Does Early Voting, Absentee Voting, and Vote-By-Mail Increase Turnout?

Image source: Gotham Gazette. November 7, 2018. “Voter Turnout Booms in New York.” http://www.gothamgazette.com/state/8056-voter-turnout-booms-in-new-york

Among the countless stories that inevitably come out of elections is reporting of voter turnout. Throughout this election season, a myriad of stories focused on enthusiasm from both Democrats and Republicans and predicted that a surge of voter turnout was coming this year. Such predictions came to fruition, as an unusually high rate of voters for a midterm election came out. Inevitably, some of this was due to opposition against (and support for) Donald Trump.

A couple things about turnout are important to note. For one, there are three major ways that turnout is calculated. One way is to divide the number of voters that turned out by the total voting age population, which includes people who are old enough to vote but have not registered as voters and would be allowed to. The second way to calculate turnout is to divide the number of voters that turned out by the voting eligible population. Not everyone who is over 18 years old is allowed to vote, one reason being because some are in prison. Someone who is, for example, trying to analyze how involved Americans are as a whole will find use in this calculation.

But the turnout calculation that makes most sense by far in most situations is to divide the number of voters that turned out by the total amount of registered voters. It is the most reasonable way because registered voters are the only ones who may actually vote. These distinctions are important to keep in mind, as some resources may use one or another way to calculate turnout.

The other thing to remember about turnout is the midterm turnout levels are lower than in presidential elections. Some people only care about voting for president, so they aren’t going to show up for a midterm. Rough approximations of voter turnout in presidential elections (I’m using the number of total votes divided by registered voters) is 75% and in midterms it is 50%. Of course, there will be variations from election to election, state to state, and county to county. But those approximations are a good ballpark to start off at.

So what else could have contributed to the high turnout this year? Early voting also saw a large increase, but that has been a natural trend in most places. It is a similar case for absentee voting. Did those help in increasing overall turnout this year?

Perhaps it contributed a little bit. But there are skeptics as to whether these electoral reforms increase voter turnout in general. That is where we turn our attention to next.


Do Electoral Reforms Really Increase Turnout?

Electoral reforms like early voting were first being enacted in the 1970s. The idea behind these reforms was to give people the opportunity to vote in the election that otherwise would not be able to. Some people have to work on Election Day, have no means of transportation, are disabled and/or poor, etc. As a result, early voting, absentee voting, vote-by-mail, and eventually (but relatively very few instances of) Internet voting. These reforms were also intended to reduce socioeconomic biases and increase demographic representativeness by helping out lower-income people.¹

Did these reforms do what they were designed to do? According to political scientist Adam Berinsky, barely anything if at all. Instead of bringing in new voters as intended, Berinsky argues that they retain voters who are already more likely to come out or have already demonstrated some consistency in turnout.

Before we go into his findings, let’s take a look at how much these electoral reforms are used today. Note that some of this has changed since 2005, when Berinsky’s article was published.

  • Early Voting: Early voting is currently allowed in 34 states and D.C.² Early voting periods set aside time ahead of Election Day for voters to fill out ballots. The days that are set aside vary by state, and within states they can vary by county.
  • Absentee Voting: There are two types of absentee voting — excuse and no excuse. No-excuse states allow anyone to request an absentee ballot for any reason, while excuse states require voters to give a reason why they are requesting one (e.g. they are in the hospital or their job prevents them from going to the polls otherwise). An excuse is required for 20 states, while 27 states and D.C. do not require any excuse for requesting an absentee ballot.³
  • Vote-By-Mail: The term is self-explanatory. There are 22 states that provide for certain elections to be conducted completely by mail. There are three states, however, that conduct all of their elections entirely by mail — Oregon (2000), Washington (2011), and Colorado (2013).⁴ At the time of Berinsky’s writing, only Oregon had implemented this system.
  • Internet Voting: Internet voting was very rare when Berinsky wrote, and it still isn’t widely used. Some states, however, allow for voters to return ballots by email or fax, though often it is limited only to certain voters.⁵ ⁶ ⁷

So what did Berinsky find? His results are disheartening for those who believed electoral reforms would lead to higher turnout rates.

  • Early Voting: There is a potential slight increase in turnout, but demographic representativeness was not increased. His Texas data indicated no difference between early and Election Day voters, and in Tennessee early voters were more politically engaged and partisan.¹ The politically engaged and partisan voters already are more likely to vote anyway. Very partisan voters also usually have their mind made up early, as they are going to vote mostly or entirely for their party up and down the ballot. Thus, early voting makes sense to them.
  • Absentee Voting: ANES data indicated that these voters tended to be older, more educated, and politically active.¹ All of these groups are already more likely to vote anyway compared to younger, less educated, and non-politically active voters.
  • Vote-By-Mail: Since Oregon was the only state doing all vote-by-mail at the time, Berinsky used data from that state, breaking voters up into three groups — traditional (Election Day), Vote-By-Mail (VBM), and nonvoters. While there were slight differences between traditional and VBM voters, they were far more alike than they were to nonvoters. VBM also increased turnout among white, higher-income, older, and more educated voters.¹ All four of these groups are more likely to vote compared to non-white, lower-income, younger, and less educated voters. Thus, VBM had not only failed to increase demographic representativeness, but decreased it.
  • Internet Voting: Berinsky admits that this is mostly speculative since there were so few examples. While Internet voting did seem to increase turnout, another study Berinsky cites concluded that social class bias was increased.¹ This was likely because lower social classes are not as likely to have access to Internet.

Given that Berinsky’s article was published in 2005, it’s quite possible that there is new data that would contradict his conclusions. To my knowledge, there hasn’t been a study since that has looked at this, so for the sake of argument let’s assume that his conclusions still hold true. If the electoral reforms of early voting, absentee voting, vote-by-mail, and Internet voting as a whole (in certain places the goals may have been accomplished) do not increase turnout or reduce the socioeconomic biases of who votes, where does that leave us?


How Do We Increase Turnout?

Berinsky does give some suggestions for how turnout may be increased. He says that we should “reframe the puzzle of participation” by focusing on how to make people want to vote. He lists three specific ideas.

  • “Perhaps political activists could present political grievances in ways that engage new groups of citizens…”¹
  • “…elites could reframe issues in ways that speak more directly to the concerns of the disadvantaged.”¹
  • “…political leaders and political party leaders could expand efforts at mobilization in ways that more directly appeal to the political interests of ordinary citizens.”¹

In fairness, trying to solve how to increase turnout was beyond the scope of his research. In my opinion, while I believe he is on the right track, I also have some criticisms of his ideas. The aforementioned suggestions are sort of vague, but more than that they have already been tried many times— with mixed results at best. For the proposal about expansions of mobilization efforts in particular, that would depend on how much it is in politicians’ best interests. In many races, a campaign strategy is to simply focus on mobilizing their party bases, and political leaders see little political incentive to make broader appeals.

So if the above proposals don’t work all that well either, where does the solution lie?

I would argue that efficacy may be the key issue, especially external efficacy. There are two types of efficacy: internal and external. Internal efficacy refers to how much one feels they know and understand about politics. External efficacy refers to how much one believes they can get government to be responsive to their actions.

One way to measure external efficacy is by examining the trust and confidence people have in government. Gallup, for example, tracks this data and frames it in various ways, such as trust in each branch of the federal government and trust in state/local governments. Pew Research Center also compiles data on external efficacy, such as by directly asking how much influence people believe their votes have on the government. The numbers are not exactly inspiring.

Even if efficacy isn’t the main issue (a lot of people simply just don’t care about politics, either), it is reasonable to assume that increasing both internal and external efficacy would lead to some increase in turnout. More people would feel that they know what they are doing and can impact government actions. How to go about this, however, isn’t clear. Perhaps more of an emphasis on civics is needed. Maybe a reframing of how people can impact government or particularly persuasive arguments for participation is the trick. In any case, anyone looking to increase voter turnout would arguably find efficacy as the best starting point.

  1. Adam J. Berinsky. 2005. “The Perverse Consequences of Electoral Reform.” American Politics Research 33 (July): 471–491.
  2. Ballotpedia. “Early voting.” https://ballotpedia.org/Early_votinghttps://ballotpedia.org/Early_voting (accessed November 8, 2018).
  3. Ballotpedia. “Absentee voting.” http://www.ncsl.org/research/elections-and-campaigns/all-mail-elections.aspx (accessed November 8, 2018).
  4. National Conference of State Legislatures. August 15, 2018. “All-Mail Elections (aka Vote-By-Mail). http://www.ncsl.org/research/elections-and-campaigns/all-mail-elections.aspx (accessed November 8, 2018).
  5. National Conference of State Legislatures. July 15, 2018. “Electronic Transmission of Ballots.” http://www.ncsl.org/research/elections-and-campaigns/internet-voting.aspx (accessed November 8, 2018).
  6. Verified Voting. “Internet Voting,: https://www.verifiedvoting.org/resources/internet-voting/ (accessed November 8, 2018).
  7. eBallot. May 24, 2018. “These States Allow Online Voting for Citizens, Is Your State One of Them?” https://www.eballot.com/blog/these-states-allow-online-voting-for-their-citizens-is-your-state-one-of-them (accessed November 8, 2018).