Vladimir Putin’s twenty year long regime has time and time again faced scrutiny and criticism from international election observers and citizens alike. Countless videos of ballot stuffing have emerged on the internet and many suspect that the election results reported by the Russian government are fraudulent. Building on previous election forensics projects studying Russian elections, this thesis aims to expand understanding of the geographic spread of electoral fraud in Russia. Using a stochastic kernel density resampling method, I estimate regional levels of contamination in Russian presidential elections from 2000-2018. The indicator of interest is round integer (multiple of 5) reports of voter turnout and incumbent vote share. Mapping the relative frequency of round integer reports of vote share across the election years, I determine that Russia’s ethnic republics most consistently report abnormally high turnout and vote share. I find that the amount of election manipulation has increased throughout the course of Putin’s regime. Moreover, statistical analysis points to the bifurcation of the electorate, signifying that regions which had previously reported abnormal results are doing so at much higher rates in subsequent election years. This is reflected in the Kernel Density Estimates for each year, which demonstrate abnormal spikes at round integers and show a remarkable shift towards a non-normal distribution over the years of Putin’s regime.