Voter file data has revolutionized studying American political participation, but it obscures crucial information about nonregistrants. This creates inferential problems when analyzing turnout among registrants using posttreatment registration data.
A new sensitivity analysis approach helps researchers evaluate potential bias from differential registration rates—a phenomenon where changes in eligibility affect registration likelihood differently than assumed. Our framework specifically addresses studies estimating turnout among registrants but is also applicable to those targeting broader voting-eligible populations via secondary sources.
We demonstrate our method with two examples examining how young voter eligibility impacts subsequent turnout. While these cases show eligibility boosts turnout, the results are highly sensitive to differential registration bias assumptions. This reveals that observed effects might not reflect true causal mechanisms if registration propensity differs systematically among eligible groups.
