GEOGRAPHIA POLONICA 2011, 84, 2, 93-113 http://dx.doi.Org/10.7163/GPol.2011.2.7 POLISH PRESIDENTIAL ELECTION OF 2010: A STUDY OF THE POWER OF VOTERS IN BIG AND MEDIUM-SIZED TOWNS *ROMAN MATYKOWSKI and **KATARZYNA KULCZYŃSKA Institute of Socio-Economic Geography and Spatial Management, Adam Mickiewicz University, Dzięgielowa 27, 61-680 Poznań, Poland E-mails: *mat(§Jamu.edu.pl, **katakul(§Jamu.edu.pl Abstract: In the literature on the subject, urbanisation is regarded as one of the most impor- tant factors shaping electoral behaviour. The effect of this factor has also been corroborated by studies in Poland, where one can speak of urban- and rural-oriented parties. To determine the significance of the urban electorate in Poland, use was made of the procedure of backward elimi- nation of voters in the successive biggest towns. The next step involved identifying the structure of support for the leading presidential contenders in the 2010 election at each stage of the rank elimination of the towns. It was already in the parliamentary elections at the start of the 21st century that big cities and the larger of medium-sized towns turned out to be their 'engines': with their highest voter turnouts, they crucially affected the results at the national scale. That is why an analysis was made of voter alignment in towns of this size category over the years 2001-2007, and on this basis various electoral types of towns were distinguished. Key words: big and medium-sized towns, electoral types of towns, Poland INTRODUCTION is a quantity "significantly affecting other It is a shared aim of many geographical stud- quantities, whether it is a classificatory or ies of the spatial variability of a phenomenon an ordering one" (Chojnicki and Czyż 1978, to seek and identify factors responsible for p. 11). Sometimes a separate category—that this heterogeneity. That is why their routine of determinants (conditions)—is distin- procedure is to establish and list factors (var- guished from the general category of factors. iables) affecting the spatial structure of the Factors are assigned properties that are ac- phenomena under analysis. A less popu- tive and readily controllable, while condi- lar approach is to dwell on the definition tions are seen as rather passive and hard to of a factor and its categorisation, although control (cf. Chojnicki 1998). in this case use is often made (e.g. Rogacki In electoral studies—depending on a so- 1988) of the definition formulated by Cho- cio-political or a spatial context—representa- jnicki and Czyż (1978) for the purposes tives of various disciplines often accuse their of factor analysis employed in spatial stud- opponents of ignoring those dimensions, so ies. In their approach, a factor distinguished there have appeared proposals to solve the http://rcin.org.pl 94 Roman Matykowski and Katarzyna Kulczyńska dispute by differentiating between spatial as 1989, a well-known article by Florczyk et and structural factors (cf. Zarycki 1997). As al. published in the Tygodnik Solidarność Zarycki (1997, p. 49) observes, "one of the announced a "return of history" and mac- ways to differentiate spatial factors from ro-regional divisions, also those still run- structural ones is using the theoretical sche- ning along the former partition boundaries. ma adopted by Rokkan and Lipset"; still, There were further elaborations of this ap- when discussing studies of conflicts carried proach to macro-regional heterogeneity, e.g. out in accordance with this schema, he had conceptions of a historical background (cf. to admit that the effect of a structural fac- Bartkowski 2003), civilisation-determined tor can "turn into a spatial factor connected macro-regions and civilisation-determined with the specificity of a place" (p. 51). split of the country, or electoral "geology" That is why in geography attempts have (Kowalski 2003). A more comprehensive been made to approach factors in two ways. model of the effect of historical determinants First, a factor is treated as a pre-theoretical (events in earlier ages treated as analogous kind of notion and as such is ascribed some to sedimentation layers) on modern political impact, e.g. factors in migration models, culture in Poland was presented by Zarycki factors of location of an economic activity, (2001). He claims that what inspired him to or factors of urban growth (not always con- seek this type of explanation of political be- nected with concrete theoretical models or haviour patterns was Dogan's (1967) study theories). Secondly, unlike representatives of electoral behaviour in France and Italy. of other disciplines dealing with elections, In the conceptions of society moderni- geographers rely on a chorological approach sation, the most important category of fac- and often seek to establish the force of im- tors—from a micro-analytic perspective— pact of various factors —e.g. on the level includes those describing the processes of electoral support —on the basis of the co- of both urbanisation and industrialisation occurrence of variables studied in territorial (cf. Anduiza-Perea 1999, Cześnik 2007). In terms. To assess the strength of co-depend- the conditions of Central and East Euro- ence of the variables, they usually employ pean states, this category of factors includes correlation analysis, although the measure parameters describing their socio-political they also sometimes use is that of geographi- and economic transformation (e.g. the un- cal distribution (a modification of Florence's employment rate). Also in Poland, from the index). What raise doubts, in turn, are at- very first free, democratic Sejm (parliamen- tempts at interpreting the co-occurrence tary) elections of 1991, a factor controlling of electoral variables and socio-economic the variability of regional support for indi- variables in terms of a cause-and-effect re- vidual parties has been the urbanisation lationship. level (cf. Matykowski et al. 1995). It should be noted that in geography the As Parysek et al. (1991) found on the ba- notion of factors often serves to build theory sis of multivariate regression analysis, the or pre-theoriesthat also deal with spatial dif- pattern of differences in regional support for ferences in the electoral behaviour pattern. Tadeusz Mazowiecki in the 1990 presiden- From a macro-analytic perspective on terri- tial election depended in a significant and torial systems ranging from the subregional direct proportion on such social factors as: to the macro-regional level, three groups the share of females, people of the working of conceptions of the impact of factors and age, people from regions with a high propor- determinants can be distinguished: (1) those tion of non-private land, persons with sec- of historical-cultural conditions; (2) mod- ondary and higher education, and persons ernisation conceptions, or those of the ef- with lower-secondary vocational education. fect of some socio-economic factors; and (3) In turn, using the principal components those connected with rivalry and conflicts method to characterise the presidential elec- within a country's political system. As early tions of 1990 and 1995 as well as the parlia- http://rcin.org.pl Polish presidential election of 2010: a study of the power of voters in big and mediam-sized towns 95 mentary elections of 1991 and 1993, Zary- cerning units with a status of independent cki (1997) distinguished two cleavages for communes, but in a few cases (e.g. Nysa, all those variables, consistently interpreted Chrzanów, Wołomin) where a town formed as factors contrasting the electoral system: an urban-rural commune together with the 'town—countryside' and 'the right—the left'. surrounding area, additional calculations Next he employed this method to character- had to be made relying on data from con- ise the regional socio-economic system, and stituencies. distinguished two factors that contrasted it: It is usually assumed that a population 'town—countryside' and 'old regions—new of 20,000 provides the statistical threshold regions'. In this way Zarycki sought to match separating small and medium-sized towns endogenous factors with exogenous ones (cf. Zuzańska-Zyśko 2006). However, Hef- making use of Lipset-Rokkan's approach. fner (2008) argues that most settlement Thus, the factor shaping the alignment units of up to 25,000 population can be char- of voters in Poland at various territorial lev- acterised by the same variables as the group els is competition along the axes of 'right/ of small towns. Hence in the present re- left-oriented parties' and 'urban/rural-ori- search the lower population limit for a medi- ented parties' (cf. Zarycki 1997, Matykowski um-sized town was set arbitrarily at 35,000. 2007). Despite the radical changes in the This step was prompted by the desire to se- Polish political area at the start of the 21st lect the larger of the medium-sized towns century, some political parties—both new while excluding those with fluid properties and old—have still preserved, or assumed, characterising, in Heffner's opinion, both the features of urban or rural groupings. small and medium-sized towns. In December 2009, Poland's 897 towns accounted for 61.0% of the country's total RESEARCH AIM AND ASSUMPTIONS population (after the GUS Regional Re- search Statistics), with 28.7% of the total The aim of this paper is to analyse the effect population living in big towns and 13.3% in of the electorate of big and medium-sized medium-sized ones (the two groups together Polish towns on the results of the presiden- accounting for 42.0% of the total popula- tial election of 2010, and to characterise tion). However, in the parliamentary elec- voter alignments in towns of this size cat- tions of 2007 the residents of those towns egory. The analysis will cover the following made up as much as 48.7% of all voters in detailed issues: the country, and in the 2010 presidential (1) the effect of the hundred biggest towns election, 45.9% of voters in the 1st round and on the national results of the election; 44.3% in the 2nd round.