussels. Although now the party does not lead the government, many of its suggestions get support from the coalition partners and are enshrined in laws.
The analysis shows a moderate positive spatial correlation in the voting results in 2010, 2011 and 2018 and a high correlation in 2014 (in 2010, Moran's I was 0.521, in 2011 — 0.687, in 2014 — 0.722, in 2018 — 0.562). It is noteworthy that as the National Alliance enjoyed growing support, the neighbourhood effect increased as well, and in 2018 it dropped sharply.
On the LISA maps, a large cluster of low support in Latgale can be seen throughout the examined period. The non-Latvian population of this region apparently cannot accept nationalist slogans (Fig. 7). There is a cluster of high support in Vidzeme, however, it splits in 2014, as well as the cluster of high support in Zemgale that splits in 2018.
* dp
j
' A* '
Fig. 7. LISA maps of the voting results of the National Alliance at 2010, 2011, 2014 and 2018 elections (left to right, top to bottom)
Note: blue are the clusters of low variable value, red are those of high value, light blue and pink are the regions where the logic of neighbourhood does not apply.
Other parties got into the parliament only once over the studied period. After 2010, it was the alliance «For a Good Latvia» (Moran's I — 0.314), after 2011 — Zatlers' Reform Party (Moran's I — 0.383), after 2014 — "For Latvia from the Heart" (Moran's I — 0.377) and the Latvian Regional Alliance (Moran's I — 0.141), after the 2018 elections — "Development/For" (Moran's I — 0.061), the New Conservative Party (Moran's I — 0.495) and KPV LV (Moran's I — 0.646). Thus, except for the Latvian Regional Alliance, the spatial correlation is moderate and positive.
It is worth noting that, for five out of seven of these "Latvian" parties, LISA maps show clusters of low support in Latgale. Only "For a Good Latvia" and "For Latvia from the Heart" have clusters of support in this region with the considerable or prevailing Russian-speaking population (Fig. 8). A possible explanation is that LPP/LC, one of the constituent parts of "For a Good Latvia", had previously posed as a party both for Latvians and non-Latvians. Such an approach seems to be at least partially productive. Moreover, the leader of "For Latvia from the Heart" Inguna Sudraba was almost openly deemed the "arm of the Kremlin" by Latvian politicians and was strongly criticized for the lack of desire to make resolute statements on ethnic matters, which apparently attracted some voters.
Fig. 8. LISA maps of the voting results of the alliance "For a Good Latvia" in 2010 and for the party "For Latvia from the heart" in 2014
Note: blue are the clusters of low variable value, red are those of high value, light blue and pink are the regions where the logic of neighbourhood does not apply.
Prospects and conclusions
The analysis defined the spatial structure of the ethnolinguistic cleavage in the electoral behaviour of Latvians more thoroughly and identified sustainable neighbourhood clusters in voting for different parties. For the "Russian" parties (Harmony and the Latvian Russian Union), the configuration of these clusters is more or less the same. For "Latvian" parties, there is more diversity.
As for the regions neighbouring Russia, the analysis allows defining clear spatial clusters of voting results that correlate with parties' attitudes to Russian-speakers and the Russian Federation. "Russian" parties (Harmony and the Latvian Russian Union), as well as parties showing some sympathy for the Rus-
sian-speaking population ("For a Good Latvia", "For Latvia from the Heart") have clusters of high support in this area, whereas the "Latvian" parties have the clusters of low support. However, this correlation is better explained not by the closeness to the Russian border but by the high proportion of the non-Latvian population in Latgale, which, in turn, stems from close historic ties with Russia and special conditions of development in this border region.
It would be interesting to analyse how the defined trends would manifest in the next elections. However, it is impossible to replicate the research as an administrative reform has reshaped municipal borders.
Another direction for further studies is connected with the fact that this one continues the series of electoral geography research by the Center for Spatial Analysis in International Relations of the Institute for International Studies of MGIMO University. Its materials allow comparing spatial trends in electoral voting in Latvia with other neighbours of Russia [38; 39]. The list includes the countries that Latvia is rarely compared with within political studies, but in this case, the common methodology makes such a comparison possible (for instance, the study of electoral behaviour in Norway: [40]).
The research has been funded by the Ministry of Science and Higher Education of the Russian Federation (the Agreement on the assignment of the grant №075-15-2020-930 from 16.11.2020).
References
1. Okunev, I. Yu. et al. 2020, Atlas mezhdunarodnykh otnoshenii. Prostranstvennyi analiz indikatorov mirovogo razvitiya [Atlas of International Relation: spatial analysis of global development indicators], Moscow, Aspect Press, 447 p. (In Russ.).
2. Taylor, P., Johnston, R. 1979, The Geography of Elections. Harmondsworth, Pen-guinBooks, 528 p.
3. Okunev, I. Yu. 2020, Osnovy prostranstvennogo analiza [Spatial Analysis Fundamentals], Moscow, Aspect Press, 255 p. (In Russ.).
4. Ikstens J. 2014, The democratic role of political parties. In: Rozenvalds, J. How democratic is Latvia? Audit of democracy 2005—2014, p. 115 — 128.
5. Ikstens, J. 2006, Eastern Slavic Political Parties in Latvia. In: Muiznieks, N. (ed.) Latvian-Russian Relations: Domestic and International Dimensions, Riga, p. 41—52.
6. Okunev, I. Yu. Zhirnova, L.S., 2019, Ethnic cleavage in the voting behavior in Latvia, Obozrevatel — Observer, no. 11 (358), p. 78 — 90 (In Russ.).
7. Vorotnikov, V. V. 2018, Party 'Harmony' in the parliamentary elections in Latvia in 2018: victory without a winner, Nauchno-analiticheskii vestnik Instituta Evropy RAN [Scientific and Analytical Herald of IE RAS], no. 5 (5), p. 85 — 90. doi: https://doi. org/1010.15211/vestnikieran520188590 (In Russ.).
8. Auers, D. 2018, Populism and Political Party Institutionalisation in the Three Baltic States of Estonia, Latvia and Lithuania, Fudan J. Hum. Soc. Sci., no. 11, p. 341—355. doi: https://doi.org/ 10.1007/s40647-018-0231-1.
9. Sartori, G. 2005, Parties and Party Systems, Great Britain, ECPR Press, 368 p.
10. Golosov, G. V. 2010, The Effective Number of Parties: A New Approach, Party Politics, vol. 16, no. 2, p. 171-192. doi: https://doi.org/1010.1177/1354068809339538.
11. Meleshkina, I. Iu. 2001, The Research of Electoral Behaviour: Theoretic Models and Issues of their Usage, Politicheskaya nauka [Political Science], no. 2, p. 187 — 212 (In Russ.).
12. Siegfried, A. 1913, Tableau politique de la France de l'Ouest sous la III Republique, P., Colin, XXVIII, 535 p.
13. Lipset, S., Rokkan, M. 1967, Cleavage structures, party system, and voter alignments: An introduction, Party system and voter alignments, N. Y., p. 27—45.
14. Johnston, R., Pattie, Ch. 2017, Local context, social networks and neighborhood effects on voter choice. In: Fisher, J. et al. (eds.) The Routledge Handbook of Elections, Voting Behavior and Public Opinion, Routledge, p. 383—400. doi: https://doi. org/10.4324/9781315712390.
15. Turovskii, R. F. 1996, Political Stratification of Russian Regions (History and Formation Factors), Partiino-politicheskie elity i elektoral'nye protsessy v Rossii. Kruglyi stol biznesa Rossii. Analiticheskie obozreniia Tsentra kompleksnykh sotsial'nykh issledo-vanii i marketinga. Seriia: politologiia [Party and Political Elites and Electoral Processes in Russia. The Round Table of the Russian Business, Analytical Overview of the Center for Complex Social Studies and Marketing. Series: Politology], no. 3 (17), p. 37 — 52 (in Russ).
16. Akhremenko, A. S., 2009, Spatial Electoral Analysis: Characterization of the Method, Possibilities of Cross-National Comparative Studies, Polit. Nauka [Political Science], no. 1, p. 32 — 59 (in Russ).
17. Sidorenko, A. 2007, Latvia, Electoral Geography 2.0, available at: https:// www.electoralgeography.com/new/ru^category/countries/l/latvia (accessed 30.09.2021) (in Russ).
18. Ikstens, J., Balcere, I. 2019, Latvia: Office-seeking in an Ethnically Divided Polity. In: Ilonszki, G., Bergman, T., Muller, W. C. (eds.) Coalition Governance in Central Eastern Europe, Great Britain, Oxford University Press, p. 252—302.
19. Zepa, B., Süpule, I. 2006, Ethnopolitical Tension in Latvia: Factors Facilitating and Impeding Ethnic Accord. In: Muiznieks, N. (ed.) Latvian-Russian Relations: Domestic and International Dimensions, Riga, p. 33—40.
20. Nakai, R. 2018, Does Electoral Proximity Enhance National Pride? Evidence from Monthly Surveys in a Multi-ethnic Society — Latvia, Studies in Ethnicity and Nationalism, p. 198 — 220. doi: https://doi.org/10.1111/sena.12285.
21. Higashijima, M., Nakai, R. 2016, Elections, Ethnic Parties, and Ethnic Identification in New Democracies: Evidence from the Baltic States, Studies in Comparative International Development, no. 51, p. 124—146. doi: https://doi.org/10.1007/s12116-015-9187-1.
22. Bennich-Bjorkman, L., Johansson, K. 2012, Explaining moderation in nationalism: Divergent trajectories of national conservative parties in Estonia and Latvia, Comparative European Politics, no. 10, p. 585 — 607. doi: https://doi.org/10.1057/cep.2011.28.
23. Bloom, S. 2012, Does a Nationalist Card Make for a Weak Hand? Economic Decline and Shared Pain, Political Research Quarterly, vol. 65, no. 1, p. 166 — 178. doi: https://doi.org/10.1177/ 1065912910388187.
24. Rozenvalds, J. 2012, The problem of "(de) sealing" of the political elite of Latvia and Estonia: Russian-speaking minority perspectives, Sravnitel'naja politika [Comparative Politics], no. 3 (9), p. 149 — 161. doi: https://doi.org/10.18611/2221-3279-2012-3-3(9)-149-161 (in Russ.).
25. Ijabs, I. 2016, After the Referendum: Militant Democracy and Nation-Building in Latvia, East European Politics and Societies, vol. 30, no. 2, p. 288 — 314. doi: https://doi. org/10.1177/ 0888325415593630.
26. Commercio, M. E. 2008, Systems of Partial Control: Ethnic Dynamics in Post-Soviet Estonia and Latvia. Studies in Comparative International Development, no. 43, p. 81 — 100. doi: https://doi.org/10.1007/s12116-007-9013-5.
27. Ivlevs, A. 2013, Minorities on the move? Assessing post-enlargement emigration intentions of Latvia's Russian speaking minority, The Annals of Regional Science, no. 51, p. 33 — 52. doi: https://doi.org/10.1007/s00168-012-0534-0.
28. Duina, F., Miani, C. 2015, Fitting in the Baltics: National identity, minorities and compliance with EU accession requirements in Lithuania and Latvia, Comparative European Politics, no. 13, p. 535 — 552. doi: https://doi.org/10.1057/cep.2014.5.
29. Buzaev, V. V., Nikiforov, I. V. 2009, Sovremennaya evropeiskaya etnokratiya: narushenie prav natsional'nykh menshinstv v Estonii i Latvii [Contemporary European ethnocracy: violation of ethnic minorities' rights in Estonia and Latvia], Moscow, 279 p. (in Russ.).
30. Tevdoi-Burmuli, A. I. 2018, Etnopoliticheskaya dinamika Evropeiskogo soyu-za [Ethnopolitical dynamics of the European Union], Moscow, Aspekt-Press, 224 p. (in Russ.).
31. Oskolkov, P. V. 2020, Estonia's party system today: electoral turbulence and changes in ethno-regional patterns, Balt. Reg., vol. 12, no. 1, p. 4—15. doi: https://doi. org/10.5922/2079-8555-2020-1-1.
32. Solopenko, A. V. 2014, Elektoral'noe povedenie zhitelei Latvii na parlamentskikh vyborakh 1993— 2011 gg. [Electoral behavior of Latvians in parliamentary elections in 1993 — 2011], Germany, LAP Lambert Academic Publishing, 52 p. (in Russ.).
33. Nemeth, A., Dovenyi, Z. 2019, Patterns of Ethnic Homogenisation, Fragmentation and Polarisation and the Vote Shares for Nationalist Parties in Latvia, Europe-Asia Studies, vol. 71, no. 5, p. 774—804. doi: https://doi.org/10.1080/09668136.2019.1604945.
34. Meleshevich, A. 2006, Geographical Patterns of Party Support in the Baltic States, Russia, and Ukraine, European Urban and Regional Studies, vol. 13, no. 2, p. 113 — 129.
35. Paiders, J., Paiders, J. 2014, EU border proximity effect on political choice in parliamentary elections of Latvia, Regional Formation and Development Studies. doi: http:// dx.doi.org/10.15181/rfds.v9i1.59.
36. Paiders, J., Paiders, J. 2012, Basic factors of parliament election results in the rural areas of Latvia, Research for Rural Development, no. 2, p. 184—190.
37. Okunev, I. Yu., Shmatkova, L. P. 2020, Transformation of electoral behavior in CIS countries: spatial analysis, Sovremennye tekhnologii nablyudeniya za vyborami: modelirovanie i primenenie v usloviyakh tsifrovizatsii [Modern Technologies of Election Observation: Modeling and Usage under Conditions of Digitalization. Materials of the International Scientific Conference], Chisinau, p. 100—110 (in Russ.).
38. Okunev, I. Yu., Shestakova, M. N., Bibina, E. S. 2020, Neighborhood with Russia: Implications for regional differentiation of public opinion in Belarus sociological and spatial analysis, Russia in Global Affairs, vol. 18, no. 4, p. 10 — 36. doi: http://dx.doi. org/10.31278/1810-6374-2020-18-4-10-36.
39. Zakharova, E. A. 2021, Electoral processes in Norwegian fylke through the prism of spatial analysis, Pskovskii regionologicheskii zhurnal [Pskov Journal of Regional Studies], p. 110—125 (in Russ.).
The author
Lidia S. Zhirnova, Research Fellow, Center for Spatial Analysis in International Relations, Institute for International Studies, MGIMO University, Russia. E-mail: l.zhirnova@inno.mgimo.ru https://orcid.org/0000-0003-2609-5389
[SUBMITTED FOR POSSIBLE OPEN ACCESS PUBLICATION UNDER THE TERMS AND CONDITIONS OF THE ¿a^^jï^JcREATIVE COMMONS ATTRIBUTION (CC BY) LICENSE (HTTPy/CREATIVECOMMONS.ORG/LICENSES/BYM.O/)
