During the tabulation of votes in the 2000 presidential election, the world was shocked at the technological inadequacy of electoral equipment in many parts of the US. In reaction to public dismay over "hanging chads", Congress quickly enacted the Help America Vote Act (HAVA), legislation to fund the acquisition of advanced vote-counting technology. However, the intention was to enable, rather than mandate, choices of new electoral equipment. This paper takes advantage of a unique historical opportunity to test whether electoral equipment follows the pattern predicted by well-established models of innovation diffusion, merging electoral data with census data on socioeconomic characteristics. We infer that fiscal constraints to acquisition are strong but are not the only limitations to technology adoption, particularly within certain types of easily identifiable populations.