The Relation Between Covid-19 Vaccination and Voting Trends in Lithuania: A Spatial Analysis

英语. In the face of global challenges such as regional conflicts, climate change and economic recession, the emergence of the Covid-19 pandemic added a unique dimension to the complex landscape that governments had to navigate. Unlike other challenges, the pandemic's solution was theoretically straightforward – the only way to stop it was mass vaccination. However, the issue of vaccine hesitancy by a portion of the population created a division in society, shaping two distinct groups with differing attitudes toward health and civic responsibility. This research investigates the intersection of civic engagement in terms of early vaccination and participation in election. It uses data of Lithuania’s population vaccination patterns, citizens’ voting location and voting results of Lithuania’s 2023 Municipal election. The research firstly looks at the connection between the early vaccinated voters, the level of education and activity in the election. The correlation analysis indicates that individuals with higher education levels are more likely to be early vaccinated with high significance and slightly more active in the election than non-vaccinated voters. The research follows to explore the spatial distribution of the early vaccinated society. Global Moran's I and Median Local Moran's I are implemented to examine the influence of geographical location on the likelihood of people being vaccinated and voting in the elections. It confirmed that both early vaccinated and non-vaccinated voters are spatially autocorrelated. Investigation of clustering geography revealed that early vaccinated voters exhibit a strong urban concentration, while those who voted without vaccination show clustering in circles around cities, contrary to anticipated rural dispersion. Lastly, research aims to predict political preference by location and the fact of vaccination. Using the K-Nearest Neighbours (KNN) method, two major winning parties, Social Democrats (LSDP) and Conservatives (TS-LKD) were analysed with early vaccinated, late vaccinated, not vaccinated voters and higher education level as independent variables. These models exhibited distinct spatial patterns, with TS-LKD support concentrated in urban areas and LSDP in central rural regions with high accuracy level. However, the relatively low Cohen's Kappa values for LSDP indicate the need for model refinement and consideration of additional covariates. Such models can be used as a foundation to train deeper prediction model with more independent covariates and with voting results from earlier elections. In summary, this research offers spatial insights beyond cartographic representation, enhancing our understanding of electoral geography based on vaccination preferences and education level within distinct voting wards. By addressing the gaps in understanding the spatial dynamics of civic engagement during a global pandemic, the research contributes to the broader discourse on electoral geography, public health and societal behaviour. The findings offer policymakers insights into potential interventions, emphasizing the role of education and geography in shaping civic engagement behaviours.

英语. How does early Covid-19 vaccination relate to voting results? By analyzing comprehensive data on Covid-19 vaccinations and the results of 2023 Municipal elections in Lithuania, this research looks at how vaccination timing, education levels, and voting location are connected. In the face of global challenges such as regional conflicts, climate change and economic recession, the emergence of the Covid-19 pandemic added a unique dimension to the complex landscape that governments had to navigate. Unlike other challenges, the pandemic's solution was theoretically straightforward – the only way to stop it was mass vaccination. However, the issue of vaccine hesitancy by a portion of the population created a division in society, shaping two distinct groups with differing attitudes toward health and civic responsibility. This research investigates the intersection of civic engagement in terms of early vaccination and participation in election. The reseacrh uncovers a significant correlation: individuals with higher education are more likely to be early vaccinated and show slightly higher electoral engagement. Spatial analysis highlights patterns with early vaccinated individuals clustering in urban areas, while non-vaccinated voters form circles around cities, challenging rural dispersion expectations. Furthermore, the reseach explores political preferences, employing machine learning techniques to predict support for two major political parties, LSDP and TS-LKD, based on vaccination status and education levels. TS-LKD gains traction in urban areas, while LSDP sees support in central rural regions. However, LSDP lower performance values indicates the need for model refinement. These models lay the groundwork for more accurate predictions, integrating additional factors and historical data for deeper analysis. These findings highlight important social trends in Lithuania, demonstrating how education and health behaviors like vaccination can influence civic engagement and political preferences. This knowledge can guide policymakers in designing targeted public health campaigns and electoral strategies.