Until the late 1980s, Americans had two ways to vote: (1) in person on Election Day, or (2) absentee ballots sent through the mail or voted in person at county election departments prior to Election Day. Early voting—in-person voting in a limited number of locations prior to Election Day—was first implemented by a state (Texas) almost 30 years ago and has been pushed by proponents as a way of increasing turnout by making voting more convenient.
But while voters may find early voting more convenient, turnout data show that early voting may actually decrease turnout, not increase it. Early voting raises the costs of political campaigns, since expensive get-out-the-vote efforts must be spread out over a longer period of time. There is also no question that when voters cast their ballots weeks before Election Day, they do so without the same access to knowledge about the candidates and the issues as those who vote on Election Day. When there are late-breaking developments in campaigns that could be important to the choices made by voters, those who have voted early cannot change their votes.
The Development and Legality of Absentee Ballots
In-person voting on Election Day had been the traditional way of voting since the beginnings of the republic. Initially, “the Colonies mostly continued the English traditions of voting by a show of hands or by voice—viva voce voting.” Voting remained “public until 1883 when the States began to adopt the Australian secret ballot,” and the states gradually changed to paper ballots.
Limited absentee balloting started during the Civil War. Wisconsin was the first state to legalize absentee voting in 1862 to allow its soldiers to vote wherever they were stationed. Ohio passed its law allowing absentee soldiers to vote in 1863.
Nineteen states enacted “laws allowing soldiers the right to vote by absentee ballot.” This was particularly important during the 1864 presidential election and explains why 19th-century Democrats generally opposed allowing absentee ballots: “[T]hey suspected loyal soldiers would vote for the party of Lincoln.” Lincoln won 78 percent of the ballots cast by soldiers in 1864, even though their ballots were only a small percentage of the total votes cast. Of the 50,000 Ohio soldiers who voted absentee in 1864, 41,000 voted for President Lincoln.
Absentee voting slowly advanced throughout the 1900s. Today, every state will mail an absentee ballot to a voter who requests one, and those ballots can be returned by mail or in person. At its beginning, states required a reason for absentee balloting, such as an unavoidable absence or a religious observance. Twenty states still require an excuse for absentee voting such as illness or employment that prevents the voter from getting to a polling location on Election Day. No-excuse absentee balloting is only a relatively recent phenomenon. Twenty-seven states and the District of Columbia now allow “no fault” absentee voting: No excuse is required. Three states mail ballots to every eligible voter.
It is important to understand, however, that absentee voting is a privilege, not a right. The U.S. Supreme Court established in McDonald v. Board of Election Commissioners that the “right to vote” does not include a constitutional “right to receive absentee ballots.” States have, through legislation, made absentee ballots available to voters, and the federal government has done the same through federal law for members of the uniformed services and citizens of the United States who reside abroad.
The Development and Legality of Early Voting
Texas was the first state to adopt early voting in 1988. It has now spread to 37 states and the District of Columbia (including three states that mail ballots to all voters).
According to the National Conference of State Legislatures, the early voting period may start from as long as 45 days before an election to as late as the Friday before Election Day. The amount of time provided for early voting ranges “in length from four days to 45 days.” The average is 19 days.
The number of voters casting their ballots through early voting has risen steadily. In the early 1990s, about 7 percent of voters cast their ballots early. In the 2016 election, according to the annual report to Congress of the U.S. Election Assistance Commission, 41.3 percent of all ballots were cast before Election Day. Of the total turnout, 17.2 percent of ballots were cast through in-person early voting and 23.7 percent were cast through by-mail absentee voting.
In the 2016 election, more than 60 percent of the ballots cast in Arizona, Florida, Montana, North Carolina, Nevada, Oregon, and Texas were through in-person early voting. The number of early voting sites in various states varied greatly. There were on “average, 6.1 early voting sites per 100,000 voters.”
As with absentee ballots, the “Constitution does not require any opportunities for early voting and as many as thirteen states offer just one day for voting: Election Day.” In a lawsuit filed by the Ohio Democratic Party, it claimed that the state legislature violated both the Voting Rights Act and the Equal Protection Clause of the Fourteenth Amendment when it reduced the number of early voting days from 35 to “only” 29, even though 29 days of early voting is the tenth-longest among all the states. Because Ohio initially provided 35 days of early voting, the Democratic Party argued that “this prior accommodation…, which also created a six-day ‘Golden Week’ opportunity for same-day registration and voting—established a federal floor that Ohio may add to but never subtract from.”
The Sixth Circuit Court of Appeals dismissed this claim, calling it “an astonishing proposition.” As the court explained:
Adopting plaintiffs’ theory of disenfranchisement would create a “one-way ratchet” that would discourage states from ever increasing early voting opportunities, lest they be prohibited by federal courts from later modifying their election procedures in response to changing circumstances. Further, while the challenged regulation may slightly diminish the convenience of registration and voting, it applies even-handedly to all voters, and, despite the change, Ohio continues to provide generous, reasonable, and accessible voting options to all Ohioans. The issue is not whether some voter somewhere would benefit from six additional days of early voting or from the opportunity to register and vote at the same time. Rather, the issue is whether the challenged law results in a cognizable injury under the Constitution or the Voting Rights Act. We conclude that it does not.
In fact, the undisputed factual record in the case showed “that it’s easy to vote in Ohio. Very easy, actually.” Having only 29 days to vote or a period in which one can register and vote at the same time “can hardly be deemed to impose a true ‘burden’ on any person’s right to vote.” This is particularly true when compared to other states like Kentucky and Michigan that do not allow any early voting and are under no constitutional requirement to do so.
The fact that there was evidence in the case that “some African-American voters may prefer voting” early or “avoiding the mail, or saving on postage, or voting after a nine-to-five day” did not provide a basis for a federal lawsuit or a violation of the law. To the extent the reduction in early voting days impacted such preferences, “its ‘burden’ clearly results more from a ‘matter of choice rather than a state-created obstacle.’” The Equal Protection Clause “simply cannot be reasonably understood as demanding recognition and accommodation of such variable personal preferences, even if the preferences are shown to be shared in higher numbers by members of identifiable segments of the voting public.”
The court rejected the challengers’ claim that Ohio was engaging in invidious discrimination because it was denying them “a more convenient method of exercising the franchise.” Such a claim would “disregard the Constitution’s clear mandate that the states (and not the courts) establish election protocols, instead reading the document to require all states to maximize voting convenience.” The Sixth Circuit warned that under this conception of the federal courts’ role, “little stretch of imagination is needed to fast-forward and envision a regime of judicially-mandated voting by text message or Tweet (assuming of course, that cell phones and Twitter handles are not disparately possessed by identifiable segments of the voting population).”
Similarly, in 2012 a federal district court ruled against a claim that a reduction in the early voting period in Florida was a violation of the Voting Rights Act or the Constitution. Although the parties in the case agreed there “is no fundamental right to an early voting option,” the plaintiffs challenged the reduction of early voting days from 12 to eight. But the fact that more minority voters preferred early voting did “not demonstrate that the changes will deny minorities equal access to the polls.”
Furthermore, many states do not have any form of early voting. The court noted:
[By extending] [p]laintiffs’ theory to its next logical step, it would seem that if a state with a higher percentage of registered African-American voters than Florida did not implement an early voting program a Section 2 [of the Voting Rights Act] violation would occur because African-American voters in that state would have less of an opportunity to vote than voters in Florida. It would also follow that a Section 2 violation could occur in Florida if a state with a lower percentage of African-American voters employed an early voting system…that lasts three weeks instead of the two week system currently used in Florida. This simply cannot be the standard for establishing a Section 2 violation.
In a questionable opinion, however, a three-judge panel of the Fourth Circuit Court of Appeals threw out North Carolina’s reduction in the number of early voting days from 17 to 10 (among other election reforms) on the unsubstantiated claim that it was discriminatory both in purpose and effect. In effect, the Fourth Circuit panel took the “preferences” that the Sixth Circuit said were simply matters of choice of voters—such as whether to vote on Election Day or during an early voting period—and converted them into a legal right.
The Effect of Early Voting
For proponents of early voting who believe that giving voters more time to vote will increase turnout, various studies show that the exact opposite seems to be true: Early voting may actually hurt turnout.
In 2008, American University released a report on the general election that concluded that the efforts of states to increase turnout by implementing different forms of “convenience” voting such as no-excuse absentee balloting and early voting were a “failure.” The campaign of President Barack Obama spurred “the highest general election voter turnout since 1960” and an increase of 2.4 percentage points over 2004. Yet of the 12 states that saw turnout declines in 2008 over the 2004 election, “ten had some form of convenience voting.” Of the 13 states with the largest increase in turnout, “seven had none of the forms of convenience voting.”
These findings by American University corroborated what it had found in prior elections (with the exception of 1998) that states that “adopt these reforms have a worse performance in the aggregate than those which do not.” In fact, “in years of turnout increase, the increases in states with convenience voting…are lesser than the states which have not so adopted. And in years of decrease, the decreases in these states are greater.”
In 2013, another study released by professors from the University of Wisconsin came to a similar conclusion by comparing early voting states to those without early voting. A statistical analysis of turnout in the 2004 and 2008 presidential elections showed that early voting led to lower turnout. The “clearest finding” was that “early voting lowers the likelihood of turnout by three to four percentage points.” In fact, the longer the window of early voting, the greater the effect on lowering turnout.
As the study concluded, this “result is counterintuitive, and it certainly runs against the grain of conventional wisdom.” However, the fact that early voting “actually decreases turnout…[is] an unanticipated consequence that has significant implications for policy and for theories of how state governments can influence turnout.”
As an example, Nevada implemented early voting at the beginning of the 1990s. By the 2000 election, voters in Clark County were casting more votes during the early voting period than on Election Day. Today, twice as many voters in Nevada vote early as vote on Election Day. Yet in the 2016 election, the turnout in Nevada of the voting eligible population was only 57.3 percent, almost 3 percentage points below the national average of 60.2 percent. In 2012, the state’s turnout was 1.6 percentage points below the national average, and in 2008 it was 4.6 percentage points below the national average. As the Las Vegas Review-Journal has pointed out, “Nevada trends don’t look much different” than what the Wisconsin study showed.
Interestingly enough, the Sixth Circuit Court of Appeals pointed out in Ohio Democratic Party v. Husted that the turnout data in that case did not support the claims being made by the challengers that reducing the early voting period in Ohio would reduce turnout, specifically of African-American voters. In fact, the “statistical evidence” from the 2014 election when the reduction was in place ran “directly contrary to the [lower court’s] speculative conclusion that the [law] would have a disparate adverse impact on African Americans’ participation.” Instead, African-Americans registered at higher rates than whites, and their turnout, according to an expert cited by the Sixth Circuit, “either exceeds or is the same as white turnout in Ohio.” Most importantly, the challengers were unable to dispute that those who had previously voted on an early voting day that had been eliminated “were not less likely to vote in 2014 than someone who had voted on a preserved day.”
Similarly, in the North Carolina case where the Fourth Circuit ruled against the state’s reduction of early voting from 17 to 10 days, turnout actually increased while the reduction was in force. As the district court (which had ruled in favor of the state) pointed out, the reduction in the early voting period was in effect in the 2014 primary and general election. In the May 19 primary, the turnout of registered white voters “increased from 15.6% to 17.4%; among registered African American voters, it increased from 11.4% to 13.4%; and among registered Hispanic voters, it increased from 2.9% to 3.3%,” when compared to the 2010 midterm primary election.
The same results were seen in the 2014 general election. In comparison to the 2010 election, “voter participation increased: among registered white voters, it increased from 45.7% to 46.8%; among registered African American voters, it increased from 40.4% to 42.2%; and among registered Hispanic voters, it increased from 19.9% to 20.5%.” In fact, with 10 days of early voting as opposed to 17, African-American turnout increased “more than other groups in 2014,” and the general election “saw the smallest white-African American [sic] turnout disparity in any midterm election from 2002 to 2014.”
Even the experts retained by the challengers in the North Carolina case admitted that early voting does not increase turnout. The district court pointed out that one of the experts opined, in a peer-reviewed publication, that the “research thus far has already disproved one commonly made assertion, that early voting increases turnout. It does not.” Early voting may be more convenient but it “pal[es] in significance to such effects as feelings of citizen empowerment, interest in and concern about the election, and political mobilization by parties, candidates, and other political organizations.”
The court also cited another expert hired by the plaintiffs who wrote that early voting results “in lower net turnout…. Our unambiguous empirical claims are based on multiple data sources and methods: despite being a popular election reform, early voting depresses net voter turnout.”
The reasons that early voting seems to hurt turnout have not been conclusively determined. However, one reasonable inference is that allowing voters to vote over an extended period of time before Election Day has “the effect of diffusing mobilization activities.” Campaigns and political parties spend an enormous amount of time and resources on get-out-the-vote (GOTV) efforts just before Election Day. If those GOTV efforts are spread out over several weeks, they will not have the same intensity and may not be as effective in reminding and convincing individuals to cast a ballot.
The Wisconsin study suggested the same thing, that early voting reduces “the civic significance of elections for individuals” and alters “the incentives for political campaigns to invest in mobilization.” As the report says, “rather than building up to a frenzied election day in which media coverage and interpersonal conversation revolve around politics, early voting makes voting a more private and less intense process.” This lessens the “social pressure” to vote, as well as “guidance on how and where to vote.” All of these “reductions in stimulation—both strategic and nonstrategic mobilization—are greater than the modest positive benefits of additional convenience that accrue largely to those who would vote in any case.” And that seems to be the key factor—early voting just provides more convenience for those who are going to vote anyway instead of stimulating nonvoters to vote.
The Other Dangers of Early Voting
Early voting also poses another danger: “[T]he most significant is the danger that something may occur on the last few days of the electoral season” after tens of millions of citizens have cast an irrevocable vote. Early voters are voting with a different set of facts than those who vote on Election Day:
They may cast their ballots without the knowledge that comes from later candidate debates (think of the all-important Kennedy–Nixon debates, which ran from late September 1960 until late October); without further media scrutiny of candidates; or without seeing how they respond to unexpected national or international events—the proverbial “October surprise.”
A recent example of this danger was demonstrated in a special election for Montana’s lone congressional race in 2017. Just one day before the May 25 election, one of the two candidates, Republican Greg Gianforte, was charged with misdemeanor assault against a reporter for the Guardian newspaper. Two of the state’s largest newspapers, the Billings Gazette and the Missoulian, withdrew their endorsements that same Wednesday evening before the Thursday election. But by that time, 70 percent of Montanans had already cast their vote and had no opportunity—if they thought this incident was important to their choice—to change their votes.
Gianforte won the election by 5.6 percentage points and a little over 21,000 votes. Tom Nichols, a professor at the U.S. Naval War College, wrote a commentary in the New York Times whose title captured the concern that early voting raises: “Now Montana Knows Why Early Voting Is Bad.”
Similarly, 2016 Republican candidate Senator Marco Rubio (FL) dropped out of the Republican nomination race a week before the Arizona presidential-preference primary. Yet because Arizona allows early voting by mail, he still came in third. John Kasich, who came in fourth, was behind Rubio by only 6,339 votes. As CNN put it, Kasich was beaten by “Rubio’s ghost in Arizona,” leading “some to question the utility of allowing weeks of early voting in a highly volatile primary in which candidates tend to abruptly leave the race if they have a poor showing in a key state.” (Rubio dropped out because Donald Trump had beaten him in his home state of Florida.)
The 2016 general election saw three presidential debates between the Republican and Democratic Party nominees, Donald Trump and Hillary Clinton, respectively, starting with the first on September 26, 2016, and the other two occurring in October. That meant that millions of voters throughout the country cast their ballots in early voting states before they had even seen all of the debates between the candidates. As J. Christian Adams of the Public Interest Legal Foundation says, “[E]arly voting produces less-informed voters.” After they cast their early ballots, “they check out of the national debate. They won’t care about the televised debates, they won’t consider options, and won’t fully participate in the political process.”
It also seems straightforward that early voting will increase the cost of political campaigns. When so many citizens vote early, any candidate who limits spending on voter mobilization to the last few days before Election Day (instead of engaging in expensive GOTV efforts during the entire early voting period) will be at a serious disadvantage.
Contrary to what might be expected, studies show that early voting not only does not increase turnout but may actually decrease it. When combined with the fact that it spreads out the voting period over a longer period of time during which voters may be casting their ballots without all of the information about candidates and issues that may become available by Election Day, early voting is an election reform that should be reconsidered by states. Its disadvantages seem to outweigh its benefits.
Curtis Gans, the long-time analyst of American elections who founded the Committee for the Study of the American Electorate, once said that, with the exception of those who cannot physically get to the polls, “the nation would be safer if everyone voted on the same day.” The failure to do so “weakens civic cohesiveness.” Or as another election expert says, “[E]arly voting destroys one of America’s last surviving common cultural experiences—turning out as a single nation on a single day to elect our leaders.”
As American University said in its 2008 report, early voting and other forms of “convenience” voting address a real problem—low turnout—but “with the wrong solutions.” The “participation problem is, at heart, not procedural but motivational.”
The problem is that “in a variety of ways, events, politics, leaders, education, communications, and values have damped the religion of civic engagement and responsibility. We will not get that back by treating would-be voters as spoiled children…. These devices are extremely popular, but popularity is not the same as wisdom[,] and in this case, it is antithetical. It’s time to consider rolling them back.”
—Hans A. von Spakovsky is a Senior Legal Fellow in the Edwin Meese III Center for Legal and Judicial Studies, of the Institute for Constitutional Government, at The Heritage Foundation.