Policy Wonk

Electronic voting machines: Brazil

Headline result
+12.3%
share of valid votes after switching from paper to electronic voting
Quasi-experimental
Implemented by
Superior Electoral Court of Brazil (TSE)
Type
National government
Where
Brazil
Region
Latin America & Caribbean
When
1994–2002
Evidence
Quasi-experimental

The challenge

Brazil's paper ballots were hard to fill in for people with little education, and a quarter or more of voters cast residual votes, error-ridden or uncounted.

The case

Brazil phased in electronic voting over three elections, in 1994, 1998 and 2002. In 1998, because devices were limited, only municipalities with more than 40,500 registered voters used the new machines, which let economist Thomas Fujiwara compare places just above and below the cutoff. Electronic voting reduced error-ridden and uncounted votes and effectively enfranchised millions of mainly less educated citizens, about a tenth of the electorate. Using the staggered timing across states, he found it raised the share of state budgets spent on health care by 3.4 percentage points, 34% more spending over eight years. Less educated mothers made more prenatal visits, and low-weight births among them fell 6.8%.

Why it worked

Making it easier for poorer, less educated citizens to cast a valid vote shifted government spending toward services they need. It's a striking example of how voting technology can change who is effectively represented.

Sources