Vote-by-Mail Doesn't Deliver

Michael Slater and Teresa James

June 29, 2007

Teresa James is an election counsel and Michael Slater is deputy director of Project Vote, a national nonpartisan, nonprofit organization that provides research, guidance and technical assistance to voter participation and voting rights organizations.

The shift in partisan control in Congress and in many state legislatures has brought about renewed interest in policies that make voting more accessible or convenient. One policy under consideration is vote-by-mail (VBM). In the U.S. Senate, Ron Wyden, D-Ore., introduced a bill that creates financial incentives to states shifting to vote-by-mail while in the House Susan Davis, D-Calif., introduced a bill requiring states to offer all residents the option of voting by mail. In the states, there have been at least 32 bills introduced in 18 legislatures that propose to replace traditional polling places with all-mail elections in some or all elections. This July, secretaries of state will find VBM on the agenda as they gather in Portland for their annual summer conference.

Oregon, of course, votes entirely by mail, but VBM elections have also been conducted in Alaska, California, Colorado, Florida, Kansas, Minnesota, Missouri, Montana, Nebraska, Nevada, New Mexico, New York, North Carolina, North Dakota, Utah, and Washington. In fact, most counties in Washington now hold elections only by mail.

Supporters of VBM systems point to cost savings, greater counting accuracy, and voter satisfaction. They also believe that VBM will increase overall voter turnout. Some supporters suggest that making voting easier may reduce socio-economic disparities in voter participation. And, in response to critics who suggest VBM elections are vulnerable to fraud, supporters note that Oregon’s VBM elections have been fraud-free.

Based largely on Oregon’s successful experience with VBM, some advocates and policymakers are eager to import all-mail elections to other states and abandon America’s centuries-old practice of polling place voting. We think that might be unwise. Here’s why.

Vote by mail is only as reliable as the mail delivery.
First class mail, as its name might imply, does not treat everyone equally. In fact, it discriminates against low-income communities and dense urban areas where residents move more frequently and not every adult shares the same family name. This bias is codified in U.S.P.S. Domestic Mail Manual, which provides that if “the addressee of certain mail is unknown to the delivery employee, the mail may be withheld pending identification of the claimant.” In Oregon’s Multnomah County (Portland), for example, 6 percent of mail ballots were returned as undeliverable in the 2004 election.

There are many common scenarios in which a mail carrier may not know that a person resides at a given address, particularly when delivering to apartment or condominium complexes. The most common scenario is a new resident. America has a famously mobile population. The most recent U.S. Census Bureau report indicates that “43 million people or 16 percent of the population aged 1 and older living in the United States moved between March 1999 and March 2000.” Minorities, young people, singles, and divorced people moved at above-average rates.

Project Vote is especially concerned about the effects of mobility among people of lower income levels. Twenty-one percent of households with incomes under $25,000 have moved in the past year, compared to 12 percent of households with incomes greater than $100,000. Almost one in three renters moved, compared to one in 11 homeowners.

Other plausible scenarios include an adult child may move in with a parent, a woman in the process of divorce may move into a friend’s home or an elderly adult may move in with an adult child. These individuals are eligible to vote if they have resided at their new address for even a short period of time, but the mail carrier may simply return a mail ballot if the primary resident hasn’t notified the mail carrier of the new occupant.

This situation came into play during the 2006 election in Baltimore, Maryland. Election officials announced intentions to cancel 2,300 new registrations because voter registration cards were returned as undeliverable. Officials concluded that returned cards indicated that applicants had failed to complete their applications accurately. Project Vote staff investigated and, through interviews with mail carriers, learned that the non-forwardable mail that the board of elections used could not be left at an address where the addressee was not known or listed as a resident. Further investigation confirmed that applicants lived at the addresses they provided to election officials.

Vote by mail’s effect on voter turnout is at best neutral, but may favor affluent voters.
VBM supporters suggest that turnout will increase with all mail elections and point to early studies showing an increase in Oregon voter turnout of up to 10 percentage points. Subsequent research contradicts these findings or presents a far more nuanced picture of VBM’s effects on voting. Most recently, political scientists Thad Kousser and Megan Mullin conducted a rigorous analysis of two recent elections in California, where election law allows officials to designate small precincts as VBM precincts for specific elections. The researchers first paired each VBM precinct with a polling place precinct with similar demographics, averaged the turnout in all the VBM precincts and all the polling place precincts, and then compared the results. They found that turnout in VBM precincts were 2.6 and 2.9 percentage points lower than in polling place precincts.

The most recent examination of VBM’s effect on Oregon’s turnout concludes that, far from the initial 10 percentage point increase some researchers found, VBM increased turnout by four percentage points and only in presidential election years.

More important than the size of any increase in turnout is who is voting under VBM who would not otherwise have voted. Here, researchers have reached a consensus. To the extent that VBM increases turnout in Oregon it does so by retaining voters who are occasional rather than habitual voters. Further, these voters are demographically similar to habitual voters. In other words, VBM does nothing to expand the electorate in ways that make it more representative of the voting age population. In fact, as MIT political scientist Adam Berinsky writes: “VBM in Oregon accentuated the stratification of the electorate. Specifically, VBM mobilized those already predisposed to vote—those individuals who are long-term residents and who are registered partisans—to turn out at higher rates than before.”
 
Vote by mail is more susceptible to corruption than voting at polling places.
There is widespread consensus among all but die-hard partisans that there is little polling place fraud (which is why the debate over voter ID requirements is a false one). There are, however, more than a few cases of absentee ballot fraud. Generally speaking absentee fraud reports arise most often in local county or community elections.

Absentee ballot fraud takes four general forms, (1) forging signatures or signing fictitious names; (2) coercing or influencing a vote; (3) vote buying; and (4) misappropriating absentee ballots. Absentee ballot fraud by members of both major political parties has been substantiated in several high-profile civil and criminal cases.

A 1997 Miami mayoral election was overturned by a Florida appeals court in a highly publicized case upon a finding that absentee ballots cast for Republican incumbent Xavier Suarez were tainted by fraud. The court installed Democratic candidate Joe Carollo as the winner of the mayoral race after throwing out all 4,740 absentee votes. As a result of the investigation, 21 Miami residents were accused of acting as false witnesses to absentee ballots. In a similar case, a state judge nullified the results of a 1993 mayoral election in Hialeah and ordered a new vote.

In a blatant instance of vote buying, Democratic and Republican supporters of candidates in a Dodge County, Georgia sheriff’s race were found guilty of paying voters for their absentee ballot votes, each side bidding against the other and operating from tables on opposite ends of the courthouse hall. In a more recent case, two defendants in Tallahatchie County, Mississippi were convicted of absentee ballot fraud for providing money and beer to voters to get them to vote by absentee ballot.

Vote by mail is amenable to manipulation by election officials.
The National Voter Registration Act (NVRA) prohibits election officials from purging voters for failing to vote. It also requires notice letters and a waiting period before officials can purge voters whom they believe have moved. A registered voter should be able to go her polling place on Election Day and find her name on the list, even if she skipped an election. Vote by mail is different. Election officials in many states can decide to mail ballots to only a subset of registered voters, leaving other voters waiting by the mailbox.

This is exactly what happened in Denver recently. Colorado law requires election officials to place voters who have missed a single general election on an “inactive list.” Voters on the inactive list do not receive a mail ballot. In this case, Denver voters who missed the 2006 general election did not receive their ballots in the mail. More than 117,000 voters were left out of the election as a consequence. Ironically, many Denver voters were unable to cast a ballot in 2006 because of the city’s well-reported failure of its electronic poll book system.

Distinguishing between “active” and “inactive” voters and then mailing ballots to only active voters is the practice Oregon, Washington, and California, in addition to Colorado. No federal law protects voters against this administrative sleight of hand.

Conclusion
Thanks largely to Oregon’s experience, many reform-minded advocates and policymakers have become persuaded that vote-by-mail stimulates increased voter turnout with few drawbacks. We think the facts don’t support their arguments. VBM reinforces the stratification of the electorate; it’s more amenable to both fraud and manipulation than voting at polling places; and it depends too much on the reliability of the U.S. Postal Service.

Americans deserve an equal opportunity to participate in democracy; vote by mail doesn’t deliver that.