Data, et cetera



The majority who participate: voting in America

December 19, 2019

Voting Booth
Photo credit: John Keane

Email Facebook Print Twitter

In ten months, Americans will go to the polls to choose their next president. If past presidential elections are any indicator, turnout will be high: the proportion of the voting-age population that voted in the 2018 midterm was an astonishing 11% higher than the one before it. Indeed, experts expect historic turnout in 2020, with one analyst predicting the highest voter turnout since 1908.

Even in a country of more than 330 million people, marginal increases in voter turnout can have significant impact on elections. In 2016, Donald Trump won the presidency with a margin of just 107,000 votes in three states (or less than 1% of all votes cast). In the 2000 presidential election, George W. Bush famously beat Al Gore in Florida with a margin of just 537 votes (out of six million votes cast), contributing to his winning the election.

As a result, voter turnout receives considerable attention from political campaigns, with 'get out the vote' efforts attempting to both increase voter registration and encourage voters to go to the polls on election day. Some states have attempted to increase turnout through automatic voter registration or by accepting ballots through web portals, voting apps, email or fax (although often limited to military members stationed overseas or people with disabilities).

On the flip side, the United States has a long history of voter suppression, from poll taxes and literacy tests, to felony disenfranchisement, voter roll purging and voter caging, and voter intimidation.

So we wanted to look at the data: How many people are voting on election day? What are some of the reasons that prevent people from voting?

How many Americans vote?

Since voter turnout is generally much lower in years between presidential elections, we decided to focus on the most recent presidential election. In 2016, according to the U.S. Census Bureau, there were people 245.5 million people aged 18 or older living in the United States. Of these, 224.1 million were U.S. citizens. Since 1996, the voting by non-citizens in federal elections is both a crime and deportable offense; as such, this analysis is only focused on U.S. citizens.

The Current Population Survey, conducted by the U.S. Census Bureau and the Bureau of Labor Statistics, collects voter data biannually. These data are collected immediately after the November elections and are drawn from interviews in a probability-selected sample of about 60,000 households.

According to this dataset, there were a total of 157.6 million people registered to vote in the 2016 presidential election. This left 66.4 million, or just under 30%, of all U.S. citizens not registered to vote.

Another 20.0 million people (or 13% of all people registered to vote) did not vote, despite being registered and eligible to do so. This meant that overall, 137.5 million registered voters (or 61.4% of U.S. citizens) went to the polls in 2016. This rate is not statistically different from the turnout rate in the last presidential election, and is down from the thirty-year high of 1992 (67.7%).

Voter registration and voting, 2016

Source: U.S. Census Bureau. Voting and Registration in the Election of November 2016. May 2017.

Variation by state: registration and voting

Typically, potential voters may be lost at one of two points: first, they must fail to register to vote. Second, those who are registered may be unable or unwilling to submit their vote.

Geographically, we find that registration rates are not uniform across the United States. A notable outlier, 46% of Hawaiians are not registered to vote. Similarly low, more than one-third of all citizens are unregistered in West Virginia, California, New Mexico, and New York. Regionally, we find that states in the southern half of the country have a higher proportion of citizens who are unregistered.

By contrast, just 18% of citizens are unregistered in the District of Columbia. In Maine, Mississippi, Minnesota, Washington, and Wisconsin, fewer than one-quarter of potential voters are not registered.

Percent of citizens not registered to vote, 2016

Source: U.S. Census Bureau. Voting and Registration in the Election of November 2016. May 2017.

Among people registered to vote, the proportion voting on election day is the lowest in the South and Southeast state. In West Virginia, the state with the lowest voting rate among registered voters, 21% of registered citizens did not go to the polls. Tennessee, Texas, Kentucky, South Dakota, and Alabama each had more than 17% non-voting among those on the voter rolls.

The highest voter turnout rates were in Colorado, Wisconsin, New Hampshire, Wyoming, Maine, and Massachusetts, where fewer than 10% of those registered failed to vote.

Percent of registered citizens that did not vote, 2016

Source: U.S. Census Bureau. Voting and Registration in the Election of November 2016. May 2017.

Why don't Americans vote?

Understanding why the 86.5 million U.S. citizens don't make it the polls is an important first step in increasing voter turnout. Evidently those who are not registered are not eligible to cast a vote. Equally important is understanding why those who are eligible to vote choose not to.

In the 2016 election, there was extensive media coverage of voter suppression, and in particular the impact of voter ID laws. From 2010 to 2016, 20 states passed laws tightening regulations on voting, which were criticized for their disproportionate impact on minority voters. Indeed, in the months leading up to the election, voter ID laws were struck down in North Carolina, Texas, and Wisconsin on the grounds of discriminatory intent.

To understand whether turnout is impacted by laws or policies that make voting more difficult, we turned to U.S. Census data. As part of its data collection, the U.S. Census asks registered voters why they did not vote.

We find that one in four registered voters that didn't vote chose not to because they didn't like the candidates and/or the campaign issues. This is nearly double the rate in 2012, when just 13% chose not to vote because they did not like the candidates. This is unsurprising since the two candidates in 2016, Donald Trump and Hillary Clinton, were both historically unpopular.

Yet the 2016 election is apparently part of a longer-term trend, with the rate of registered people choosing not to vote out of a dislike of the candidates and issues having tripled from 2000 (8%) to 2016 (25%).

The second most commonly-reported reason for not voting, being uninterested, has hovered at about 15% in the last two presidential elections. This reason for not voting has increased from 12% in 2000.

Other reasons for not voting have mostly declined since 2000, with reasons such as being too busy, being out of town, having registratoin problems, or inconvenient polling places declining in relative importance over time.

As such, in 2016, 40% people who chose not to vote did so because they disliked, or were not interested in, the election, candidates, and issues. This is double the rate in 2000, when about 20% of registered voters didn't vote for these two reasons.

Reasons for registered voters not voting, 2000-2016

Source: U.S. Census Bureau. Reasons for Not Voting, by Selected Characteristics. 2000-2016. "Bad weather conditions" not shown (<1% in all years).

What does it mean for the 2020 election?

According to data from the Census, voter turnout has fluctuated over the last three decades from a low of 58% in 1996 to a high of 68% in 1992.

By far, the majority of people that do not make it to the polls do so because they were not registered. We find that once registered, the majority of people do vote.

Indeed, many of the reasons for not voting have been successfully addressed in the past 15 years: fewer people in 2016 did not vote because they had scheduling conflicts, were ill or disabled, were out of town, forgot to vote, or had too inconvenient of a polling place.

The main barriers to voting that remain are two-fold: registration remains a significant bottleneck, with nearly one in three U.S. citizens not eligible to vote. This apparently is a greater barrier in states in the Southwest and Southeast, and remains despite 21 states and the District of Columbia now permitting same-day registration (of which six were enacted after the 2016 election).

Second, voters are increasingly finding the political candidates so uninteresting or distasteful that they are unwilling to vote. This is in part a phenomenon specific to the 2016 election; however, the proportion of people who do not vote because they do not like the candidates or campaign issues has trended upward every year since 2000, more than tripling in the last five presidential elections.

As such, increasing American political engagement will require more than voter drives, early voting, or providing more accessible polling places. It will require removing more barriers to registration and improving the political discourse such that the U.S. electorate feels engaged, optimistic, and empowered enough to vote.




Email Facebook Print Twitter

Copyright © 2019 Data, Et cetera