Left-right voting in 10 th Saeima and Riigikogu elections in 2011

Distribution of parties along ideological continuum historically has been the issue of governmental accountability, has served as a competition platform and communication code between political candidates and potential voters. Debate on the framework of the left-right continuum and its impact on voting behaviour are still ongoing in the Western countries, generally recognising the split as such, and rather debating its changing nature. The aim of the paper is to highlight the left-right continuum as a meaningful notion in the research of electoral behaviour both in Latvia and Estonia according to the thesis that voters in general tend to choose a party in elections which corresponds (or lies closest) to their selfplacement in the left-right continuum (Downs, 1957); and the Laponce’s thesis concerning the left-right continuum as the “political Esperanto” – a language allowing parties and voters to understand each other (Laponce, 1981). Results of this research show that although the opinion of the parties themselves and experts concerning placement of the parties along the left-right scale coincides, the left-right continuum does not serve as the “political Esperanto” between the parties and voters. Still the idea of voting for the party, which is closest to one’s own opinion, is particularly well pronounced in Latvia.

One of the most significant processes of political involvement in the model of representative democracy is selection of the state political elite and delegation of power to the state administration authorities.Ideological continuum of the party system can be regarded as framework of competition and one of the main angles of different policy development initiatives.It also serves as a tool for determining voters' behaviour at ballot boxes; because the left-right scale historically has quite often stood for representation of interests of different groups or identified a spectrum of different ethnic, cultural, economic and social issues (Lipset, 1960;Inglehart, Klingemann, 1976;Conover, Feldman, 1981;Laponce, 1981;Fuchs, Klingemann, 1989;Kroh, 2003;Tadosijevic, 2004;Anduiza-Perea, 2006;Dalton, 2008;Haupt, 2010;etc.).In examining the ideological continuum within the context of parties' competition in elections, Downs (1957) developed a spatial modelling concept which assumes that voters in general tend to choose a party in elections which corresponds (or lies closest) to their self-placement in the left-right continuum.Laponce (1981) entitled it "political Esperanto" -a language allowing parties and voters to understand each other and serving as a communication code among representatives of parties, mass media and residents (Freire, 2008).Its meaning and social basis can change over time (Inglehart, 1985).However, other researchers believe that along with putting forward a thesis of "end of ideology" (Bell, 1960) the left-right placement today has lost its meaning or has become insignificant (Lipset, 1981;Mair, 1997, etc).Even though socioeconomic factors are often considered as milestones of political behaviour explanation (Verba, Nie, 1978;LeDuc, Niemi, Norris, 1996;Freire, 2006Freire, , 2008, etc., etc.), yet socioeconomic factors cannot fully explain determinants of society's behaviour (Murray, Vedlitz, 1977) especially in ethnically divided society (Tam Cho, 1999).There is an opinion that ethnic favouritism can be observed in electoral behaviour of ethnically divided society because political elite expects support from voters in elections just like voters with certain ethnic origin expect more social, economical and political advantages as they vote for the same political elite (Chandra, 2006).Ethnic belonging even may lie with social, economic or political benefits it brings (Da, 2006).
Meaning of placement of political attitudes on left-right continuum will not be discussed in this paper, and left-right continuum is used without specifying whether it is accepted in the country as classificatory of classically conservative liberal ideology or role of the state in economics, whether it shows the issues of social and economical inequality in the society or important social and cultural issues for residents and political elite.The aim of the paper is to highlight the left-right continuum as a meaningful tool in the research of electoral behaviour both in Latvia and Estonia, based upon the Downs thesis that voters in general tend to choose a party in elections which corresponds (or lies closest) to their self-placement in the left-right continuum (Downs, 1957); and the Laponce's thesis concerning the leftright continuum as the "political Esperanto" -a language allowing parties and voters to understand each other (Laponce, 1981).The focus of this study is to test: 1) whether the left-right continuum is evident and meaningful in Latvia and Estonia; 2) whether voters have fairly accurate perceptions of parties' left-right positions in Latvia and Estonia and whether they vote according to their self-placement on left-right scale.
Specifying local contect it should be emphasized that during the time period under review traditionally the parties representing interests of the Latvia'a largest ethnical minority -Russians (or the extended group "Russian speaking people") -have not been involved in coalitions.Data of international surveys indicate a relatively high proportion of the respondents, who admitted their affiliation to ethnic minorities.The survey by EUREQUAL indicates that only 56% of respondents affiliate themselves with the majority ethnic group of Latvia -Latvians (the general cluster of 1001 respondent).Within the framework of the present research the above statements are important, because the ethnic split, when ethnic minorities tend to place themselves rather towards the left side of the scale and representatives of the basic nation place themselves rightwards, is one of the arguments referred to in relation to insignificance of the left-right scale.Preceding research (Vikmane, Kreituse, 2011), which is based on correlation of the data from EUREQUAL, World Value Survey, and European Social Survey, confirms that self-evaluation of the Latvian voters concerning placement on the left-right scale has not changed considerably since 1996, still the results differ slightly depending on the research and also the number of respondents, who have responded.The research confirms that ethnic and linguistic factors are closely interrelated, still data of different surveys present different results as regards the impact of ethnic and linguistic factors to the self-evaluation of people as to placement on the left-right scale.

Materials and methods
Data on actual election results for 10 th Saeima elections (2011) and Riigikogu (parliamentary) Elections (2011) was obtained from "Centrālā vēlēšanu komisija" (The Central Election Commission of Latvia) and "Vabariigi Valimiskomisjon" (Estonian National Electoral Committee).
Data on party placement was obtained from experts (N = 5 for Latvia and N = 5 for Estonia) who placed parties along 10 point scale and party top management (board members, secretaries-general and alike) opinion poll (N = 7 for Latvia and N = 6 for Estonia) through they placed their own party and other most probably to be elected parties along the 10 point left-right scale, answering the question "Many evaluate political attitudes as "right" or "left"; where would you put your party and n party (particular parties indicated) along this scale?".Fieldwork was done in the period 08/02/2010-05/03/2010.It is important to emphasise that some parties and unions of parties united during the preelection period.In calculations the mean arithmetic value of placement in case of the left-right scale is taken into account.In order to establish whether there is coherence between the mutual appraisal of political parties within the framework of the left-right continuum of political orientation and votes casted by voters regarding their self-placement on the left-right political orientation, left-right average (arithmetic mean) was calculated for each party placed on the 11 point left-right and 10 point scale by respondents.Potential electoral outcome was calculated by giving each party their percentage of votes according to the percentage of voters placing themselves on the same 11 point scale or 10 point scale accordingly.Research by Kroh (2004) shows that the difference between credibility in 11 point scale and 10 point scale is insignificant (credibility in 11 point scale is 0.97 but 10 point scale it is 0.93) if compared with application of 101 point scale (0.76).
Further, correlation analysis was applied to test: 1) whether voters have fairly accurate perceptions of parties left-right placement in Latvia id est voters and parties have mutual intuitive understanding which could serve as "Political Esperanto" or communication code; 2) whether mathematically calculated results correlate with actual results thus verifying Downs' thesis.Correlation analysis between actual results and answers on direct question regarding voting intentions and additional data on feeling close to the party, used to control the Down's left-right voting hypothesis.

Results
Respondents (Figure 1) in Latvia identify themselves relatively more orientated to the right than respondents in Estonia while less number of respondents refused to respond or could not identify themselves in the left-right continuum (18% of respondents in Latvia, A −23.5% and B −26.4% in Estonia).
Despite the fact that Estonian respondents assess themselves as rather leftwing, according to their evaluation parties are rather rightwing (Table 1, (3)), because none of the six parties with a potential to gain Parliamentary seats has received a score below 5 (on the 10 point scale).More diversity on the scale can be seen in Latvia where parties have received scores ranging from 3 to 8 (on the scale of 11 points) (Table 2, (3)).If comparing the results of both scales, the peak of the mean value of EES2009 (Figure 1, B) is clearly more pronounced than that of EUREQUAL (Figure 1, A).At this very little range of evaluation of parties along the left-right scale differences in respondents' self-assessment in different surveys can have essential impact upon the interpretation of the results.
Results of this research indicate that after recalculation of the votes cast by respondents, assuming that they vote for the party, which is placed closest to self-placement along left-right scale on the leftright continuum, the correlation is very strong, however, reversed (r = −0.93),which means that voters have not voted for parties, which are placed closest to their self-assessment, but vice versa.In the result, both parties, which have been assessed as less rightwing, have not overcome the 5% threshold and
have not gained any Parliamentary seats.In comparison, in Latvia when the respondents' votes are recalculated assuming that they vote for the party, which is placed closest to self-placement along leftright scale on the left-right continuum, the correlation is insignificant (r = 0.20), thus, confirming that the actual election results at the 10 th Parliamentary election are not related to the voters' feeling as to the placement of parties along the left-right scale.
The above is confirmed also by additional correlations between the results of the 10 th Parliamentary election and the survey data on the intention in general election and the data on feeling close to a party.In this case (Table 2, (2) and ( 5)) correlations are very close (r = 0.90 and r = 0.88, correspondingly).An additional survey of the parties' leaders and experts has been carried out with the objective to clarify, whether the fact that the voters' feeling as to the placement of parties along the left-right scale is not related to the actual election results (or, in case of Estonia, this relation is very strong and reverse), which in any case does not comply with the Down's thesis concerning support to the party, which is closest to the voter's self-assessment, is related to the low importance of division of the parties based upon their placement along the left-right scale in Latvia and Estonia, or to the actual existence of the 00038-p.4], where 1 -actual results in 10.Saeima Elections, 2010 (%); 2 -vote intention in general elections (%), 3 -party placement on left-right scale by respondents EES2009; 4 -respondents' votes per parties (%) if they voted for the closest party to their own placement along the left-right scale, 5 -respondents feeling close to a party (%), EES2009.left-right division, which still does not serve as the political "Esperanto" -as clear means of communication between the parties and voters.

Int. Conf. SOCIETY. HEALTH. WELFARE; Congr. of Rehabilitation Doctors of Latvia
Results of the survey indicate that opinions of the parties and experts (recognised local academicians and professors, political technologists, sociologists and alike) concerning placement of the parties along the left-right scale coincide.If there are some differences in opinion in Estonia (r = 0.96), in case of Latvia the opinion of parties and experts fully coincides (r = 1.00) (Table 3).However, when the placement along the scale is recalculated to express it as the respondents' votes, assuming that they vote for the party, which is closest to their self-placement along left-right scale, which they have stated in EUREQUAL survey, on the left-right continuum, in Estonia (Table 3, (c) and (e)) a very strong reverse correlation can be seen again (r = −0.94 and r = −0.67,correspondingly), and in Latvia (Table 3, (c) and (e)) there is insignificant correlation between calculated and actual election results (r = 0.17).

Discussion
Results of party placement in Latvia and Estonia contradict observations from "Mapping policy preference II" (Fig. 1.6 and Fig 1 .7. p.6) (Klingemann, 2006), where during the preceding decade (in Latvia from 1993 to 2002, in Estonia from 1992 to 2003) there has been an opposite trend in the leftright movements of parties: in the beginning of this millennium parties in Latvia became clearly centrist, the left-right range became relatively broad in Estonia.It should be noted that the research "Mapping policy preferences II" was carried out on the basis of analysis of pre-election manifestos of the parties, but in the surveys EES2009 -only the voters' feeling as to the placement of parties on the left-right scale is reflected, thus questioning manifestos as successful communication tool between the parties and their voters.For the sake of providing the context, it should be added that in Latvia the left-right scale serves also as an indication of the ethnically diverse society -parties, which are placed rather leftwing according to the evaluation, traditionally focus on the Russian-speaking voters, and the parties, which are rather rightwing according to the evaluation, are focusing on the voters with Latvian identity.Results of the preceding research (Vikmane, Kreituse, 2010) show that no correlation between self-placement and ethnic-linguistic factors can be found using EUREQUAL data, though Latvia is considered to be one of the ethnically most divided societies in Europe, other data, as those from the World Value Survey 1996 and European Social Survey 2008, show importance of ethnic and linguistic factors upon the selfplacement of inhabitants of Latvia.Thus, it can be assumed that concerning evaluation of the parties the  respondents' answers are influenced by the common stereotype in the society and also the correlation of the impact of ethnic-linguistic origin found in some surveys, and conclusions of the division of the parties on the basis of ethnic origin of the voters can be drawn.The question concerning selfassessment as to the placement on the left-right scale is asked in the survey prior to the call to evaluate the parties along the left-right scale, besides, in the survey there is no explanation as to what the leftright scale means from the content viewpoint, respondents, possibly, express their feelings concerning self-assessment based upon some other principles (for example, based upon their position as to social benefits or allowances or their position concerning taxation policy).
Although the opinion of the parties themselves and experts concerning placement of the parties along the left-right scale coincides, still neither in Latvia, nor in Estonia the left-right continuum serves as the "political Esperanto" or the successful communication tool between the parties and their voters.Similar results, which were obtained in surveys carried out by two different institutions based upon different methodology and the field work at different times (EES (2009) and EUREQUAL), allow to gain a credible confirmation that the Down's concept, which assumes that voters in general tend to choose a party in elections which corresponds (or lies closest) to their self-placement in the left-right continuum, cannot be observed in Latvia or in Estonia.Still the idea of voting for the party, which is closest to one's own opinion, can be fully applied to Latvia, in particular, and also to Estonia.This Int.Conf.SOCIETY.HEALTH.WELFARE; Congr. of Rehabilitation Doctors of Latvia is confirmed by the tight correlation between the research data and actual election results.It should be noted that approximately one year has been in between the survey fieldwork and election in Latvia and in Estonia this term has been approximately two years, which confirms the stability of the voters' choice and also allows to note that the term one year prior to the election is suitable and sufficient for making credible forecast as to the election results, if there are no essential political, economic or other distractions, which could change the voters' opinion radically, in the country.

Conclusions
Results of this research show that the left-right continuum is still topical and significant in terms of research even if understanding of the left-right continuum or even its content has changed.The research has also allowed concluding that if the meaning of the left-right scale is not defined in surveys, various concomitants, as well as public stereotypes have essential impact upon the results.
The Down's left-right concept is not evident in Latvia and Estonia, at least based upon the methodology of the above surveys, when the meaning of the left-right scale was not defined for the respondents in advance.It is especially important in Latvia, which is one of the ethnically most divided societies in Europe, where there is a well-grounded public stereotype concerning placement of the parties along the left-right scale based upon the voters' ethnic-linguistic origin, which exists along with other definitions of the content of the left-right scale.But the tight correlation between the research data on direct voting intention and closeness to a party, and actual election results show that the idea of voting for the party, which lies closest, is evident in Estonia, and particularly well pronounced in Latvia.
Definition of the question field would restrict the interpretation options considerably and, possibly, would open up a field for forecasting the election by utilising the left-right scale as a tool for forecasting the election results.
Study is developed with the support of European Social Fund (ESF).
Data on political orientation of according to the left-right self-placement in 10 point scale was obtained from EU Commission Funded FP6 EUREQUAL project (PI professor Stephen Whitefield, University of Oxford), "Social Inequality and Why it Matters for the Economic and Democratic Development of Europe and its Citizens: Post-Communist Central and Eastern Europe in Comparative Perspective".General cluster covers 1058 respondents in Estonia (N = 1058) and 1001 respondents in Latvia (N = 1001) out of which 753 forms in Estonia and 976 forms in Latvia have respondent's political attitude measured within the framework of 10 point left-right continuum.Survey's field work took place in 2007.Data on the voter left-right political orientation and the left-right placement of the parties was obtained from EES (2009), European Parliament Election Study 2009, [Voter Study], Release, 30/06/2010, (www.piredeu.eu).The PIREDEU data collection has received funding from the European Community's FP7 Programme (grant agreement 211810).Fieldwork in Latvia was done by "Latvian Facts" during the period 10/06/2009-28/06/2009.Fieldwork in Estonia was done by "Saar Poll LLC" during the period 07/06/2009-25/06/2009.General cluster covers respondents 1000 respondents in each country -700 f2f and 300 WebCATI at home of respondent (2009 EES Technical Report, Latvia -EUI-GALLUP, 2009 EES Technical Report, Estonia -EUI-GALLUP, www.piredeu.eu).

Figure 1 .
Figure 1.Percent (%) distribution of respondents according to self-placement in the left-right 10 and 11 point scale in Latvia and Estonia.Source: A -EU Commission Funded FP6 EUREQUAL project (PI Professor Stephen Whitefield, University of Oxford), "Social Inequality and Why it Matters for the Economic and Democratic Development of Europe and its Citizens: Post-Communist Central and Eastern Europe in Comparative Perspective", B -EES (2009), European Parliament Election Study 2009, [Voter Study].
Source: "Vabariigi Valimiskomisjon" (Estonian National Electoral Committee); "Centrālā vēlēšanu komisija" (The Central Election Commission of Latvia); data from author's opinion poll; data from EU Commission Funded FP6 EUREQUAL project (PI Professor Stephen Whitefield, University of Oxford), "Social Inequality and Why it Matters for the Economic and Democratic Development of Europe and its Citizens: Post-Communist Central and Eastern Europe in Comparative Perspective", where a -Actual results in Elections (%), b -parties placement along the scale by parties, c -respondents' votes per parties (%) if they voted according to their self placement (EUREQUAL) for the closest party according to thei placement along the left-right scale by parties, d -parties placement along the scale by experts, e -respondents' votes per parties (%) if they voted according to their self placement (EUREQUAL) for the closest party according to the placement along the left-right scale by experts.

Table 1 .
Election results for 2011 Riigikogu elections, data on vote intention, self-placement of parties, and feeling close to a party from EES (2009).

Table 2 .
Results for 10 th Saeima elections 2010, data on vote intention, self-placement of parties, and feeling close to a party from EES (2009).

Table 3 .
Results for 2011 Riigikogu and 10 th Saeima (2010) elections potencial outcome, placement of parties by parties and experts along the left-right scale.