A Network Comparison of the Diffusion of State Electoral Laws

Abstract

While much research has focused on state adoptions of specific electoral laws, less attention has been paid to the broader structure of election law policies. To address this, we estimate latent diffusion networks using data on the adoption of 19 election policies over three decades, including registration laws (e.g., motor voter, same-day registration, AVR), voting laws (e.g., voter ID), and election reforms (e.g., term limits). These networks reveal patterns in policy diffusion between states, allowing for a comparison of election laws with broader policy diffusion networks to test different diffusion mechanisms. We find clear evidence of policy learning: states with a high Election Performance Index (EPI) score, which reflects election administration quality, are more likely to send out election policy ties, while this measure is unrelated to tie formation in the broader diffusion network. Partisan and ideological factors seem to have little impact on the network structure, and the initiative process does not predict election policy ties either. The results suggest the election law policy network is driven by policy learning as states are more likely to model their election laws off well-performing states and highlight the value of investigating non-policy specific patterns of election law diffusion.