The Neighborhood Effect on Perceptions of Corruption: Comparative Spatial Autocorrelation Analysis

Abstract

This article aims to investigate the factors influencing the perception of high corruption levels in a country and examines the potential existence of spatial dependence in the distribution of these factors. It specifically focuses on exploring the interdependence between the Corruption Perception Index (CPI) and forty four other factors categorized into six distinct categories. The central hypothesis posits that the interdependence between the CPI and these factors is more pronounced at the domestic level compared to its influence on indicators in neighboring countries. To assess the degree of interdependence between the CPI and each of the other indicators, the Pearson's Coefficient of Determination is employed, enabling an evaluation of corruption levels based on domestic state-specific factors. Furthermore, Moran's Bivariate Spatial Autocorrelation Index is utilized to elucidate the extent to which the CPI in one country influences one of the forty-four indicators in neighboring countries. Additionally, the Index of Spatial Interdependence is employed to ascertain the significance of domestic and international factors for each indicator. The research findings provide several noteworthy conclusions. Firstly, the neighborhood effect proves to be particularly significant for indicators that hold universal relevance for all governments, such as demographic and standard of living indicators. Conversely, indicators influenced by institutional, historical, and cultural differences exhibit stronger interrelations within the state. Lastly, the study establishes that the Pearson's Index holds greater significance than the Bivariate Moran's Index of Spatial Autocorrelation and the Index of Spatial Interdependence, thereby confirming the proposed hypothesis.