Spatiotemporal dynamics of corruption propagation: A local-level perspective

Abstract

What are the patterns and predictors of corruption across space and time? Existing studies have approached this question either through descriptive statistics or spatiotemporal autocorrelation analysis. Moreover, they have primarily relied on ad hoc predictor sets. This paper, however, implements emerging hot spot analysis and geographically and temporally weighted regression (GTWR) to examine corruption heterogeneity across Vietnam. In so doing, it draws on ecological complexity theory and informal payments data over the 2006–2020 period. Subsequently, emerging hot spot analysis indicates corruption is declining in every Vietnamese region. At the same time, GTWR elaborates on corruption's spatiotemporal heterogeneity: (a) only two predictors (i.e., net migration and proactivity) have consistent positive or negative associations with corruption; and (b) the variance of most predictors' corruption impact is high. Furthermore, whereas labor has the strongest corruption impact in North Vietnam, provincial leaders' proactivity takes precedence in Central and South Vietnam. These findings, then, underscore the significance of incorporating ecological complexity theory into the study of corruption.