We estimate a causal effect of QE1 on quarterly wage income in every county in the U.S including Puerto Rico. For each of the 3,195 counties, we run difference-in-differences regressions with controls for county-level American Recovery and Reinvestment Act transfers, the local tax rate, manufacturing employment shares, population, and time fixed-effects. We then use these estimates to (a) construct a national index of the geography of ZLB-constrained monetary policy effectiveness in the United States, (b) construct regressions with county-characteristics related to industry, education, inequality, and political preference, and© check the robustness of our findings by building VAR models for every county, state, and region in the United States to estimate whether monetary policy has become more or less effective over time. There is substantial variation in counties’ causal effects and we find our coefficients for geographic differences in wage income response to be significant at the 95% level for 2,013 of the 3,195 counties we test. We characterize the types of counties where QE1 was more effective: Counties with lower local tax rates per capita, lower government expenditures per capita, less population loss, less racial segregation, more colleges per capita, more Democrat votes, and more urban areas were more likely to have stronger wage-income responses to QE1. Of particular note, counties with an economic reliance on mining and farming had significantly less wage-income effects to QE1 and counties which changed their Presidential Election vote from Republicans to Democrats from 2008 to 2016 were overwhelmingly counties with stronger wage-income responses to QE1. Ultimately, this paper suggests that monetary policy constrained by the zero-lower bound is significantly less effective at stimulating wage income for some regions in the United States than others, and these distributional impacts can and already have exacerbated the growing divide within our country.