This article analyzes voter turnout in the Czech Republic on a very detailed spatial structure and an extended yearly time series (1994–2018). Its main goal is to examine the spatial dimension of the disparities in voter turnout in local elections at the level of all (more than 6,000) Czech municipalities. To achieve this goal, global and local spatial autocorrelation methods are used. Municipality-level cartographic presentations then provide spatial evidence of highly stable patterns of electoral participation in Czech municipalities. In the long term, there is no substantial inter-electoral change of the clustering of voter turnout in the different municipalities, except for an overall decline of the homogeneity of the clusters with low or high electoral turnout. In short, the article provides an understanding of electoral turnout in Czech local elections that other approaches have not achieved.

The crisis of liberal democracy in the countries of post-communist Central and Eastern Europe has brought issues related to political participation to the fore. Despite some opposing views (Rosema, 2007), most authors agree that political participation is one of the basic indicators of democratic quality, people’s active involvement in the political process is a necessary condition of successful democratic functioning, and low voter turnout is symptomatic of a crisis of democracy (Norris, 2002). A higher level of participation thus makes government activities more responsive to broader segments of the population (Altman & Pérez-Liñán, 2002)—something that can be best achieved when participation is as widespread as possible, because different social groups participate differently in elections and a low level of electoral participation systematically affects less-affluent individuals. This results in unequal influence of different population groups on political decision making and violates one of the fundamental normative assumptions of democracy, namely that every citizen of the democratic polity should have equal influence on political decision making (Dahl, 1989).

Empirical research has shown lower levels of political participation in East-Central European countries. Most citizens of new democracies have traditionally refused to participate in politics and their refusal goes beyond elections (Kostelka, 2014). More importantly, most authors paid particular attention to the so-called first-order elections (Linek & Petrúšek, 2016), and only a much smaller number of studies focused on subnational levels of government (Cancela & Geys, 2016; van Houwelingen, 2017; Kouba et al., 2021), which is also true of the Czech Republic. Finally, although several studies analyzed voter turnout in Czech local elections (Kostelecký, 2011; Kostelecký & Krivý, 2015), some using at least partly GIS methods (Balík et al., 2015; Hájek & Balík, 2020), a complex spatial analysis of local voter turnout in the Czech Republic is still missing. This, however, is also true for studies from other areas across the world, with geographical approaches to local turnout being relatively rare (e.g., Mansley & Demšar, 2015) and most studies concerned with presidential or parliamentary elections are performed at larger-scale aggregations (Darmofal, 2006; Tam Cho & Rudolph, 2008; Kevický, 2020; Fiorino et al., 2021; Manoel et al., 2021).

We focus on the local elections rather than general elections given the growing importance of the local tier in the multilevel governance of many contemporary democracies (Loughlin et al., 2012). The patterns and dynamics of local voting can be considered as a missing link in the study of multilevel elections (Gendźwiłł & Steyvers, 2021), and local democracy plays a very important role in effective daily functioning of local communities (Mansley & Demšar, 2015). At the same time, a subnational comparative method offers several advantages: (1) increasing the number of observations and thus mitigating the limitation of a small-N research design to enable more meaningful and controlled comparisons; (2) strengthening the capacity to accurately code cases and thus make valid causal inferences; (3) better handling the spatially uneven nature of major political processes (Snyder, 2001).

The relevance of studying Czech local elections to scientists as well as the general public, then, is based on several reasons. First, spatial analysis of local turnout addresses the above-mentioned research gap, namely the lack of works using geographical approaches to analyzing local turnout at a lower scale of aggregation. Second, the Czech Republic is a highly suitable case for answering the question whether there is some form of spatiotemporal (in)stability of voter turnout between elections as it belongs to the very small group of European countries with extremely fragmented settlement (municipality) structures (Maškarinec, 2015; Jüptner & Klimovský, 2022), resisting the ongoing worldwide trend to merge municipalities (see van Houwelingen, 2017, p. 409).1

The extremely fragmented settlement structure of the Czech Republic thus, on one hand, entails more demanding data collection and processing but, on the other hand, it offers a considerably higher number of observations than studies in countries with much more concentrated municipality structures, enabling us to put research questions to more reliable tests. At the same time, the methodological advantages of going local are particularly important in spatial econometric analysis because precisely the choice of spatial research techniques is strongly determined by the preferred scale of aggregation. It is indeed the municipality level that provides suitable conditions for applying local analysis methods because at higher levels of aggregation (district, regional, etc.), the methods become devoid of their local nature and highly imprecise results tend to be obtained.

For the above reasons, the aim of this article is to present the possibilities offered by maps, or geospatial analytical methods, for exploring the spatial interaction effects of voter turnout patterns between Czech municipalities. At the same time, an important ambition of the article is to show to what extent a given (low/high) level of voter turnout in a municipality in one election shifts (partly in the form of a contagion effect) into surrounding municipalities in the following elections. Precisely this question can be answered using suitable spatial data analysis techniques that help us not only identify transformations of the clustering patterns of voter turnout in local elections. At the same time, spatial autocorrelation or spatial regression models can show us whether there does exist some level of spatial autocorrelation of voter turnout between a concrete municipality and its surroundings, suggesting the existence of a contagion effect. Indeed, while existing studies of aggregate local turnout find several factors influencing the level of electoral participation,2 with municipal size appearing as a key explanatory variable of local turnout (Dahl & Tufte, 1973; Denters et al., 2014; van Houwelingen, 2017; Górecki & Gendźwiłł, 2021),3 more empirical tests are needed to verify the possible influence of proximity of municipalities with various levels of turnout on the amount of turnout in a concrete municipality, especially ones using spatial (geographical) approaches at a fine-grained level of aggregation.

For instance, Tam Cho and Rudolph (2008) show that political participation in the US is geographically clustered and that this clustering cannot be explained entirely by social network involvement, by individual-level characteristics, or by aggregate-level factors. Their analysis suggests that the spatial structure of participation is consistent with a diffusion or contagion process. And while Darmofal’s (2006) examination of county-level turnout in US presidential elections from 1828 to 2000 did not find clear evidence of a contagion effect, Fiorino et al. (2021) mentioned a kind of spillover effect across regions in their study of turnout in European Parliament elections between 1999 and 2014 (in 155 regions of EU-12).

In this article, we analyze data on voter turnout in Czech local elections on an extended time series between 1994 and 2018. We use the data for all 6,249 Czech municipalities. Former analyses of social-spatial differentiation (Netrdová & Nosek, 2017), electoral behavior (Bernard et al., 2014; Lysek et al., 2021), geoparticipation (Pánek et al., 2021), and women’s descriptive representation (Maškarinec, 2020) emphasized the need to work at the lowest scale possible, since analysis at higher levels of aggregation (regions, districts, etc.) may obscure substantial intra-regional differences, whereas similar spatial disparities can be expected in the case of voter turnout as well. In this article, voter turnout is defined as a share of registered voters in a given municipality. Data on voter turnout are available from the Czech Statistical Office election server.

We use several spatial techniques to study spatial effects and analyze the dynamics of voter turnout in Czech municipalities. First, our exploration of the spatial structure of voter turnout begins with the formal detection of spatial autocorrelation using Moran’s I statistic (Cliff & Ord, 1981). However, Moran’s I is an overall measure of linear association, whose single value is valid for the entire study area. Since the aim of this article is to identify potentially different patterns of electoral participation in local elections within larger units and their transformation between elections, a local indicator of spatial association (LISA) is used to obtain a more detailed insight into the ways voter turnout is clustered throughout the Czech Republic’s territory (Anselin, 1995).

Bearing in mind that our goal is to compare differences in the geographical clustering of electoral participation in local elections, we use both univariate LISA indicators, which can show the clustering of support for voter turnout in one election, and bivariate LISA indicators, which allow us to compare how voter turnout transformed between pairs of elections. In the case of bivariate correlation, we present maps of so-called outward diffusion indicating the ways voter turnout in a given unit in elections t shifts toward voter turnout in surrounding units in elections t + 1 (Anselin, 2005).

The coefficient of correlation based on Moran’s I ranges between −1 and +1. The value of −1 indicates perfect negative autocorrelation, +1 indicates perfect positive autocorrelation, and 0 indicates a random pattern of spatial clustering in the data.4 In other words, when high levels in one unit are accompanied by high levels in neighboring units (or if there are neighboring places with low levels), there is a positive spatial autocorrelation or spatial clustering. Conversely, when places with low levels are surrounded by places with high levels (or vice versa), there is a negative spatial autocorrelation, which helps us identify spatial outliers: cases where the observed phenomenon is spatially random (Fotheringham et al., 2002).5

The LISA indicators are calculated for each unit (municipality), and statistically significant values can be mapped or categorized (by spatial autocorrelation type) into four groups, each represented by one quadrant of the Moran diagram (Figure 1). This helps us identify units with positive or negative spatial dependence—high values of the variable in a unit accompanied by similarly high values in surrounding units (hot spots) or, in contrast, a low-value unit surrounded by units with other similarly low values (cold spots)—or spatial outliers with high values surrounded by low values and vice versa (see Anselin, 1995). In our case, we present the patterns of voter turnout in Czech municipalities in four classes of clusters: high–high (H–H), low–low (L–L), high–low (H–L), and low–high (L–H).

Figure 1.

Moran’s diagram.

Figure 1.

Moran’s diagram.

Close modal

However, before calculating the spatial autocorrelation indicators, we must address one of the methodological issues of spatial data analysis. The problem is that different approaches to operationalizing spatial proximity (in terms of defining neighboring spatial units) may lead to highly divergent results (Unwin & Unwin, 1998). The crucial choice here concerns the spatial weighting function (wij), or construction of the spatial weight matrix (W), which operationalizes the position and proximity of spatial units (Getis & Aldstadt, 2004). There are generally two basic types of weight matrix (W): discrete and continuous (Unwin & Unwin, 1998; Fotheringham et al., 2002). We opted for a spatial weighting scheme with a threshold distance of 10 km, which had been demonstrated as adequate for Czechia’s settlement structure by previous studies (Blažek & Netrdová, 2009; Maškarinec & Novotný, 2020).

However, as the local Moran’s I statistic is useful especially for descriptive analysis of the given phenomenon, and our aim is also to demonstrate a possible contagion effect on local turnout, we use a spatial econometric strategy that helps us integrate spatial interaction in a regression model, the so-called spatial lag model. The spatial lag model incorporates the dependent variable’s spatial effects directly, as an additional independent variable (Anselin, 2002), and the new variable (here, turnout in local elections – turnout [W] variable) indicates the level of spatial autocorrelation of the variable between a given spatial unit (municipality) and its surroundings (neighboring municipalities). As suggested by some authors, spatial lag models are most appropriate when the spatial patterning is a function of neighboring observations and there is evidence that the political behavior is somehow driven by neighbors (surrounding municipalities in our case) (Wing & Walker, 2010; Tam Cho & Rudolph, 2008).

Furthermore, besides the local turnout [W] variable, we use some other factors that may matter for electoral participation.6 First, we include municipality size (logarithm of the number of inhabitants) as most studies confirmed its decisive effect on voter turnout, which considerably declines with rising local population. Second, we use turnout in previous elections—previous turnout variable—as we expect that the degree of turnout in a specific local election reproduces at a similar level (low of high) in the following election.

However, as the previous turnout variable could be an endogenous predictor of the percentage of voters who cast a vote in following elections (i.e., the dependent variable), we control for the effect of this independent variable including its lagged version, too. As we expect citizens to be more likely to go to the polls where their fellow citizens more often go to the polls in the long-term perspective, we include the non-lagged version of the previous turnout (previous turnout [t – 1] variable), as well as its versions lagged by one (previous turnout [t – 2] variable) and two election cycles (previous turnout [t – 3] variable); one election cycle spans four years.

The analysis of voter turnout in Czech local elections shows a considerable stability of electoral behavior in Czech municipalities. After an initial high mobilization of voters in the context of the country’s democratic transition in 1990, when voter turnout in local elections reached 73.24% (compared to 96.79% in parliamentary elections), the first local elections after the splitting of Czechoslovakia (31 December 1992) saw a decline of voter turnout by 11 percentage points (62.26%). Turnout in the following six elections stabilized around the average level of 46.49%, with maximum oscillations of ±2 percentage points (Table 1).

Table 1.

Clustering Statistics of Global and Local Moran’s I for Voter Turnout in Czech Local Elections, 1994–2018

Local Moran’s I (Number of Municipalities)
YearsTurnoutMoran’s IH–HL–LL–HH–L
1994 62.26 0.138 1379 666 331 266 
1998 46.72 0.200 1324 967 450 361 
2002 45.51 0.150 1157 825 458 379 
2006 46.38 0.134 1045 815 452 436 
2010 48.50 0.105 934 660 432 368 
2014 44.46 0.131 1008 764 419 382 
2018 47.34 0.128 982 784 416 401 
Local Moran’s I (Number of Municipalities)
YearsTurnoutMoran’s IH–HL–LL–HH–L
1994 62.26 0.138 1379 666 331 266 
1998 46.72 0.200 1324 967 450 361 
2002 45.51 0.150 1157 825 458 379 
2006 46.38 0.134 1045 815 452 436 
2010 48.50 0.105 934 660 432 368 
2014 44.46 0.131 1008 764 419 382 
2018 47.34 0.128 982 784 416 401 

Source: Czech Statistical Office; own calculation.

More importantly, the analysis of the distribution of turnout using spatial correlation based on global Moran’s I for the same period yielded a positive global Moran’s I value for every election year. This indicated that areas with a similar density (high or low) were clustered. However, while there was a rise of the value of global Moran’s I between 1994 and 1998, the following local elections witnessed generally rather a decrease of turnout clustering.

The subsequent analysis of voter turnout based on local Moran’s I showed that in all elections, most areas had no correlation (between 50.4% in 1998 and 61.7% in 2010). Among the areas that were correlated, with the exception of 2010, most featured the H–H pattern (between 14.9% and 22.1%, with a mean value of 17.9%), followed by the L–L pattern (between 10.6% and 15.5%, with a mean value of 12.5%), the L–H pattern (between 5.3% and 7.3%, with a mean value of 6.8%), and the H–L pattern (between 4.3% and 7.0%, with a mean value of 5.9%).

In addition, while in the case of the H–H pattern there was a permanent decline of the clustering of municipalities with the highest level of electoral participation since the 2002 local elections, the clusters of municipalities with a low turnout (the L–L pattern) indicate some reversal to much more clustering in the last two local elections of 2014 and 2018. Generally, while the share of voters participating in local elections showed a decrease of almost 15 percentage points between the local elections of 1994 and 2018, the share of municipalities in both categories (H–H and L–L) underwent several oscillations only to reach, in the year 2018, a situation not far from that of 1994.

Furthermore, to systematically test the change in Moran’s I over time, we calculated bivariate Moran’s I, as bivariate spatial autocorrelation can help to quantify the assumption of spatiotemporal (in)stability of electoral participation in the long term (Table 2). Here again, we can see an initial rise of clustering in subsequent elections, with the highest value of bivariate Moran’s I between the elections of 1998 and 2002, followed by a continued decline of the level of inter-electoral clustering of turnout, with the lowest value between the elections of 2010 and 2014. However, when we compared the value of bivariate spatial autocorrelation in the entire time period observed, the differences were relatively low, suggesting that the spatial patterns of voter turnout in local elections have been relatively stable over time.

Table 2.

Bivariate Moran’s I Statistic for Voter Turnout in Czech Local Elections, 1994–2018

Spatial Lag of Municipality
Bivariate Moran’s I199820022006201020142018
Municipality 1994 0.156 0.130 0.121 0.107 0.114 0.117 
1998  0.165 0.155 0.135 0.146 0.151 
2002   0.137 0.119 0.133 0.133 
2006    0.117 0.129 0.132 
2010     0.115 0.112 
2014      0.125 
Spatial Lag of Municipality
Bivariate Moran’s I199820022006201020142018
Municipality 1994 0.156 0.130 0.121 0.107 0.114 0.117 
1998  0.165 0.155 0.135 0.146 0.151 
2002   0.137 0.119 0.133 0.133 
2006    0.117 0.129 0.132 
2010     0.115 0.112 
2014      0.125 

Source: Czech Statistical Office; own calculation.

To capture the general stability of the spatial patterns of turnout in Czech local elections, we compared spatial clustering in the local elections of 1998 (the highest level of the local Moran’s I statistic) with the latest local elections of 2018 (the second lowest level of the local Moran’s I statistic); the remaining spatial patterns do offer a similar picture. Consistently with the low values of Moran’s I, the values of local Moran’s I, too, indicate rather lower levels of spatial clustering between positive and negative values of electoral participation; at best, there were 36.7% of municipalities with positive (1998) and 14.2% of municipalities with negative autocorrelation (2006), respectively. The LISA cluster maps, then, visualize the areas with significant spatial clustering of municipalities with above-average (below-average) values of turnout in their local elections as high–high (low–low) clusters, whereas municipalities that differ in their level of turnout from their neighborhood are high–low or low–high spatial outliers.

On one hand, the results of LISA analysis (Figure 2) demonstrate the divergent character of the spatial autocorrelation of electoral participation across the territory. On the other hand, the maps for individual election years do not indicate any substantial change or transformation in inter-municipal turnout between elections, with the exception of a generally decreasing level of turnout clustering in Czech local elections. This conclusion suggests a high stability of turnout between election years, which is nevertheless consistent with the Pearson correlation coefficient values indicating the inter-electoral stability (uniformity) of the spatial patterns of local electoral participation (Table 3).

Figure 2.

Univariate LISA cluster maps of voter turnout in Czech local elections, 1994–2018; logged municipality size, 2011 (natural breaks).

Figure 2.

Univariate LISA cluster maps of voter turnout in Czech local elections, 1994–2018; logged municipality size, 2011 (natural breaks).

Close modal
Table 3.

Inter-electoral Stability (Uniformity) of Spatial Patterns of Voter Turnout in Czech Local Elections, 1994–2018 (Pearson Correlation Coefficient)

199820022006201020142018
1994 0.477 0.439 0.402 0.351 0.351 0.325 
1998  0.688 0.616 0.565 0.545 0.527 
2002   0.678 0.649 0.562 0.544 
2006    0.600 0.589 0.552 
2010     0.621 0.564 
2014      0.621 
199820022006201020142018
1994 0.477 0.439 0.402 0.351 0.351 0.325 
1998  0.688 0.616 0.565 0.545 0.527 
2002   0.678 0.649 0.562 0.544 
2006    0.600 0.589 0.552 
2010     0.621 0.564 
2014      0.621 

Source: Czech Statistical Office; own calculation.

Generally, in 1998, municipalities with high levels of turnout in their local elections were concentrated especially in the extensive area covering almost all parts of the Vysočina Region (with its predominantly rural and dispersed settlement structure). The area’s uniformity is only interrupted by a contiguous set of municipalities surrounding the regional capital of Jihlava. Subsequently, the compact cluster of municipalities with the H–H pattern extends from Vysočina to most surrounding regions (Central Bohemia, Pardubice, South Moravia), but only to a limited extent: it is only in the regional borderland that the majority of municipalities are affected by the spillover. Other parts of the country see only sporadic occurrences of the H–H pattern, especially in the borderland of three regions: Pilsen, South Bohemia, and, to a lesser extent, Central Bohemia.

In contrast to the H–H pattern, the clustering of the L–L pattern was much more evenly spread throughout the Czech Republic’s territory (Figure 1A). Here, too, several extensive areas can be identified where Czech citizens exhibit rather low willingness to participate in local elections. The first such area covers large parts of the country’s northern borderland, the formerly German-inhabited Sudetenland that had been re-populated by new settlers after World War II. This area begins in western and northwestern Bohemia, more specifically in the Karlovy Vary Region, and also includes large parts of the Ústí nad Labem Region, as well as a northeastern section of the Liberec Region, although both latter regions exhibit large areas without any form of spatial autocorrelation. Overall, low-turnout municipalities in both regions are clustered in the country’s borderland, as well as along their shared regional border (as opposed to their borders with other regions, with exceptions); at the same time, though, this cluster also includes both regional capitals (Ústí nad Labem and Liberec) as well as some of the other largest cities. The area of low turnout (the L–L pattern) extends from this territory to the south (Central Bohemian Region) as far as to Prague.

The last large area of low turnout clustering was concentrated in Northern Moravia, including almost the entire Moravian-Silesian Region (slightly extending to the Olomouc Region, including the regional capital of Olomouc) and, especially, the northeastern industrial agglomeration of Ostrava/Karviná/Frýdek-Místek, as well as borderland peripheries of northwestern Moravia. The remaining clustered areas of the L–L pattern (much smaller in size), then, were found in some peripheral (borderland) areas of South Moravia and South Bohemia.

The following three local elections (2002, 2006, 2010) witnessed a decline of Moran’s I values, which was only interrupted in the two most recent elections of 2014 and 2018, albeit that rise of Moran’s I value was only a slight one. However, the changes in Moran’s I values did not translate into a significant overall transformation of turnout clustering patterns across the Czech territory; the only main difference was the long-term decline of turnout clustering in local elections. On the other hand, in the local elections of 2018, too, the large cluster of the H–H pattern remain concentrated in the Vysočina Region, together with a narrow strip on the borders of the regions of Pilsen, South Bohemia, and Central Bohemia.

More importantly, the decline of clustering was much more visible in the case of the L–L pattern. The relatively large area of neighboring low-turnout municipalities that used to cover almost the entire northern Bohemia was reduced to a few smaller clusters. Similarly, there was a considerable decrease of the number of municipalities comprising the compact low-turnout cluster in Central Bohemia, especially around Prague. However, it was as early as in the 2006 elections that Prague ceased to belong to that cluster, just as the metropolis of northeastern Bohemia, Liberec, ceased to belong to the L–L pattern.

As previous analysis (Kostelecký, 2011; Balík et al., 2015) confirmed municipality size as the main determinant of electoral turnout (turnout decreased with growing size), we also present maps of municipality size using the natural breaks method (Figure 2) as well as a LISA map (Figure 3).7 Here, we can see that although the claim of higher turnout in smaller municipalities is generally valid, especially the map showing the clustering of municipalities according to their size (Figure 3) shows a relatively large number of municipalities deviating from this pattern. Once again, for the Vysočina Region, where the degree of clustering of the H–H pattern is the most extensive, the maps demonstrate that clusters of high turnout contain both smaller and larger municipalities; and municipalities deviating from the general pattern occur in some other parts of the Czech Republic as well.

Figure 3.

Bivariate LISA cluster maps of voter turnout in Czech local elections, 1994–2018; univariate LISA cluster map for logged municipality size, 2011.

Figure 3.

Bivariate LISA cluster maps of voter turnout in Czech local elections, 1994–2018; univariate LISA cluster map for logged municipality size, 2011.

Close modal

Furthermore, all the results of bivariate Moran’s I have consistently identified rather low levels of global spatial clustering of voter turnout between elections and its spatial representation in the form of bivariate local clusters (LISA cluster maps) shows much larger levels of regionalization of municipalities with low or high levels of turnout lasting (or shifting) between elections. More importantly, the identified clusters of bivariate local indicators of spatial autocorrelation are largely identical with the maps of univariate LISA indicators for the different municipal elections presented above. A comparison of selected election years at the beginning and end of the observed time period, but also temporally distant pairs of elections, confirms a high level of spatiotemporal stability of electoral participation in the different municipalities—despite their, once again, decreasing number.

Table 4 presents results from the spatial regression models estimating effects on local voter turnout. A basic comparison of the regression models indicates very similar amounts of total variance explained. More importantly, in all models, the effect of most variables is consistent with theoretical assumptions, with very similar effects of independent variables across local elections.

Table 4.

Predictors of Electoral Participation in Czech Local Elections, 2006–18 (Spatial Lag Model)

2006201020142018
Turnout (W0.109 (0.022)* 0.059 (0.024)** 0.182 (0.022)* 0.168 (0.023)* 
Constant 26.790 (1.972)* 21.600 (2.093)* 19.871 (2.040)* 18.326 (1.963)* 
Municipality size (log) −2.888 (0.276)* −1.986 (0.293)* −4.060 (0.294)* −2.888 (0.276)* 
Previous turnout (t – 1) 0.397 (0.013)* 0.373 (0.013)* 0.345 (0.013)* 0.325 (0.012)* 
Previous turnout (t – 2) 0.172 (0.013)* 0.156 (0.014)* 0.168 (0.014)* 0.178 (0.013)* 
Previous turnout (t – 3) 0.046 (0.008)* 0.139 (0.013)* 0.121 (0.014)* 0.136 (0.013)* 
Log-likelihood −22,372.4 −22,420.4 −22,633.6 −22,488.1 
AIC 44,756.7 44,852.9 45,279.3 44,988.1 
BIC 44,797.2 44,893.3 45,319.7 45,028.6 
6,243 6,243 6,243 6,243 
R2 0.528 0.482 0.490 0.474 
2006201020142018
Turnout (W0.109 (0.022)* 0.059 (0.024)** 0.182 (0.022)* 0.168 (0.023)* 
Constant 26.790 (1.972)* 21.600 (2.093)* 19.871 (2.040)* 18.326 (1.963)* 
Municipality size (log) −2.888 (0.276)* −1.986 (0.293)* −4.060 (0.294)* −2.888 (0.276)* 
Previous turnout (t – 1) 0.397 (0.013)* 0.373 (0.013)* 0.345 (0.013)* 0.325 (0.012)* 
Previous turnout (t – 2) 0.172 (0.013)* 0.156 (0.014)* 0.168 (0.014)* 0.178 (0.013)* 
Previous turnout (t – 3) 0.046 (0.008)* 0.139 (0.013)* 0.121 (0.014)* 0.136 (0.013)* 
Log-likelihood −22,372.4 −22,420.4 −22,633.6 −22,488.1 
AIC 44,756.7 44,852.9 45,279.3 44,988.1 
BIC 44,797.2 44,893.3 45,319.7 45,028.6 
6,243 6,243 6,243 6,243 
R2 0.528 0.482 0.490 0.474 

Note: Unstandardized regression coefficients, standard errors in parentheses, statistical significance levels: * p < 0.001, ** p < 0.05.

Source: Czech Statistical Office; own calculation.

In the first step of the analysis, the new turnout [W] variable in the spatial lag model indicates the effect of autocorrelation on variance of the dependent variable. The effect varied in time, more in the direction of moderate long-term growth of the effect on the dependent variable of spatial autocorrelation of voter turnout between a concrete municipality and its surroundings. In all cases, growing autocorrelation had a positive effect on turnout: a unit (1%) growth of spatial autocorrelation between a municipality and its surroundings was accompanied by a turnout growth between 0.06% (the local elections of 2010) and 0.18% (the local elections of 2014), with a mean value of 0.13% for all elections studied. In substantive terms, this evidence of a slightly positive effect of spatial autocorrelation on turnout levels in a given municipality suggests that a low turnout rate as well as high turnout rate is partly driven by a contagion effect, that is, by factors common to both the unit in question and its surroundings (as positive autocorrelation identifies both areas where high-level turnout in one unit is accompanied by high levels in neighboring units and areas where low turnout in one unit is associated with low levels in neighboring places). However, as this effect is only moderate, in contrast to the high levels of inter-electoral stability (uniformity) of turnout in individual municipalities, we can hypothesize that idiosyncratic factors (specific to the municipality in question and different across municipalities) are very important for explaining local-level electoral participation.

Furthermore, our results also confirmed municipal size as the key explanatory variable of local turnout; its negative effect varied to a large extent. This is a starting point for a more detailed investigation as to why this heterogeneity exists and what kind of social processes it relates to. However, perhaps more interesting results relate to the question whether the degree of turnout in a specific local election reproduces at a similar level (low or high) in the following local election. However, as previous turnout would be an endogenous predictor of the percentage turnout in the following elections, we controlled for the effect of this independent variable, including its lagged versions.

We found support for the assumption that historical level of local political participation (voter turnout) influences the degree to which citizens are willing to participate in the local electoral race in the future. Specifically, even if the previous turnout percentage is lagged (regardless of whether by one or two terms), it is still significant in the positive direction. This confirms the importance of local context. We found a decreasing effect of previous turnout in local elections (with growing time since the election), possibly a sign of a gradual change in the electorate (although more in-depth investigation would be required to prove this). In spite of that, the persistent significant positive effect of the variable shows that exposure to a local political culture characterized by higher or, in contrast, lower voter turnout levels generally play an important role in reproduction of that political culture, one which (fails to) encourage voters to cast their ballots to influence the workings of the local political process in their community.

More importantly, as presented above, there is a high stability of turnout in local elections between election years (Table 3), with levels of correlation between consecutive elections consistently above 0.6 (with the single exception of the very first local elections of 1994). Even higher inter-electoral stability (uniformity) of the spatial patterns of electoral participation is exhibited by parliamentary elections, where the levels of correlation between consecutive elections exceeded 0.7 for the years 2002/2006 and 0.8 for 2010/2014.

In contrast, the correlation between turnout levels for parliamentary and local elections only exceeded 0.6 once, in the year 2002 (parliamentary and local elections were held in the same year from 1998 to 2010), and declined consistently to levels below 0.5 in the last two parliamentary and local elections. These findings suggest that despite some links between the local and national levels of governance, the long-term evolution of local election turnout only partially reflects external factors, and an important role in turnout reproduction is played by local dynamics of political processes in specific municipalities such that may not be fully consistent with national-level political developments.

The main goal of the article was to examine long-term trends and differences in the spatial distribution of voter turnout in Czech local elections. Although the dataset is limited to one country, we believe that the results of our analysis are interesting for general audience as well, as the extremely fragmented settlement structure of the Czech Republic provides a highly suitable context for applying spatial data analysis techniques. Thus, an analysis of spatial distribution using the spatial autocorrelation method was practically applied to study the small-scale level of all 6,249 Czech municipalities for seven local elections between 1994 and 2018. Overall, the mapping of voter turnout resulted in several main conclusions.

The analysis of voter turnout distribution using global Moran’s I showed that the spatial pattern of electoral participation in Czech municipalities proved to be relatively highly stable. At the same time, both the univariate LISA cluster maps (depicting the spatial clustering of and spatial deviations from voter turnout in a single election) and the bivariate LISA cluster maps (that help us analyze the spatiotemporal [in]stability of voter turnout between elections) demonstrated that in spite of the disparate character of the spatial autocorrelation of electoral participation across the Czech Republic’s territory, there is no substantial inter-electoral change of the clustering of voter turnout in the different municipalities, except for an overall decline of the homogeneity of the clusters identified. This suggests a high level of spatiotemporal stability of voter turnout between local elections in the different municipalities, whether they were places of higher or lower participation of citizens in their local politics.

More importantly, the above-mentioned continuing decline of turnout clustering is consistent with existing evidence of slight but long-term approximation of turnout levels between the smallest municipalities and major cities (see Balík et al., 2015, pp. 8–10; Hájek & Balík, 2020, pp. 111–112). Precisely such geography and cartographic comparison at the lowest level of aggregation is a good demonstration of how that approximation is reflected in the spatial dimension, too. The empirical findings based on geospatial techniques thus may also have their substantial local policy implications as our results suggest that the levels of citizens’ political participation in local political decision-making processes are largely endogenous phenomena. This is evidenced (1) by the results of spatial regression (spatial lag models) because, regardless of whether the previous turnout variable was lagged one or two terms, its permanently positive effect (although decreasing with growing time since the previous election) suggested that a political culture characterized by higher civic participation or, in contrast, higher electoral abstention has a tendency to reproduce in time, and (2) in terms of inter-electoral stability (uniformity) of the patterns of electoral participation between local and parliamentary elections. These patterns exhibit a long-term weakening trend, suggesting that external factors and national-level political development only partially affect local turnout while local political environment and its dynamics have a tremendous power to strengthen or erode local turnout—although, given the extent of the topic, a more detailed analysis of the effect size of internal/external factors on patterns of local electoral participation remains a challenge for future research.

Finally, we found some positive spatial autocorrelation of turnout between a concrete municipality and its surroundings, suggesting the existence of a contagion effect. However, as this effect was not too strong and, as shown by the LISA maps, it quickly waned with growing distance, idiosyncratic factors (specific to the individual municipality) seem highly important for explaining local-level electoral participation. So, as one of the central conclusions of our research is that voter turnout attained in a municipality in a specific election (namely as far back as in the early 1990s in many municipalities) reproduces at a similar level in the following election in the same municipality and its close vicinity, rather than diffusing to more remote surrounding municipalities in the form of a contagion effect, future research (using GIS methods) should seek to identify the main reasons why in some parts of the Czech territory the effect of size—that is, that of the country’s extremely fragmented settlement structure—contradicts the pattern of higher turnout in smaller municipalities.

This study was prepared under a grant project supported by the Czech Science Foundation, Grant No. 20-04551 S, “Patterns of Quality of Democracy at Regional Level in the V4 Countries: Looking Inside the Black Box.”

Published online: July 27, 2022

1.

With regard to the structure of Czech settlements, communities with a population under 1,000 account for more than 77% of the Czech Republic’s 6,253 municipalities but only approximately 17% of the country’s total population.

2.

For instance, Cancela and Geys (2016) showed that turnout in subnational elections is primarily affected by size and composition of the geographical unit, concurrent elections, and the electoral system, whereas Kouba et al. (2021) suggested that compulsory voting, concurrent first-order elections, and autonomy of local elected authorities drive turnout in local elections.

3.

Gendźwiłł and Kjær’s (2021) analysis of horizontal variation in the national–local turnout gap confirmed a strong and consistent positive correlation with size, as the gap between turnout levels in local versus national elections increases in larger municipalities.

4.

We do not use specific results from other countries as a benchmark against which the present results could be assessed as the Moran’s I values themselves are not directly comparable between individual countries: they depend upon the weights, which vary with the number of observations (i.e., level of aggregation).

5.

The way Moran’s I is calculated is much like Pearson’s correlation coefficient, and the same applies to interpreting the values of Moran’s criterion.

6.

Regression analysis includes only a limited number of independent variables with regard to the main research question—that is, an effort to examine the spatial dimension of the disparities in voter turnout in local elections—and questions whether low/high level of voter turnout in a municipality in one election shifts into surrounding municipalities in the following elections, partly in the form of a contagion effect. At the same time, to avoid problems with multicollinearity, some other variables were excluded from the analysis, as, for instance, municipality size in the Czech Republic is strongly correlated with some other indicators of party competition, for example, competitiveness (Ryšavý & Bernard, 2013).

7.

We use the logarithm of the number of inhabitants to take account of a considerable gap in population size between a few municipalities and the remaining municipalities.

Altman
,
D
. &
Pérez-Liñán
,
A
. (
2002
)
Assessing the quality of democracy: Freedom, competitiveness and participation in eighteen Latin American countries
.
Democratization
.
9
(
2
),
85
100
.
Anselin
,
L
. (
1995
)
Local Indicators of Spatial Association—LISA
.
Geographical Analysis
.
27
(
2
),
93
115
.
Anselin
,
L
. (
2005
)
Exploring Spatial Data with GeoDaTM: A Workbook
.
Urbana-Champaign
:
University of Illinois
.
Balík
,
S
.,
Gongala
,
P
. &
Gregor
,
K
. (
2015
)
Dvacet let komunálních voleb v ČR
[
Twenty years of municipal elections in the Czech Republic
].
Brno
,
CDK
.
Bernard
,
J
.,
Kostelecký
,
T
. &
Šimon
,
M
. (
2014
)
Existují prostorové kontextové vlivy na volební chování i v relativně nacionalizovaném stranickém systému? Příklad Česka [Are there spatially contextual influences on voter behavior in a relatively nationalized party system? The case of Czechia]
.
Geografie
.
119
(
3
),
240
258
.
Blažek
,
J
. &
Netrdová
,
P
. (
2009
)
Can development axes be identified by socio-economic variables? The case of Czechia
.
Geografie
.
114
(
4
),
245
262
.
Cancela
,
J
. &
Geys
,
B
. (
2016
)
Explaining voter turnout: A meta-analysis of national and subnational elections
.
Electoral Studies
.
42
,
264
275
.
Cliff
,
A. D
. &
Ord
,
J. K
. (
1981
)
Spatial processes: Models and applications
.
London
,
Pion
.
Dahl
,
R. A
. (
1989
)
Democracy and its critics
.
New Haven, CT
,
Yale University Press
.
Dahl
,
R. A
. &
Tufte
,
E. R
. (
1973
)
Size and democracy
.
Stanford, CA
,
Stanford University Press
.
Darmofal
,
D
. (
2006
)
The political geography of macro-level turnout in American political development
.
Political Geography
.
25
(
2
),
123
150
.
Denters
,
B
.,
Goldsmith
,
M
.,
Ladner
,
A
.,
Mouritzen
,
P. E
. &
Rose
L. E
. (
2014
)
Size and local democracy
.
Cheltenham
,
UK, Edward Elgar
.
Fiorino
,
N
.,
Pontarollo
,
N
. &
Ricciuti
,
R
. (
2021
)
Spatial links in the analysis of voter turnout in European parliamentary elections
.
Letters in Spatial and Resource Sciences
.
14
(
1
),
65
78
.
Fotheringham
,
A. S
.,
Brunsdon
,
C
. &
Charlton
,
M
. (
2002
)
Geographically Weighted Regression: The analysis of spatially varying relationships
.
Chichester
,
UK, John Wiley & Sons
.
Gendźwiłł
,
A
. &
Kjaer
,
U
. (
2021
)
Mind the gap, please! Pinpointing the influence of municipal size on local electoral participation
.
Local Government Studies
.
47
(
1
),
11
30
.
Gendźwiłł
,
A
. &
Steyvers
,
K
. (
2021
)
Guest editors’ introduction: Comparing local elections and voting in Europe: Lower rank, different kind…or missing link?
Local Government Studies
.
47
(
1
),
1
10
.
Getis
,
A
. &
Aldstadt
,
J
. (
2004
)
Constructing the spatial weights matrix using a local statistic
.
Geographical Analysis
.
36
(
2
),
90
104
.
Górecki
,
M. A
. &
Gendźwiłł
,
A
. (
2021
)
Polity size and voter turnout revisited: Micro-level evidence from 14 countries of Central and Eastern Europe
.
Local Government Studies
.
47
(
1
),
31
53
.
Hájek
,
L
. &
Balík
,
S
. (
2020
)
Komunální volby v roce 2018
[
Municipal elections in 2018
].
Brno
,
Masarykova univerzita
.
Jüptner
,
P
. &
Klimovský
,
D
. (
2022
)
Vertical and horizontal intergovernmental relations during the first wave of the COVID-19 crisis: Experience from the extremely fragmented CEE countries
.
Local Government Studies
.
48
(
2
),
232
250
.
Kevický
,
D
. (
2020
)
Volebná účasť na Slovensku v roku 2020 z pohľadu geografie – zmena alebo stabilita [Turnout in Slovakia in 2020 in terms of geography – change or stability?]
.
Acta Geographica Universitatis Comenianae
.
64
(
2
),
161
178
.
Kostelecký
,
T
. (
2011
) Politická participace: analýza účasti ve volbách v roce 2010 a jejích determinant [Political participation: Analysis of turnout in 2010 and its determinants]. In:
Čermák
,
D
&
Vobecká
,
J
. et al. (eds.)
Spolupráce, partnerství a participace v místní veřejné správě: význam, praxe, příslib [Cooperation, partnership and participation in local public administration: Meaning, practice, promise]
.
Praha
,
SLON
, pp.
60
83
.
Kostelecký
,
T
. &
Krivý
,
V
. (
2015
)
Urbánno-rurálne rozdiely voličského správania v Česku a na Slovensku [Urban-rural differences in voter behavior in the Czech Republic and Slovakia]
.
Sociológia
.
47
(
4
),
390
413
.
Kostelka
,
F
. (
2014
)
The state of political participation in post-communist democracies: Low but surprisingly little biased citizen engagement
.
Europe-Asia Studies
.
66
(
6
),
945
968
.
Kouba
,
K
.,
Novák
,
J
. &
Strnad
,
M
. (
2021
)
Explaining voter turnout in local elections: A global comparative perspective
.
Contemporary Politics
.
27
(
1
),
58
78
.
Linek
,
L
. &
Petrúšek
,
I
. (
2016
)
What’s past is prologue, or is it? Generational effects on voter turnout in post-communist countries, 1990–2013
.
Electoral Studies
.
42
,
78
90
.
Loughlin
,
J
.,
Hendriks
,
F
., &
Lidström
,
A
., eds. (
2012
)
The Oxford handbook of local and regional democracy in Europe
.
Oxford
,
Oxford University Press
.
Lysek
,
J
.,
Pánek
,
J
. &
Lebeda
,
T
. (
2021
)
Who are the voters and where are they? Using spatial statistics to analyse voting patterns in the parliamentary elections of the Czech Republic
.
Journal of Maps
.
17
(
1
),
33
38
.
Manoel
,
L
.,
Costa
,
A. C
. &
Cabral
,
P
. (
2021
)
Voter turnout in Portugal: A geographical perspective
.
Papers in Applied Geography
.
Available from: doi:10.1080/23754931.2021.1958251
.
Mansley
,
E
. &
Demšar
,
U
. (
2015
)
Space matters: Geographic variability of electoral turnout determinants in the 2012 London mayoral election
.
Electoral Studies
.
40
,
322
334
.
Maškarinec
,
P
. (
2015
)
Nationalisation of the Czech local party system: Case study of the 2010 local elections in municipalities with extended powers
.
Sociológia
.
47
(
6
),
625
656
.
Maškarinec
,
P
. (
2020
)
Spatial dimensions of female political participation: Regional perspective of women’s descriptive representation in Czech local councils, 1994–2018
.
ISPRS International Journal of Geo-Information
.
9
(
12
),
729
.
Maškarinec
,
P
. &
Novotný
,
L
. (
2020
)
Von links nach rechts: Wandel der regionalen Unterstützung für die technokratisch-populistische Bewegung ANO 2011 bei den tschechischen Parlamentswahlen 2013 und 2017 [From the right to the left: Transformation of the regional support of the technocratic populist ANO 2011 movement in Czech parliamentary elections of 2013 and 2017]
.
Austrian Journal of Political Science
.
49
(
4
),
1
14
.
Netrdová
,
P
. &
Nosek
,
V
. (
2017
)
Exploring the variability and geographical patterns of population characteristics: Regional and spatial perspectives
.
Moravian Geographical Reports
.
25
(
2
),
85
94
.
Norris
,
P
. (
2002
)
Democratic Phoenix. Reinventing political activism
.
Cambridge
,
Cambridge University Press
.
Pánek
,
J
.,
Pásztó
,
V
.,
Burian
,
J
.,
Bakule
,
J
. &
Lysek
,
J
. (
2021
)
What is the current state of geoparticipation in Czech municipalities?
GeoScape
.
15
(
1
),
90
103
.
Rosema
,
M
. (
2007
)
Low turnout: Threat to democracy or blessing in disguise? Consequences of citizens’ varying tendencies to vote
.
Electoral Studies
.
26
(
3
),
612
623
.
Ryšavý
,
D
. &
Bernard
,
J
. (
2013
)
Size and local democracy: The case of Czech municipal representatives
.
Local Government Studies
.
39
(
6
),
833
852
.
Snyder
,
R
. (
2001
)
Scaling down: The subnational comparative method
.
Studies in Comparative International Development
.
36
(
1
),
93
110
.
Tam Cho
,
W. K
. &
Rudolph
,
T. J
(
2008
)
Emanating political participation: Untangling the spatial structure behind participation
.
British Journal of Political Science
.
38
(
2
),
273
289
.
van Houwelingen
,
P
. (
2017
)
Political participation and municipal population size: A meta-study
.
Local Government Studies
.
43
(
3
),
408
428
.
Unwin
,
A
. &
Unwin
,
D
. (
1998
)
Exploratory spatial data analysis with local statistics
.
The Statistician
.
47
(
3
),
415
421
.
Wing
,
S. I
. &
Walker
,
J. L
. (
2010
) The geographic dimensions of electoral polarization in the 2004 U.S. presidential vote. In:
Paez
,
A
.,
Le Gallo
,
J
.,
Buliung
,
R
., &
Dall’Erba
,
S
. (eds.)
Progress in spatial analysis: Theory and computation, and thematic applications
.
Berlin
,
Springer
, pp.
253
285
.

Appendix

Figure 1A.

Administrative Division of the Czech Republic. Note: Bohemia in light gray, Moravia in gray, Silesia in dark gray. 1, Capital city of Prague; 2, Central Bohemian Region; 3, South Bohemian Region; 4, Pilsen Region; 5, Karlovy Vary Region; 6, Ústí nad Labem Region; 7, Liberec Region; 8, Hradec Králové Region; 9, Pardubice Region; 10, Vysočina Region; 11, South Moravian Region; 12, Zlín Region; 13, Olomouc Region; 14, Moravian-Silesian Region.

Figure 1A.

Administrative Division of the Czech Republic. Note: Bohemia in light gray, Moravia in gray, Silesia in dark gray. 1, Capital city of Prague; 2, Central Bohemian Region; 3, South Bohemian Region; 4, Pilsen Region; 5, Karlovy Vary Region; 6, Ústí nad Labem Region; 7, Liberec Region; 8, Hradec Králové Region; 9, Pardubice Region; 10, Vysočina Region; 11, South Moravian Region; 12, Zlín Region; 13, Olomouc Region; 14, Moravian-Silesian Region.

Close modal