Colorado Universal Vote-by-Mail

Summary

Colorado's 2013 universal vote-by-mail law is one of the largest natural experiments in US election administration in the modern era, and the quasi-experimental evidence on its effects is now sufficiently robust to be considered settled within the academic literature. The intervention is straightforward in mechanism — instead of requiring voters to request an absentee ballot or travel to a polling place on Election Day, the state mails a ballot to every registered voter — and the design impact is mechanical (lower transaction costs for voting). The size and durability of the turnout effect, however, was contested in advance. Some predicted small or zero effects on the theory that universal mail ballots would mostly shift the channel of voting rather than the population voting. Others predicted asymmetric partisan effects, with Democrats or Republicans differentially benefiting. The empirical answers were: a modest but real turnout increase of 2–3 percentage points in high-salience federal elections, much larger effects (9–14 pp) in low-salience local elections, and effects that have been remarkably partisan-neutral — both Republicans and Democrats vote at higher rates under universal VBM than they did under traditional voting. The largest beneficiaries are low-propensity voters (young people, infrequent voters, voters who recently moved). The Colorado evidence has shaped both policy adoption in other states and the design of pandemic-era voting accommodations in 2020. It remains a useful counterpoint to the assumption that election-administration reforms are either inconsequential or politically zero-sum.

Research question

"Does adopting universal vote-by-mail (mailing a ballot to every registered voter automatically) increase voter turnout, and does it differentially expand turnout across partisan, demographic, and prior-voter-engagement subgroups?"

Methodology

Result

Positive

Estimates from rigorous quasi-experimental studies (Bonica & Grumbach; Pew Charitable Trusts; Stanford's Healthy Elections Project) find approximately +2 to +3.3 percentage point increase in general-election turnout and substantially larger effects (+9 to +14 pp) in low-salience local and primary elections. Effects largest among low-propensity voters and roughly partisan-neutral.

Evidence strength

Moderate

Quasi-experimental design; causal interpretation requires care.

Replication status

Partially replicated