Voter Turnout in Local Elections Paper for the Research Conference “Civil Society and Democracy” AUSTRIAN RESEARCH ASSOCIATION – WORKING GROUP ON DEMOCRACY Vienna, November 9-10, 2012 Armin Mühlböck Voter turnout in local elections Armin Mühlböck 1. Introduction Today politics, media and society call attention to a critical development of democracy. Nolte 1 determines that in society, politics and science a debate is in progress which raises questions for the future of democratic systems. Various considerations like decline of voter turnout, weakening trust of the citizens in integrity and alignment to public welfare of their ambassadors, increasing not legitimized influence of private interests, a banalization caused by acting of the politicians themselves and the statements in the mass media arising from these facts, an economic pressurized middle class and a political acting that lags behind economic processes instead of shaping policy play a significant role in this discussion.2 Communities carry great weight as “schools of democracy” and “centres of legitimization” for superior political levels and they encourage democratic development in political systems. 3 But there is a big burden on communities and as a consequence this fact could have negative influence on legitimization of political systems and the receipt of democratic level in local authorities.4 An excessive financial demand of local authorities, delegating responsibilities from national or regional level to local level, the decrease of infrastructure in rural areas, demographic change and problems by recruiting people who are willing to be representatives express this specific situation.5 Recently financial difficulties and the problems of recruiting ambassadors resulted in a variety of amalgamations of communities in Switzerland. The local Swiss media already talks about a “boom”.6 In Austria you can notice an unmistakable trend to merge small and very small communities. 1 Paul Nolte: Von der repräsentativen zur multiplen Demokratie, in: Aus Politik und Zeitgeschichte 1-2/2011, S. 5 – 12, hier S. 5. 2 Asiye Öztürk: Editorial, in: Aus Politik und Zeitgeschichte 1-2/2011, S.2. 3 Angelika Vetter: Lokale Politik als Ressource der Demokratie in Europa? Lokale Autonomie, lokale Strukturen und die Einstellungen der Bürger zur Politik. Opladen: Leske + Budrich, 2002, S. 16; Roland Roth: Handlungsoptionen zur Vitalisierung der Demokratie. Expertise für die Bertelsmann Stiftung, Online-Dokument: http://www.b-be.de/fileadmin/inhalte/aktuelles/2009/09/nl20_bertelsmann_vitalisierung.pdf; Zugriff am 17. 8. 2011. S. 16f.. 4 :Projektkonzept zur lokalen Demokratieforschung entnommen: Michael Ritter, Armin Mühlböck: Politische Stabilität auf kommunaler Ebene – Forschungskonzept. Nicht veröffentlichtes Manuskript, 2012. Armin Mühlböck: Die Gemeinden. Beitrag zum Buchprojekt „Landesgeschichte Salzburg – Die Ära Katschthaler und Schausberger“, im Erscheinen. 5 Armin Mühlböck: Die Zukunft der Gemeinden am Land. In: Herbert Dachs, Roland Floimair (Hg.): Salzburger Jahrbuch für Politik 2010. Wien, Köln, Weimar: Böhlau, 2010, S. 120-142, hier S. 120. 6 http://www.tagesschau.sf.tv/Nachrichten/Archiv/2009/01/07/Schweiz/Boom-bei-Gemeindefusionen, abgerufen am 7.6.2010 In Styria a large-scale project of reforms in regional and local concerns is already in implementation.7 Against the background of these facts you can initiate a discussion about the quality of development of democracy in the communities. In this paper we make a detailed consideration at participation in local elections. How has the participation in local elections been developing in the longer term and which factors can explain the extent of voter turnout? 2. Voter turnout in communities If you take a look at participation in elections in terms of voter turnout a reference to the recurring discussion in literature about the importance of voter turnout for democratic development will be necessary. In this discussion questions are raised how far high voter turnout is a basic requirement for democracy or if decreasing voter turnout is a substantial problem for the politics in a democracy.8 The result is traditionally ambivalent and can be affiliated to theoretical principles of democracy. 9 Now the discussion should not be expanded and continued. It should be a hint at the evidence of discussion and the fact that the debate provides the basis for voter turnout. Today engaging in questions of voter turnout seems to be out of fashion because elections are scored as antiquated in comparison to other forms of political participation.10 Nonetheless elections and voter turnout at the local level appear to be very interesting because you can recognize special characteristics in electoral behaviour at the local level but not at the superior level (in Austria: regional or national elections).11 The willingness to differ from the usual party preference should be much more pronounced and the electorate is more contend to choose another 7 Armin Mühlböck: Die Zukunft der Gemeinden am Land. In: Herbert Dachs, Roland Floimair (Hg.): Salzburger Jahrbuch für Politik 2010. Wien, Köln, Weimar: Böhlau, 2010, S. 120-142. Zur lokalen und regionalen Strukturreform in der Steiermark siehe z.B.: http://www.gemeindestrukturreform.steiermark.at; Up to now in Styria fusions of districts have been fixed, in terms of fusions communities the referendums on 20.11.2011 and 15.01.2012 led to negative results, in the referendum on 30.09.2012 the referendum led to a positive result. Thus three communities (Gai, Hafning bei Trofaiach und Trofaiach) will be amalgamated on 01.01.2013. 8 See for example Arend Lijphart: Unequal Participation: Democracy's Unresolved Dilemma. In: American Political Science Review 19, 1997, S. 1-14; Sidney Verba: Would the Dream of Political Equality Turn out to Be a Nightmare? In: Perspectives in Politics 1, 2003, S. 663-679. 9 See for example. Manfred G. Schmidt: Demokratietheorien. Eine Einführung. Opladen, 2000, S. 451 ff. 10 Angelika Vetter: Lokale Politik als Ressource der Demokratie in Europa? Lokale Autonomie, lokale Strukturen und die Einstellungen der Bürger zur Politik. Opladen: Leske + Budrich, 2002, S. 50. 11 Anton Pelinka: Kommunalpolitik in Österreich. In: Rüdiger Voigt (Hg.): Handwörterbuch zur Kommunalpolitik. Opladen, 1984, S. 259-261, hier S. 260. party or to deny participating in elections at the local level.12 In the insufficient literature you often find references to a gap in research of elections. Furthermore questions of voter turnout at local elections are marginalized.13 The first presentation of previously undisclosed data and analysis to participation of voters in local elections should fill a gap in general in the research of Austrian local politics and in particular in electoral behaviour at local elections.14 In this survey local elections in seven years in 118 communities in the federal country of Salzburg (excepting the city of Salzburg) are submitted to examination. Since the local elections in 1979 information about 826 datasets is gathered and analysed in the following written pleadings.15 The data is conceptualized as Time-Series-Cross-Section-Analysis, where a comparison between communities and election dates has been taken place. Each ballot in each community generates an elementary unit. 3. The development of voter turnout in the communities of Salzburg since the end of the 1970ies. Graph 1 shows the development of the average of voter turnout in 118 communities in the federal country of Salzburg (excepting the city of Salzburg) beginning with the local elections in 1979.16 The results convey that a slight decline of voter turnout is starting at the middle of the 1980ies, from 86 Percent in 1984 to 81 Percent in 2009. The values never deceed the point of 80 percent and voter turnout keeps at high stage.17 12 Heinz Zielenski: Wahlverhalten in Kommunalwahlen. In: Rüdiger Voigt (Hg.): Handwörterbuch zur Kommunalpolitik. Opladen, 1984, S. 496-499, hier S. 497. 13 http://www.uni-stuttgart.de/soz/avps/forschung/wahlbeteiligung.de.pdf, abgerufen am 25.10.2012, and Austrian contributions to direct-democratic features at the local level, see for example Martin Dolezal / Universität Wien und Klaus Poier / Universität Graz um Tag der Österreichischen Politikwissenschaft im November 2012 in Graz: Rechtliche Hürden, Struktur des Parteienwettbewerbs oder sozioökonomischer Kontext? Die Anwendung direktdemokratischer Verfahren in Österreichs Gemeinden: http://www.uni-graz.at/ofreaeww/tpw/ofreaeww_tpwfreitag.htm, abgerufen am 25.10.2012. Zum Forschungsstand siehe z.B.: Angelika Vetter: Lokale Politik als Ressource der Demokratie in Europa? Lokale Autonomie, lokale Strukturen und die Einstellungen der Bürger zur Politik. Opladen: Leske + Budrich, 2002, S. 49ff; Freitag, Markus (1996). Wahlbeteiligung in westlichen Demokratien. Eine Analyse zur Erklärung von Niveauunterschieden. Swiss Political Science Review 2(4). S. 1-36. 14 Armin Mühlböck: Die Gemeinden. Beitrag zum Buchprojekt „Landesgeschichte Salzburg – Die Ära Katschthaler und Schausberger“, im Erscheinen. 15 We have excluded the city of Salzburg because of avoiding problems in the analysis. 16 We emanate from the fact that the average of voter turnout of all communities is a better solution for our aim than the value that is consisted of all citizens of voting age and all persons who used their right to vote. See Landesstatistik Salzburg: www.salzburg.gv.at/20003stat/wahlen/gvw/index.htm 17 Recently an expansion of the electorate has occurred. On the one hand citizens of the European Union gained the right to vote and on the other hand minimum age for voting declined from 18 years to 16 years. See Barbara Steininger: Gemeinden. In: Herbert Dachs et al. (Hg.): Politik in Österreich. Das Handbuch. Wien: Manz-Verlag, 2006, S. 9901007, hier S. 995. Origin: Landesstatistik Salzburg, own calculation and presentation on the base of the results of local elections in the communities in the federal country of Salzburg on 14.10.1979, 14.10.1984, 08.10.1989, 06.11.1994, 07.03.1999, 07.03.2004, 01.03.2009 The comparison of the average of voter turnout in the communities of the federal country of Salzburg in local, regional and national elections since the beginning of the 1980ies visualizes that the average of voter turnout at national elections is on a high level with values exceeding the point of 90 percent in national elections at the beginning of the 1980ies. In the sequel you can recognize a sharp drop of voter turnout at national elections. Since the middle of the 1990ies voter turnout at national elections becomes stable at a level about 75 percent. Only the elections in 1995 and 2002 have not followed this trend.18 The average of voter turnout at regional elections in the communities in the federal country of Salzburg is clearly lower than at national elections until the beginning of the 1990ies, afterwards a permutation is occurring. If you exclude the national election in 1995 the average of voter turnouts at regional elections will be higher than at national elections since 1994. 2002 and 2004 the average of voter turnouts at national, regional and local elections are almost at the same level. 18 The elections 1995 and 2002 have been snap elections. In 1995 the ÖVP quitted the coalition with the SPÖ, 2002 the problems of the FPÖ culminated in Knittelfeld and led to breach of the coalition between ÖVP and FPÖ. Origin: Landesstatistik Salzburg und Bundesministerium für Inneres (interior ministry) own calculation and presentation on the base of the results of local elections in the communities in the federal country of Salzburg on 14.10.1979, 14.10.1984, 08.10.1989, 06.11.1994, 07.03.1999, 07.03.2004, 01.03.2009, the results of regional elections in the communities in the federal country of Salzburg on 25.03.1979, 25.03.1984, 12.03.1989, 13.03.1994, 07.03.1999, 07.03.2004, 01.03.2009, the results of national election in the communities in the federal country of Salzburg on 06.05.1979, 24.04.1983, 23.11.1986, 07.10.1990, 09.10.1994, 17.12.1995, 03.10.1999, 24.11.2002, 01.10.2006, 28.09.2008 Since the year 1999 regional elections and local elections have been held on the same day in the federal country of Salzburg.19 This fact has had a positive influence on the average of voter turnout at regional elections which starts ascending at this time. Voter turnout at local elections features high stability in comparison to local or national elections.20 In all 826 examined ballots it is higher than 80 percent. Nevertheless you find some differences between the results of the communities and several election dates. 19 Despite amalgamation of the election date of local and regional elections the voter turnout at local elections deviates from voter turnout at regional election because postal voting at regional elections have (has) been introduced in the year 2009. 20 Recently has been an expansion of the electorate. On the one hand citizens of the European Union gained the right to vote and on the other hand minimum age for voting declined from 18 years to 16 years. See Barbara Steininger: Gemeinden. In: Herbert Dachs et al. (Hg.): Politik in Österreich. Das Handbuch. Wien: Manz-Verlag, 2006, S. 9901007, hier S. 995. Origin: Landesstatistik Salzburg, own calculation and presentation on the base of the results of local elections in the communities in the federal country of Salzburg on 14.10.1979, 14.10.1984, 08.10.1989, 06.11.1994, 07.03.1999, 07.03.2004, 01.03.2009 The boxplot shows that in seven ballots in the 118 communities of the federal country of Salzburg voter turnout ranges between 68,5 percent at least and 97,6 percent at most. The range of deviation represents nearly 30 percentage points. In three quarters of the ballots voter turnout is accounted for 80 percent or more. In the course of time you can notice considerable rise of the count of communities where voter turnout is lower than 75 percent (measured by the 10th percentile). (Graph 4) In the year 1979 in no community of Salzburg voter turnout at regional elections have come below 75 percent, in 1984 you find two communities where less then 75 percent have used their right to vote. In 1989 already 11 communities fall into this category and until the year 1999 the number raises to 15 communities where voter turnout does not achieve the level of 75 percent. The election in 2004 takes a special position because voter turnout for once has been increasing contrary to the trend of decline in voter turnout.21 In the elections 2009 voter turnout is decreasing again. In 2004 in two communities voter turnout has been below 75 percent, in 2009 this number has reached its peak with 22 communities. In parallel case the count of communities where voter turnout is higher than 90 percent is obviously decreasing. (Graph 5) In 1979 have been 35 communities in Salzburg where more than 90 percent have exercised their right to vote. In the sequel the number declines to seven or eight communities where voter turnout is topping the level of 90 percent in the elections 1999, 2004 and 2009. 21 The SPÖ Salzburg reached the first position in the regional elections 2004. Before that the ÖVP had been the strongest party since 1945. In 2004 a change of government took place. In advance polls predicted this fact and probably mobilized the electorate in regional and local elections because they were held on the same day. Origin: Landesstatistik Salzburg, own calculation and presentation on the base of the results of local elections in the communities in the federal country of Salzburg on 14.10.1979, 14.10.1984, 08.10.1989, 06.11.1994, 07.03.1999, 07.03.2004, 01.03.2009 4. Approach of investigation Starting from the consideration that communities as “schools of democracy” and “centres of legitimization” weighty contribute to vitalization of democratic system, we have shown the longterm development of participation in elections at the local level with a view to visualize development of democracy in the communities. We have stated that the average of voter turnout in the communities in the federal country of Salzburg is slightly declining but stable on a high level. All the values top the point of 80 percent. The number of communities with low voter turnout deceeding the level of 75 percent has considerably risen in the period from 1979 to 2009 whereas the number of communities with high voter turnout exceeding the level of 90 percent has declined. Therefore voter turnout altogether has decreased in a modest way in the period of 30 years. If you have a look at all data they will range between 68,5 and 97,6 percentage points. This fact results in a difference of 30 percentage points which constitutes a great spectrum. Further analysis shows that voter turnout differs widely between the communities but it remains relatively stable in the communities. The average range reaches the result of 9,2 percentage points in each individual community in the period from 1979 to 2009. This value is considerably lower than the variance of results in all communities. This fact concludes that voter turnout at local elections is not only influenced by individual factors but also by other conditions in the context of elections.22 Which conditions we are taking about? 22 Angelika Vetter: Institutionen und lokale Wahlen: Wo bleiben die Wähler? In: Vetter, Angelika (Hrsg.). Erfolgsbedingungen lokaler Bürgerbeteiligung. VS Verlag für Sozialwissenschaften, 2008, S. 51f. 5. Approach of explanation of height of voter turnout In the sequel we put value to influences on voter turnout at local elections on the aggregate level. Which macro-factors have a bearing on the height of voter turnout? Recently a large-scale research democracy project in Germany has been handled that attends to voter turnout in communities23. This project provides some toeholds. Although the searches in the German project are not pursued the results show that beside the size of communities mainly socio-economic factors have an important influence on participation at the local level.24 Angelika Vetter25 refers to studies with the title “Size and Democracy” which lay the focus on the variable “size of communities” (= number of inhabitants) and argues that in smaller local entities higher voter turnout can be expected. Up to now it is vague which procedures induce the fact that voter turnout in smaller entities is higher than in bigger entities.26 Graph 6: The communities in the federal country of Salzburg Origin: SAGIS, 25.10.201227 23 -Projekt 2005-2008: Wahlbeteiligung in den Städten und Gemeinden der Bundesrepublik Deutschland: Zum Einfluss institutioneller und politischer Kontextfaktoren auf die lokale Wahlbeteiligung im Vergleich der deutschen Bundesländer und im internationalen Vergleich. Projektleitung: Dr. Angelika Vetter. Siehe z.B.: http://www.uni-stuttgart.de/soz/avps/forschung/wahlbeteiligung.de.pdf , abgerufen am 25.10.2012. 24 Angelika Vetter: Institutionen und lokale Wahlen: Wo bleiben die Wähler? In: Vetter, Angelika (Hrsg.). Erfolgsbedingungen lokaler Bürgerbeteiligung. VS Verlag für Sozialwissenschaften, 2008, S. 49 – 72. 25 Vetter, Angelika (2008). Institutionen und lokale Wahlen: Wo bleiben die Wähler? In: Vetter, Angelika (Hrsg.). Erfolgsbedingungen lokaler Bürgerbeteiligung. VS Verlag für Sozialwissenschaften, S. 49 – 72. 26 Dahl, Robert A./Tufte, Edward R. (1974). Size and Democracy. Stanford, Cal: Stanford University Press. Verba, Sidney/Nie, Norman H. (1972). Participation in America: Political democracy and social equality. New York: Harper and Row Verba, Sidney/Nie, Norman H./Kim, J. (1979). Participation and political equality. A sevennation comparison. Cambridge: Cambridge University Press. Blais, Andre/Carty, R.K. (1990). Does Proportional Representation Foster Voter Turnout? In: European Journal of Political Research, S. 167-181. Blais, Andre/Dobrzynska, Agnierszka (1998). Turnout in electoral democracies. In: European Journal of Political Research 33, S. 239-261. Frandsen, A.G. (2002). Size and electoral participation in local elections. In: Environment and Planning C-Govemment and Policy 20, S. 853-869. 27 http://www.salzburg.gv.at/themen/se/sagis/sagis_gis_online.htm In the period from 1979 to 2009 the 118 communities in the federal country of Salzburg have between 257 (Tweng im Lungau in the year 1989) and 19.501 (Hallein im Tennengau in the year 2009) inhabitants, the average preserves the value of 2.927 inhabitants. In this data the city of Salzburg is excluded. If you add the city of Salzburg and then compare the average of inhabitants in communities in the Austrian federal countries excluding the capital Vienna on the reference day 01.01.2010 you will see that the federal country Salzburg has the highest average of inhabitants in relation to the number of communities. Carinthia is on the second position. Table 1: The communities in Austria Inhabitants on reference day 01.01.2010 Population Burgenland Communities Average 283.965 171 1.661 1.208.372 542 2.229 706.873 279 2.534 Niederösterreich (Lower Austria) 1.607.976 573 2.806 Oberösterreich (Upper Austria) 1.411.238 444 3.178 Vorarlberg 368.868 96 3.842 Kärnten (Carinthia) 559.315 132 4.237 Salzburg 529.861 119 4.453 6.676.468 2.356 2.834 Steiermark (Styria) Tirol (Tyrol) Total (excluding Vienna) The considerations that socio-economic factors have influence on voter turnout stand in tradition of election sociological theories which have been developed by Lazarsfeld 28, Lipset and Rokkan29.30 28 Lazarsfeld, Paul/Berelson, Bernhard/Gaudet Hazel (1944). The People's Choice. New York: Duell-Sloan and Pearce. Siehe auch z.B.: Lazarsfeld, Paul et al. 1948: The People’s Choice. How the Voter Makes up his Mind in a Presidential Campaign. New York: Columbia University Press. Vgl. u.a.: Thaidigsmann, Isabell (2004). Sozialstruktur und Wählerverhalten. Das Ende einer alten Beziehung? Arbeitspapier / Dokumentation herausgegeben von der KonradAdenauer-Stiftung, Nr.126/2004. S. 6. Und Wüst , Andreas M. (2002). Wahlverhalten in Theorie und Praxis: die Bundestagswahlen 1998 und 2002. Interntepublikation: abgerufen am 11.1.2012 unter: http://andreas.unihd.de/oberr2002.pdf , S.2 ff. 29 LIPSET, Seymour Martin and Stein ROKKAN (1967). “Cleavage Structures, Party Systems and Voter Alignements: An Introduction”, in Seymour Martin LIPSET and Stein ROKKAN (eds.). Party System and Voter Alignements: Cross Perspectives. New York/London: The Free Press, pp. 1-64. 30 For a summary of the theories see: Markus Freitag: Wahlbeteiligung in westlichen Demokratien. Eine Analyse zur Erklärung von Niveauunterschieden. Swiss Political Science Review 2(4), 1996, S. 1-36. They lay focus on the individual level of voter turnout. They justify as well as Verba, Nie and Kim that predominantly individual resources offer an explanation for participating in local elections (= socio-economic model for political participation, called “SES-Model”). The facilities of individual resources are basic requirements for participating in elections. Socio-demographic features like gender, age and especially apprenticeship play an important role among other features like language competence, social state and others.31 How do these principles bear upon the context? They argue that the extent of participation in elections is dependent on socio-economic modernization, economic efficiency and the development of the political system. Thus a high grade of socio-economic modernization should have positive effect on voter turnout. These considerations contain cognitive and material resources of individual (in this case: local) societies.32 We want to quantify local resources and economic efficiency by the indicator of financial power. The financial power is determined by the sum of property tax , council tax, commercial tax , payroll tax, beverage tax and the profit share to federal state which dues to communities excluding the casino levy33. We use financial power for five years to demonstrate the effect on voter turnout at local elections. The extent of socio-economic modernization is measured by a local index of tertiarization. The term “tertiarization” means the socio-economic transformation which has started in the 1970ies in the western industrial states. In several states this procedure differs in way and time. Tertiarization means not only that the tertiary sector gains in importance versus the primary sector (agriculture, forestry and fisheries) and the secondary sector (manufacturing industry) but also that in occupational patterns you can recognize a tendency to high-skilled tasks and intellectual work incident to a longer period of school education.34 The results in the communities of the federal country of Salzburg show that the percentage of people who passed the Austrian “matura” (similar to A-Levels in Great Britain) and graduates has been considerably increasing from 6,3% in 1981 to 17,4% in 2009. The average of employees of 31 See for example Verba, S. et al: Participation and Political Equality: A Seven-Nation Comparison, Cambrigde, 1978; M. Schneider, H. Willems: Youth Participation, Globalisation and Democracy. Coyote Youth Partnership, 14/2009, S. 39-42. 32 Markus Freitag: Wahlbeteiligung in westlichen Demokratien. Eine Analyse zur Erklärung von Niveauunterschieden. Swiss Political Science Review 2(4), 1996, S. 15ff. 33 See http://www.ris.bka.gv.at/Dokumente/Bundesnormen/NOR40094177/NOR40094177.html und http://www.nbz-online.at/index.php?option=com_content&task=view&id=234&Itemid=67, beide angerufen am 25.10.2012. 34 Baethge, Martin (2000). Abschied vom Industrialismus: Konturen einer neuen gesellschaftlichen Ordnung der Arbeit. In: SOFI-Mitteilungen Nr. 28/2000, S. 87 – 101, hier S. 87; Elkmann, Reinhard/Lauterbach, Nora/Wind, Stephan (1994). Working Paper: Tertiärisierung regionaler Wirtschaftsstrukturen: Eine empirische Analyse kreisfreier Städte und Landkreise in Hessen, Rheinland-Pfalz und dem Saarland. Arbeitspapier des Instituts für Statistik und Ökonometrie, Johannes-Gutenberg-Universität. STATOEK, No. 6. Abgerufen am 25.10.2012 unter: http://econstor.eu/bitstream/10419/49969/1/257644997.pdf. higher services sector (without trade, transport and tourism) has gone up from 18,7% in 1981 to 33,2% in 2009.35 Quelle: Statistik Austria, own calculation and presentation on the result of the population census on 12.05.1981, 15.05.1991, 15.05.2001 and the labour statistics(= Abgestimmte Erwerbsstatistik) of Statistik Austria on 31.10.2009 For acquisition of data of tertiarization we have generated an index that includes the components of higher services sector and higher education. This could be a good description of this phenomenon. We use the percentage of employees in higher services sector (without trade, transport and tourism) in relation to all employees (HSS) and the percentage of people who passed the “matura” and graduated from university in relation to the population aged more than 15 years (HE). The data of education of the population and the occupation in the higher services sector are not available for the years in which have been held elections. Therefore we use data that chronological comes close to election date.36 Although an extensive consideration of control variables is not possible we want to add the aspect to the analysis of voter turnout at local elections. The data show that the average of voter turnout in the communities in the federal country of Salzburg is slightly declining but stable on a high level. The number of communities with low voter turnout deceeding the level of 75 percent has considerably risen in the period from 1979 to 2009 whereas the number of communities with high voter turnout exceeding the level of 90 percent has declined. 35 We have chosen this definition because in these occupational areas you can expect a higher school education. Higher Services Sector = Information and Communication Sector; Financial and Insurance Sector; Sector of Property and Housing; Freelance, Scientific and Technical Services; Other Economic Services; Public Administration, Defense and National Insurance; Education and Teaching; Health and Social Affairs; Art, Entertainment and Recovery; Exterritorial Organizations. 36 Election years 1979 and 1984: data from population census 1981; Election years 1989 and 1994: data from population census 1991; Election years 1999 and 2004: data from population census 2001; Election year 2009: data from labour statistics 2009, Statistik Austria You can raise the question if altering the time factor will place great importance on the height of voter turnout. As a supplement to time factor you can talk about socio-economic modernization. In the literature the socio-economic modernization process is responsible for general decline of voter turnout. These processes give rise to dissolution of traditional cultural surroundings and social classes, a secularization and individualization of society and increasing mobility of the members of society.37 6. Empiricism 6.1. Voter turnout at local elections: Bivariate Models Origin: Landesstatistik Salzburg, own calculation and presentation on the base of the results of local elections in the communities in the federal country of Salzburg on 14.10.1979, 14.10.1984, 08.10.1989, 06.11.1994, 07.03.1999, 07.03.2004, 01.03.2009 If you have a look at the bivariate relationship between inhabitants of a community and voter turnout, you will state that voter turnout will decline if the number of inhabitants increases. Graph 9 shows this effect, (r = 0,558, sig. = 0,000). 37 Angelika Vetter: Lokale Politik als Ressource der Demokratie in Europa? Lokale Autonomie, lokale Strukturen und die Einstellungen der Bürger zur Politik. Opladen: Leske + Budrich, 2002, S. 69. Origin: Landesstatistik Salzburg, own calculation and presentation on the base of the results of local elections in the communities in the federal country of Salzburg on 14.10.1979, 14.10.1984, 08.10.1989, 06.11.1994, 07.03.1999, 07.03.2004, 01.03.2009 The analysis concerning the socio-economic features like financial power has no effect on the height of voter turnout at local elections (r = –0,065, sig. = 0,101). If you cast a glance at voter turnout and the phenomenon of “tertiarization” you will find an evident correlation between the index of tertiarization (TI) and voter turnout (VT). A relatively higher index of tertiarization (communities with a higher percentage of higher-educated persons in population and a higher percentage of employees in services sector) is associated with a relatively lower value of voter turnout. Does this result contradict the usual theory that a higher education of population leads to higher levels of voter turnout? In the view of local elections in the period from 1979 to 2009 in the communities in the federal country of Salzburg we have observed that in big communities where you find more highereducated people voter turnout is significantly lower than in small communities where the percentage of higher-educated people is slight.38 Related to the time-factor the analysis shows a light correlation inversely proportional to voter turnout (r = –0,298, sig. = 0,000). How far the time factor interferes with a multivariate model will be subject of further investigations. 6.2. A first model for explanation of voter turnout at the local level If you pull up theoretical knowledge of bivariate analysis you will have to prove in which extent the independent variables size of the community (= number of inhabitants) and index of tertiarization in a community can explain the variation of the dependent variable voter turnout. The financial power 38 The shown results have to be interpreted as a temporary statement which cannot be ensured and intensified. could have been eliminated in the bivariate model. First of all you have to check if multicollinearity can be excluded. You can notice that a correlation between size of a community (= number of inhabitants) and the index of tertiarization in a community is measurable on a low level. If the number of inhabitants is increasing the value of the index of tertiarization will also be heightening (r = 0,360, sig. = 0,000). Furthermore the correlation calculations shows that both the correlation between voter turnout at local elections and size of communities (= number of inhabitants) controlled by the index of tertiarization (r = –0,457, sig. = 0,000) and the correlation between voter turnout at local elections and the index of tertiarization (r = –0,525, sig. = 0,000) will be stable. In terms of the residues the requirement for normal distribution of the two independent variables exists. Although on the basis of the tests homoscedasticity is not given we adhere to enforcement of regression analysis without weighted least squares (WLS) because it has become clear that a weighted regression signifies no remarkable improvement. The attempt to explain voter turnout at local elections (VT) with the aid of size of communities (S), the index of tertiarization (TI) and the citation of the time factor shows the following result. Table 2: Explanation of voter turnout at local elections in the communities in the federal country of Salzburg (multivariate regression: b, beta-coefficients und Sig.) Model 1 Model 2 Constant Constant 88,85 89,554 b beta-Coefficient Sig. b beta- Coefficient Sig. Size of community (S) -0,001 -0,389 0,000 -0,001 -0,390 0,000 Index of Tertiarization (TI) -0,929 -0,467 0,000 -0,929 -0,467 0,000 --- --- --- 0,000 -0,001 0,983 Time (T) 2 Adj.R N 0,499 0,499 826,000 826,000 In the linear multivariate regression model the size of the community (S) and index of tertiarization (TI) can explain nearly half of the variance of voter turnout at local elections (Ad.R2 = 0,499). The extent of this influence on these variables seems to be impressive (Model 1). Moreover the analysis indicates that the time (T) factor looses influence on voter turnout at local elections in the multivariate setting while features like the size of communities (S) and the index of tertiarization (TI) remain firm (Model 2). For the prediction of voter turnout at local elections the following regression equation is available. 7. Summary In smaller communities voter turnout at local elections is quite higher than in bigger communities. You can recognize the coherence in this investigation, but the features that induce this process remain untold. It is not clear if the smaller distance between politicians and the electorate in comparison to the regional, national or European level has relevance. How important are the aspects that citizens on the local level have the possibility to involve in the political discourse or that the themes of local politics can be communicated easier than on the regional or national level? In which extent operate the straightforwardness of the action field at the local level and the circumstance that local politics deals with the structuring of the immediate living environment and therefore solutions that result in armchair decisions might be easier? The size of the community (= number of inhabitants) and the socio-economic modernization process play an important role to explain participating in local elections. We have noticed that in communities where the percentage of high-skilled persons is relatively high voter turnout is considerably decreasing in comparison to communities where the percentage of high-skilled persons is low. The results in this analysis do not correspond to the usual expectations from electoral analysis that a high level of socio-economic modernization in the society has a positive effect on voter turnout. About an answer to this question we only can conjecture. We have to clarify that this data provides no expressiveness on the individual level. We have not raised the question if the individual participating in elections is affected by the individual education level or not. It had not been a contradiction if the result on the aggregate level would have been reversed into the opposite on the individual level. That means a direct proportional correlation between the individual education level and the individual willingness to participate in the process of decision-making in terms of elections. The results provide indications of an interaction between features like size of the community (= number of inhabitants) and the index of tertiarization. Hence a setting in the communities is emerging that leads to high or low voter turnout at local elections. The questions which processes have specific influence on voter turnout in communities must remain unanswered. This will be the basis for further investigation.