A voter places a sticker on her sweater after voting on Election Day Tuesday, Nov. 6, 2012 in Bradfordton, Ill. (AP Photo/Seth Perlman)

Today marks the release of the latest edition of the Pew Charitable Trust’s Elections Performance Index (EPI), a measure of how effectively U.S. states administer elections. The news is surprisingly good: there has been a slow but steady improvement in election administration in this country.

This good news flies in the face of the typical stories about election problems — hanging chads, long lines at the polls, voter purges in Brooklyn, precinct consolidation in Maricopa County, Ariz. — to say nothing of claims that election outcomes are “rigged.”

Stories of electoral malfeasance are real and important, of course. But the EPI goes beyond anecdotes to gauge performance across several dimensions of election administration, including the quality of voter registration, ballot casting, and vote counting.

To do so, the EPI relies on 17 indicators, including the average wait time at polling places, voter turnout and registration rates, return and rejection rates of absentee and military/overseas ballots, and the availability of online voter registration and voting information.

These indicators are combined into a composite index that captures the degree to which voting is convenient and secure.

The improvement in election administration documented by the EPI between 2010 and 2014 is not dramatic, but it is real. The states that improved the most were those that have gotten better at simply reporting data. States that fully respond to the Election Assistance Commission’s Election Administration and Voting Survey, and thus provide a more complete picture of their election administration performance, are rewarded in the EPI.

Another way that many states improved was by adopting online tools that make voter registration easier. For instance, the number of states offering online registration grew from eight in 2010 to 21 in 2014.  The fraction of people saying they failed to vote because of voter registration problems, perhaps not surprisingly, fell between 2010 and 2014.

The following map shows the rankings of states — with darker green signifying better performance.


Map showing relative ranking of states in the 2014 Elections Performance Index released by the Pew Charitable Trust. States in dark green are ranked highest; states in dark red are ranked lowest. Shades of green and red indicate intermediate rankings.

The top states in 2014 were (in order) North Dakota, Minnesota, Virginia and Wisconsin. These states have many similarities. Each had relatively few complaints by voters about registration or absentee ballot problems. In each of these states, a large proportion of absentee ballots were returned; among those that were returned, few were rejected because of procedural defects. And all four states had a set of best-practice online tools for voter convenience.

The worst-performing states are more diverse. Alabama performs poorly because it has one of the country’s lowest turnout rates, on top of simply not reporting enough data to calculate how well they handle absentee and overseas ballots.

California is another consistently poor performer, even though it does a good job of reporting data. Unfortunately, what the data show is not always positive. In 2014, California rejected domestic and military absentee ballots at rates that were nearly tops in the nation.   Voter registration and turnout rates were also well below the national average.

Of course, the EPI does not capture everything. It doesn’t not take into account the accuracy of vote counts or voter registration rolls.  These omissions are not for a lack of trying, but due to the lack of comprehensive, reliable data to measure these areas.

Those who champion particular policy choices, such as strict voter identification laws or Election Day registration, will be disappointed that the EPI doesn’t explicitly take those into account either. This is because the EPI tries to assess election administration based on its outputs, not its inputs. If states’ policy choices affect whether eligible voters register or turn out, then the EPI does capture the consequences of laws.

This new release of the EPI updates the index through the 2014 November election. So we cannot say definitively whether states are ready for 2016. Nevertheless, state rankings have remained stable across the years, and so the states who excelled at election administration in 2014 will likely do so in 2016.

The 2016 election raises a few issues worth watching. One is the concern raised by the Presidential Commission on Election Administration that there is an impending crisis in the state of voting machines. This would show up as a spike in the residual vote rate, that is, the rate of over- and under-voted ballots. This is one of the components of the EPI in a presidential election year.

We should also expect the expansion of voter registration opportunities in 2016, either through new online portals or automatic voter registration, which should show up as increases in registration rates and decreases in complaints about registration problems.

Election administration is the quintessential “out of sight, out of mind” policy.  Despite being fundamental to the very functioning of democracy, its needs are apparent to the public only every couple of years — and even then, only when a high-profile election is close or a crisis besets an election administrator. By contrast, the performance of other high-priority government services, such as education, police protection, and roads, are in front of citizens every day.

The EPI offers a more comprehensive depiction of each state’s election infrastructure. The ultimate influence of the EPI will be measured by whether policymakers improve elections in those states that perennially fall at the bottom of the index.

Charles Stewart III is the Kenan Sahin Distinguished Professor of Political Science at MIT, co-director of the Caltech/MIT Voting Technology Project, and co-editor (with Barry Burden) of The Measure of American Elections. Stephen Pettigrew is a doctoral candidate in the Department of Government at Harvard University.

Note: Stewart has been involved with the conceptual and technical development of the EPI as a consultant to Pew and a member of the EPI’s advisory board.  Pettigrew has provided technical assistance through the preparation and analysis of the data that go into the construction of the index.