Jackie Pierce is an unlikely warrior for democracy.

A fifty-two-year old mother of three long-since-grown children, she grew up in Flint, Michigan—all four grandparents worked at the former GM plant there—and worked for most of her career as a nursing home aide. Several years ago, she and her husband, Marshall, moved to Pellston, a small village just south of the Mackinac Bridge, which leads to northern Michigan’s Upper Peninsula, to manage a Jet’s Pizza franchise.

On election night in November 2014, Pierce was watching the returns, distressed less about Michigan Democrats losing (yet again) as by the state’s dismal voter turnout: just 41 percent of registered voters. “And then I read on the Internet about Oregon—and how much higher their turnout was. I decided to make a few calls, and find out why.”

What she learned was that Oregon and two other high-turnout states, Colorado and Washington, employ a different electoral system. Instead of requiring voters to cast ballots at official polling places, or apply for absentee ballots, these states mail ballots to all registered voters at their homes. Voters then have two weeks to fill out their ballots and either mail them back or deliver them personally to any one of hundreds of official “ballot drop sites” located strategically across their states—in, for example, schools, libraries, police and fire stations, city halls, county courthouses, post office lobbies, and secure, free-standing metal boxes that are available twenty-four hours a day.

Why shouldn’t Michiganders have the same system, she thought? Several of her friends agreed. So this past summer, they founded a group called Let’s Vote, Michigan! and filed a proposed ballot initiative to amend Michigan’s state constitution to require all future elections, beginning in 2018, to be conducted via mailed-out ballots.

Early this past fall, Pierce bought herself a used 2008 Ford Focus—she calls the color “silver and rust”—and quickly logged more than 12,000 miles speaking before any group in Michigan that would have her, including nearly a dozen Republican central committees. Many of these Republicans, including her own state representative, have told her how much they like the idea. But she’s yet to get any GOP groups or leaders to endorse the concept—and doesn’t expect any. “A lot higher turnout isn’t exactly a big Republican cause these days,” she told me.

She’s also given her spiel to labor unions and more than sixty Democratic central committees in Michigan. A Democrat herself, she hopes to convince her fellow party members that vote by mail could dramatically increase turnout and reverse a decade of near-uninterrupted Republican dominance of Michigan politics.

Yet, surprisingly, her reception from Democratic-leaning groups has been only marginally better. Only a half-dozen Democratic central committees, plus a former lieutenant governor, have officially embraced her proposal. “I think the postal worker and letter carrier unions have endorsed the effort—but I’m still waiting on their official letters,” she explained. Other key Democrats and labor leaders tell Pierce they first need to raise $30,000 to conduct a statewide poll. None of the big national progressive groups fighting for voter rights and increased access to the ballot—like Demos, the Brennan Center for Justice, Common Cause, or Project Vote—have offered any support, either.

“Michigan Democrats have been losing for so long, they just aren’t thinking straight,” says an exasperated Jack Lessenberry, a Wayne State University professor of communications who writes a regular column on government and politics for a chain of local newspapers. “I can’t help but think that instead of putting millions behind other proposals that are doomed to fail, unions and progressive groups might be better off funding vote by mail first.”

Presidential elections still attract a majority of America’s voters. For all other elections, however, democracy is mostly a spectator sport. In the 2014 midterms, just 83 million registered voters cast ballots—a nosedive of almost 10 million from 2010’s already anemic levels. Almost 110 million registered voters were no-shows—for a registered voter turnout rate of just 44 percent. Add another 40 million eligible but unregistered citizens, and the rate was just 36 percent.

Turnout in primary nomination contests is even lower—for instance, just 18 percent of registered voters participated in the 2012 cycle. This is a major factor pulling both parties, but especially the GOP, to the extremes, and it should be especially worrisome to Republican and Democratic moderates and the 42 percent of Americans who now identify as “none of the above.” An estimated 90 percent of the nation’s 435 congressional seats and 7,383 state legislative seats are noncompetitive between the major parties. Win the dominant party primary, and the November election is a mere formality.

Fiscal conservatives, too, have reason to worry about low voter turnout, if for no other reason than the costs that taxpayers in many states are incurring to ameliorate it, such as keeping polls open longer. Voting as traditionally done—with 110,000 polling stations and 800,000 poll workers—is expensive. Just to upgrade or replace the hundreds of thousands of aging touchscreen voting machines could easily cost states and localities $2 billion in the next decade.

Low voter turnout, however, should really trouble progressives, because the voters who don’t show up at the polls (including the ones who vote in presidential years but not in off years) are disproportionately Democratic in their orientation and propensity.

Democrats and their progressive allies aren’t bereft of ideas to boost voter participation. Hillary Clinton and Bernie Sanders are both championing promising reforms, such as automatically registering all American citizens based on driver’s licenses or birth records. But no other solution holds anywhere near the potential to boost actual voter turnout. Evidence from Oregon, Colorado, and Washington suggests that if other states adopted universal vote by mail (UVBM), they could increase their registered voter turnout in midterm elections by 10 to 15 percent. Even more dramatically, they could double or triple their primary election turnout, which would almost certainly reduce the inordinate influence of take-no-prisoners ideologues. (See “Can Vote by Mail Reduce Partisan Extremism?”)

Universal vote by mail has many other virtues, too. Those odious photo ID laws? Rendered moot; you don’t need a voter ID to fill out a ballot at your own kitchen table. Long polling place lines? How about no lines, period—and no way for elected officials to manufacture them by (whoops!) providing too few polling places in certain neighborhoods?

Universal vote by mail has also proven to be at least as secure from fraud—and arguably more so—as traditional voting at polls. Election officials check each voter’s signature on the ballot return envelope, matching it against the voter registration card before the ballot is counted. (Since signatures can change, voters still have time to update their registration cards—and qualify their ballots—before results are officially certified.)

Universal vote by mail has the additional advantage of being less costly to taxpayers than the traditional method. Beginning in 2000, Oregon taxpayers started saving $3 million per election cycle. Or consider California’s San Diego County, where election officials found that in a 2013 special election for mayor the direct cost of operating their polling places—$360,000, for 32 percent of votes cast—far exceeded that of the “mailed out” portion—$84,000, for 68 percent of votes cast.
Getting UVBM adopted in a lot more states needn’t be held hostage to partisan gridlock in Washington and state capitals, either. Indeed, it could become law relatively quickly in twenty-one other states, including Michigan, that allow citizens like Jackie Pierce to put new laws or constitutional amendments on the ballot through the initiative process.

This, of course, presumes that politically powerful groups and individuals can bestir themselves to fund the signature gathering, publicity, and organizing needed for these initiatives to succeed. And that’s not happening—yet—in Michigan, or anywhere else for that matter. The reason is that too many election experts and democracy advocates, the very people who should be championing universal vote by mail, remain unpersuaded at best, and dismissive at worst. As a result, the single most promising reform to strengthen America’s flagging democratic franchise and change the political map is being overlooked in favor of far less effective ideas that likely will prove disappointing.

The problem of low voter turnout is least apparent in presidential contests, where the media frenzy and the billions spent by campaigns, political parties, and outside groups on advertising and get-out-the-vote efforts reach all but the most oblivious of voters. Though far lower than similar elections in most other advanced democracies, America’s 70 percent registered voter turnout in recent presidential races is stellar compared to what happens in midterm and party primary contests.

In 2014, registered voter turnout averaged a paltry 44 percent. About 83 million votes were cast—as calculated by the U.S. Election Project of University of Florida—out of 190 million total registered voters, as the fifty states reported to the federal Election Assistance Commission. The EAC reports have some reliability issues, however, and state voter registration rolls vary in how well they’re maintained and updated. (Some states, for instance, report more total registered voters than they have eligible citizens.)

To compare state-to-state registered voter turnout performance, a better denominator uses the fifty states’ 175 million “active” registered voters (ARVs), separated out from about 15 million “inactive” voters.

“Active” voters cast a ballot at least once every four years; inactive voters may go far longer in between, perhaps because they’ve moved or because they’ve simply given up on politics. Unless noted otherwise, we’ll use active registered voter data for 2014 election results, which produces an ARV turnout value of 48 percent, for both the national average and Michigan’s performance.

The three states with vote by mail, however, did far better than other states. Figure 1 shows ARV turnout for the three UVBM states, along with rates in sixteen battleground states that had a hotly contested U.S. Senate or governor’s race (or both, as in Michigan).

In 2014, Oregon’s active voter turnout rate was 70.9 percent—23 percentage points higher than the national average—despite having no hotly contested top races. Colorado, which had both a close Senate and governor’s race, logged a 71.9 percent ARV rate in its debut UVBM election. Washington State had no U.S. Senate or governor’s race in 2014, but its 54 percent ARV turnout still handily beat the national average. (In 2010, thirty-eight of Washington’s thirty-nine counties used all mail ballots, and its 71.2 percent rate that year was second only to Oregon’s.)

Figure 1. States with Universal Vote by Mail Had Higher Turnout Among Active Registered Voters in 2014 than Battleground States Without UVBM


Although Colorado was the only UVBM battleground state in 2014, the three states’ combined ARV turnout was 9 percent higher than in the sixteen polling place-centric battleground states, and 17 percent above the national average.
Had all other states matched Oregon and Colorado’s 70-plus percent turnout rate for active registered voters, an additional 40 million ballots would have been cast nationally in 2014, making 123 million total. Even at a 60 percent rate, another 22 million more votes—nearly 105 million—would have been counted.

Would those extra votes have changed electoral outcomes in 2014? That depends on the likely composition of those extra voters, of course. Tens of millions of voters who participate in presidential contests fail to vote in midterm and other important elections—and it’s no secret that these “intermittent” voters are far more likely to be younger, poorer, and more racially diverse.

Striking differences between these two electorates are revealed in exit polls conducted by the National Election Pool, a consortium of major news organizations.

In 2012, exit polls showed that the median age among that election’s 130 million voters was forty-seven. In 2014, the median age—of just 83 million voters—was fifty-three. In 2012, 46 percent of the electorate was between the ages of eighteen and forty-four, and just 16 percent was sixty-five or older. In 2014, only 35 percent of votes cast came from eighteen- to forty-four-year-olds, while the share of voters sixty-five or older leaped to 22 percent. Median income was also different: $65,000 in 2012 versus $75,000 two years later.

According to other data from the U.S. Census, black and white voters now vote at roughly the same rates in midterm elections. (In 2012, black turnout was actually higher.) Far more dramatic changes between presidential and midterm elections are evident among the nation’s more than 25 million Latinos eligible as citizens to vote.

According to the U.S. Census Bureau, 11.1 million self-described Latinos said they voted in 2012. But in 2014, the number was just 6.7 million, and there are good reasons to believe that even this latter number is over-reported. (Survey respondents significantly over-report midterm voting behavior; the 2014 survey suggested almost 10 million more votes than were actually counted, with the largest errors occurring among young voters.)

Dr. Matthew Barreto, a professor at UCLA and the cofounder of Latino Decisions, has built what many consider the most sophisticated model for tracking Latino voting patterns, using a wide range of data sources to identify individual Latino registered voters in more than twenty key states. In a March 31, 2015, blog posting, Barreto noted that in 2014, Latino registered voter turnout averaged just 31 percent in these jurisdictions. As Figure 2 shows, Colorado and Oregon were nearly double that rate, at 54.8 percent. (Only Wisconsin, with a much smaller share of Latino voters, did slightly better.)

Far better than exit polls or self-reported census surveys is actual election data from state voter files. While demographic factors such as income, education level, and marital status aren’t part of these public files, all fifty states record age or date of birth. Because so few report even this data, two colleagues and I at Portland State University’s Center for Public Service have acquired and are now analyzing voting patterns by age cohort. Our current sample includes full or partial data from twenty states—including the three vote-by-mail states of Oregon, Washington, and Colorado—across six different contests: the general and primary elections of 2014, 2012, and 2010.

Our first-of-its-kind research is ongoing, but our bottom line finding is this: in states with UVBM, turnout among younger voters (those from eighteen to thirty-four) is dramatically higher—double or more the rate in non-UVBM states. For instance, Figure 3 (page 34) shows the percentage of registered eighteen- to thirty-four-year-olds who voted in 2014 by state. Compare Oregon’s 45.6 percent turnout for this cohort—again, in a relatively dull election year—with rates in battleground states like North Carolina (22 percent), Kentucky (26 percent), Pennsylvania (19.7 percent), and Iowa (27.9 percent).

And consider this: voter turnout of eighteen- to thirty-four-year-olds in the battleground state of Ohio for the 2012 presidential contest was just 3 percentage points higher than Oregon’s midterm rate for that cohort.

Figure 2. 2014 Turnout of Registered Latino Voters in 25 States with High Latino Populations


Turnout among seniors is also higher in UVBM states than in non-UVBM states, though not by much—more on the order of 10 to 15 percent. There’s also good reason to believe that those additional seniors are disproportionately lower-income, less educated, and minority voters, and hence lean Democratic. In the 2010 midterms, 60 percent of eligible seniors voted, according to census data, giving Republicans a 21 percent edge among seniors, according to exit polls. But in 2012, seven million additional seniors voted (about 71 percent of eligible seniors), giving Mitt Romney an overall senior preference of just 12 percent.

This is compelling evidence that UVBM delivers on its key promise: higher voter turnout, especially among the very registered voters who now are most likely to show up in presidential years but skip midterms.

As a practical political matter, few if any state legislatures will enact UVBM anytime soon. Republicans detest the idea; since 2000, only one Republican-controlled legislative body has even allowed a floor vote. Among the forty-seven states that don’t have UVBM, the GOP fully controls both houses in thirty, or one house in seven. And in four more (Illinois, Maryland, Massachusetts, and New Jersey), a Republican governor wields a veto pen.

But in the nation’s twenty remaining initiative states, UVBM reform can be put before the voters whether the political powers-that-be like it or not. There’s still time to qualify the issue to the 2016 ballot in many of these states, though signature deadlines are fast approaching. These include many key battleground states. Among them are Arizona, Ohio, Illinois, Florida, Arkansas, California, Massachusetts, Nevada, Montana, Utah, Idaho, Missouri, Nebraska, and South Dakota.

It would cost $1 million to $2 million to mount a successful signature drive in larger states like Florida, Ohio, and Illinois, but far less in states like Montana, Nevada, South Dakota, and Arkansas. A $10 million effort—the same amount announced in November by a new group, iVote, to pass automatic voter registration laws—could give voters in five or ten states the chance to pass UVBM in time for the 2018 midterm elections.

Figure 3. 2014 Turnout Among 18- to 34-Year-Old Registered Voters in 17 States


In the last decade, dozens of wealthy political donors—on both sides of the political aisle—have stepped forward with that kind of cash to promote various causes, from legalizing marijuana to reforming the prison system. What if just one of them came to the conclusion that American democracy would be reinvigorated by the big surge in voter turnout that UVBM would bring?

As a freshman Democratic state legislator in Oregon, I voted against a 1989 bill to expand vote-by-mail elections from local and special elections to party primary contests. Admittedly, I was swayed by the “crunch of autumn leaves” argument—that vision of walking to my local school on a crisp, fall afternoon, exchanging neighborly hellos, and pulling the voting booth curtain under the watchful gaze of octogenarian poll workers.

I changed my mind after becoming Oregon secretary of state in 1991. Once I grasped the dramatic cost savings from no longer having to operate and staff thousands of traditional polling stations, and realized how much voter turnout had soared in local and special elections—by a factor of three or even five—I saw that I’d been confusing a particular ritual of democracy with its essence, which is participation.

Back then, most of my fellow Democrats instinctively opposed the idea. Republicans were doing far better convincing their voters to use absentee ballots—an even truer reality today, as absentee ballots have always been disproportionately used by older, suburban, and largely white voters. “Wouldn’t giving everyone an absentee ballot just make it even worse for Democrats?” I heard over and over. Such overtly partisan concerns were largely voiced in private; in public, Democratic opponents fretted about potential voter fraud and coercion. (Sound familiar?)

In the wee morning hours of the 1995 Oregon legislature’s final day, a relentless effort by a bipartisan coalition of county clerks and legislators forced a surprise vote on a bill to extend this system to all primary and general elections. Passing with plenty of votes to spare, SB 319 headed to the governor, Democrat John Kitzhaber, for his signature. Celebration quickly turned to mourning, however. The reform’s dead-of-night surprise—combined with the bill’s chief legislative sponsor being Republican Senator Randy Miller, also the state’s Republican Party chair—led Kitzhaber to veto the bill, over our strong objections.

Most national Democrats applauded. “This bill is like giving citizens vitamins—without anything nutritious to eat,” was how then DNC chair Don Fowler explained his opposition to a local reporter.

Three months later, U.S. Senator Bob Packwood suddenly resigned in a sexual harassment scandal. Under Oregon law, the contest required to fill this vacancy was considered a “special election.” The county clerks were unanimous in supporting our decision to make this the first federal election in U.S. history to be conducted entirely with mailed-out ballots.

Oregon voters hit it out of the park. Voter turnout—58 percent of registered voters in the December 1995 primary round; 66 percent in the January 1996 general election—shattered all previous national records for a special Senate election.

However, it was a Democrat— Senator Ron Wyden—who emerged victorious. While some previously skeptical Democrats now warmed to the idea of vote by mail, his narrow victory convinced far more former Republican supporters to do a 180-degree policy turn. In 1997, the GOP-controlled state legislature killed the bill.

It took more than $200,000, and a massive volunteer effort, to qualify Ballot Measure 60—which simply removed the word “not” from the clause “primary and general elections shall not be conducted by mail.” Oregon voters didn’t just say yes; they said, “Hell, yes!” The state passed the measure by a 69 to 31 percent margin in November 1998, receiving resounding support in all thirty-six Oregon counties, Democratic and Republican alike.

In 2000, Oregon held the nation’s first vote-by-mail presidential election. This was the last year Oregon was a battleground state. Had George W. Bush still won Florida but lost New Hampshire—he only beat Al Gore by 7,211 votes—Oregon’s seven electoral votes and its all-vote-by-mail system would have been in the partisan klieg lights. (Gore beat Bush by just 6,765 votes.)

Despite much hyperventilation by critics, Oregon’s 2000 contest was a “no fire, and not even a whiff of real smoke” election. There wasn’t a single credible allegation (much less proof) of attempted individual (much less organized) fraud.

Voter turnout was some of the highest in the nation. But critics didn’t seem to notice, or even care, about the details. A 2001 report by the Caltech-MIT Voting Technology Project, led by the longtime mail ballot critic MIT political scientist Charles Stewart III, recommended that states make it even harder to get mailed-out absentee ballots, asserting that the convenience of vote by mail is “bought at a significant cost to the real and perceived integrity of the voting process.”

Since 2000, Oregonians have been free of polling places for four presidential races, four midterm elections, and eight primary contests. If we assume an average of four more local and special elections every two-year cycle, the typical longtime Oregon voter has now voted in about fifty such elections. This means roughly 100 million ballots have been mailed to Oregon’s registered voters during this period. According to the Oregon secretary of state, there have been just a dozen documented cases of fraud or coercion—none of them organized, much less consequential, and all of them of the single ballot variety. Rounded to the seventh decimal point to the right of a zero, that’s .0000001 percent of all votes cast.

Critics often dismiss Oregon’s success by arguing that the state is a demographic outlier. Yet during the last quarter century, underneath the Portlandia-inspired stereotype of a state overflowing with college-educated, craft-beer-sipping hipsters, Oregon has become a good deal poorer, less educated, and even more racially diverse than most other U.S. states, thanks to a quadrupling of the Latino population. The collapse of the timber industry plunged Oregon’s per capita personal income from above the national average to about 92 percent of it. Childhood poverty rates are now above the national average; Oregon’s food stamp rate, at 20 percent, is the nation’s highest; and Oregon’s high school completion rate is America’s fourth worst.

All of these demographic forces should have driven voter turnout down in Oregon. But they didn’t. Oregon’s last polling place midterms produced just 59 percent registered voter turnout (1998); it rebounded to 69 percent in 2002 and has exceeded 70 percent ever since. During the same period, heavy ballot measure use tailed off—and Oregon lost its “swing state” status. Now reliably blue, it’s no longer targeted by the major parties or presidential campaigns.

Again, contrast Oregon with perennial battleground state Ohio, which has similar middling-to-low national rankings (in the 2010 census) in age, high school graduates, and per capita income, but only a 3 percent Latino population (compared with Oregon’s 12 percent). Oregon’s 2012 presidential election active registered voter turnout beat Ohio by 11 percent; in the 2014 midterms, by 22 percent.

Since 2000, only two other states have followed Oregon’s lead. Universal vote by mail came county by county to Washington State through the leadership of two moderate Republican secretaries of state, Ralph Munro and Sam Reed. Not until 2012 did Washington’s final holdout, Pierce County, throw in the polling place towel. Colorado adopted UVBM in 2013, debuting it for the 2014 election cycle.

That year, of course, the GOP clobbered the Democrats almost everywhere, winning fifteen of sixteen battleground states, adding 300 state legislators, and gaining control of eleven legislative chambers. Democrats lost single state senate races in both Washington and Colorado—which caused both chambers to flip Republican. Colorado Democratic Senator Mark Udall, after running what was widely seen as an inept campaign, lost to the Republican challenger, Cory Gardner. But Colorado’s Democratic governor, John Hickenlooper, eked out a surprise victory, and Democrats retained narrow margins in both the Washington and Colorado houses. In Oregon, Democrats won all statewide races on the ballot and actually increased their majority hold on the state legislature—the only state in which that occurred.

Would expanding UVBM to other states help Democrats more than Republicans? The weight of the evidence certainly suggests so. But that Democratic advantage, such as it is, won’t necessarily last forever. Voter preferences change over time as generations age. A decade ago, for instance, older voters leaned Democratic, and as recently as the 1980s most younger voters supported the GOP.

Moreover, partisan advantage shouldn’t matter at all if UVBM will help democracy and give voters far more ability to cast an informed and considered ballot, on their schedules. How many of us, voting at a traditional polling place, have felt the pressure to rush through the process, picking candidates, especially down-ballot ones, almost at random? Voting by mail, at your kitchen or dining room table, is unhurried. You can use the Internet to learn more about candidates’ policy positions and views, or look for newspaper editorials. You can reach out to knowledgeable relatives, friends, and colleagues who might know more than you do about a particular race. The result is more considered and intelligent voting.

Universal vote by mail can also counter the outsized influence of extreme ideologues who thrive in the “micro-turnout” world of current party primary elections. Official voter statistics our PSU team has compiled from all fifty states’ primaries show that just 35 million voters cast primary ballots during the 2012 election cycle—while 155 million already-registered voters didn’t. That’s an overall registered voter turnout rate of 18 percent. Separate research involving complete voter files for fifteen 2014 primary states shows the median age of those voting at about sixty-two, compared to the median age of forty-six for registered voters in those states. (For more details on our research, see “Can Vote by Mail Reduce Partisan Extremism?”)

If 10 or 12 percent of registered voters in an average-turnout state choose to vote in the dominant party’s primary, the electoral math is pretty clear: win just 5 or 6 percent of your constituents’ votes in a competitive race, in the right primary, and you can pop the champagne corks. But when more than 50 percent of eligible Republicans and Democrats vote in primary elections—Oregon’s track record since 2000—each party’s more moderate voters have a far greater opportunity to be heard.

So let’s do a quick recap. Vote by mail saves money, simplifies elections, and eliminates voting lines. It renders moot the debate over photo ID rules, and lets election officials avoid spending billions on software-enabled (but vulnerable to big-impact hacking) voting equipment. Best of all, it promises double-digit increases in registered voter turnout. So why the deafening silence among Democrats, progressive activists, and even various good-government groups fighting hard for increased voter access and against odious photo ID laws? Why the indifference—even hostility—to this far more direct route to higher turnout?

I’ve been asking these questions for the past twenty-five years. In reporting this story, I’ve immersed myself in the latest research literature on voting, and interviewed experts at various national groups that advocate for expanded voter participation. Here’s what I think are the major hang-ups:

Ignorance—and irrational fear— about voter fraud

Yes, fraud has occurred in elections involving vote by mail. It’s also happened in traditional elections. In both cases, the odds are about the same as your dying next month in a plane crash, caused by a flaming meteorite. No electoral system—including one based on UVBM—can prevent all potential fraud. Nor should it have to. Electoral systems must balance the need for basic integrity safeguards to minimize the odds of invalid ballots being counted and affecting an election
result—for example, 100 percent signature verification in a vote-by-mail system—with the importance of maximizing access to the extent practicable for citizens wishing to exercise their democratic franchise.

Photo ID laws fail miserably on all these counts. Based on no substantive evidence and reliant on smoke-like wisps of dark insinuations, they are rightly condemned by critics as little more than voter-suppression tactics masquerading as heartfelt concerns about “electoral integrity.”

Unfortunately, too many of the same liberal critics fall for the same flawed reasoning when it comes to UVBM. They freak out over theoretical risks of fraud, no matter how unlikely, while ignoring the fact that evidence of actual fraud is vanishingly small. Many progressive voting rights advocates, for instance, point to fraud in absentee balloting as a worrisome indication of what UVBM would bring. President Obama’s presidential election commission found that “fraud is rare, but when it does occur, absentee ballots are often the method of choice.” Certainly, actual examples of fraud related to absentee ballots exist. A 2010 sheriff’s race in West Virginia, for example, involved a few dozen fraudulent ballots; a similar number were involved in a late-2000s Florida school board contest. During the 1990s, several mayoral elections—two in the Midwest, one in Miami—were found to have involved illegally cast absentee ballots. None of these instances involved a presidential, congressional, gubernatorial, or state legislative race.

But there’s an even stronger argument for dispensing with the fraud bogeyman. Mail-based voting systems today are far less risky than most polling place elections, precisely because they distribute ballots (and electoral risk) in such a decentralized way. To have any reasonable chance of success, an organized effort to defraud a mail-based system and its safeguards must involve hundreds (if not thousands) of separate acts, all of them individual felonies, that must both occur and go undetected to have any chance of success.

Contrast that to the risks inherent in polling place elections that increasingly rely on direct-recording electronic (DRE) voting systems and proprietary software systems that both record and tally votes. A single successful software hack potentially could affect thousands of votes. It’s the difference between “retail” fraud and “wholesale” fraud. Or, as one county clerk once put it to me, “Ever wonder why no one bothers to counterfeit pennies? If you’re going to risk the jail time, twenties and hundreds make a lot more sense.”

Confusing absentee balloting with vote by mail

Absentee ballots are also mailed out. Hence the all-too-easy analogy: universal vote by mail is simply “mandatory absentee voting” or “absentee ballots on steroids.” And since absentee ballots today are largely the province of older, white, and/or military—and strongly Republican—voters, wouldn’t UVBM make things worse for the Democrats?

Many progressives (and academic researchers, too) instinctively recoil in such fear because they fail to grasp how profoundly wrong this analogy is. Voters must apply, and qualify for, absentee ballots. (In many states, this requires swearing a legal oath.) This takes time, effort, and foresight. Little wonder that it’s the most highly motivated voters who dominate the ranks of absentee-ballot casters.

Contrast that with the core policy principle underlying UVBM, which can more precisely be understood as “Universal Ballot Delivery (via the U.S. Postal Service).” In three states, it’s already the government’s obligation to send every registered voter their ballot. Voters don’t have to take affirmative action to “connect” with their ballot—nor are they required to “vote” by mail. (In the increasingly popular parlance of behavioral economics, UVBM is a high-participation “opt-out” rather than low-participation “opt-in” system.) While many voters will find a 49-cent stamp more convenient—and often far less costly than gas, bus fare, or taking time off work—they can choose to physically take their ballots (before or on election day itself) to any official ballot drop site.

The voting research we’ve done here at PSU provides additional, and new, evidence. While UVBM increases turnout among older registered voters, as we might expect, it increases turnout even more dramatically for younger registered voters, whose numbers are typically 30 to 50 percent higher in most states than their older counterparts. Today, that net benefit will likely go to Democrats.

Prioritizing voter registration over actual vote casting

Rather than champion universal vote by mail, liberal voting rights activists and candidates like Hillary Clinton and Bernie Sanders have put their chips on other reforms, such as automatically registering every eligible voter. I strongly support automatic voter registration laws too, and am proud that my home state of Oregon pioneered the idea in March 2015, with California following suit in October. The system makes good sense. Oregon election officials estimate an additional 300,000 new registered voters in Oregon; California expects to see more than six million. These new voters will doubtless be disproportionately younger and minority compared to already-registered voters—reason enough to explain why Democrats are so enthusiastic (and even a bit giddy) about this.

But let’s assume new automatic voter registration laws could somehow be passed in enough states to add another 20 million of the nation’s 40-plus million now-unregistered citizens to the voter rolls. That would certainly boost turnout in the 2020 presidential election. But what about the 2018 (and future) midterms and party primaries.

The sad truth is that the vast majority of those 20 million new registered voters would likely join the ranks of the 110 million already-registered voters who now sit out the midterms, not to mention the 155 million who skip party primaries.

Automatic voter registration will get more potential voters to the first step of the voting process—and that’s a great thing. But it’s the much smaller half of the turnout battle. The real payoff lies in getting registered voters connected to their ballots—the second step. Only then is it possible for them to arrive at the real destination: marking, and then voting, their completed ballots (see “Why Vote by Mail Is Better than Early Voting.”).

The tyranny of dated, superficial, and/or irrelevant, academic research

Over the years, many academic studies have purportedly found—or, perhaps more precisely, allowed others to easily conclude—that mail ballot-based systems like Oregon’s don’t, in fact, increase voter turnout. Why even risk more fraud, if UVBM doesn’t—well, deliver the voter turnout mail, so to speak?

Virtually every major study that critics typically cite has one or more of the following flaws. The research is at least eight years old, harkening back to a time before voter turnout in midterm and primary elections took a steep nosedive almost everywhere except Oregon. These studies often only focus on presidential races, or rely on high-margin-of-error census data and complex statistical models often built on absentee ballot-based data. Or they evaluate UVBM’s performance on whether it increases the turnout of unregistered citizens—something it can’t and has never pretended to do. (See “What the Critics Get Wrong About Universal Vote by Mail.”)

Perhaps most telling, only a small handful of these studies focus on Oregon’s fifteen years of experience. Not a single one I’ve found is based on Oregon’s individual voter turnout data, as our preliminary research is.

In 2000, America reached a fork in the electoral road, though it didn’t recognize it at the time. That year’s presidential debacle shined a spotlight on the many aspects of our election system in desperate need of reform: improperly purged voting rolls, the infamous “butterfly ballot,” the even more memorable “hanging chads.”

What followed was a flurry—even a frenzy—of change. Some fixes were long overdue. For example, the federal government, through the Help America Vote Act (HAVA), spent billions of dollars to help every state upgrade its voter registration rolls. But bias and ignorance fueled other changes. The new preference for digital technologies over paper led to the elimination of hanging chads, which were the inevitable result of a poor, long-outdated 1960s-era technology that had lingered far too long in many cash-strapped election jurisdictions. This high-tech bias also led many to abandon or reject far better optical-scan paper-based ballots for largely unproven and proprietary software-based DRE voting machines.

Congress spent billions of dollars to help states and localities buy these machines, but they’re already becoming obsolete and dysfunctional. The state of Virginia recently discovered that 3,000 machines in more than 20 percent of the state were vulnerable to hackers; according to a report by the Brennan Center for Justice, “an external party could access the machine’s wireless features to record voting data or inject malicious data.” The state immediately ordered the machines replaced—at a cost of $10,000 to $12,000 per precinct, amounting to hundreds of thousands of dollars for some Virginia counties.

By abolishing polling places altogether, Oregonians since 2000 have avoided all that. Unfortunately, conventional wisdom seems to be leading the country in a much different direction. Groups like the Caltech-MIT Voting Technology Project, the presidential election commission, and the Brennan Center all advocate a three-part hybrid system of ballot delivery: traditional election day precinct voting; continued (and likely greater) use of absentee ballots; and more “early in-person voting” (EIPV), where voters receive and cast ballots at special voting centers” prior to election day.

So what’s wrong with more voter choice? Almost everything, compared to delivering a ballot to every registered voter—and then providing voters with multiple choices to complete and return them. Doubling down on ballot delivery via traditional polling places—especially ever-more-expensive and software-dependent versions—is a certain prescription for billions in additional (and unnecessary) expense—and likely even further declines in voter turnout. These expensive improvements and their alternatives have grown in the last twenty years. Meanwhile, polling places have only grown more empty and overall voter turnout has decreased. (For more information on EIPV, see “Why Vote by Mail Is Better than Early Voting.”)

The traditional polling place dates back to the fifth century BC, when Athenians began assembling on the Pnyx hill to vote on weighty matters of state. Universal vote by mail would update the technology of democracy to at least the eighteenth century, when the U.S. Post Office was founded. This basic rethinking of how voters connect with their ballots will also facilitate the day—perhaps not too far on the horizon—when we may be able to receive and even cast our ballots, with comfort and sufficiently high levels of security and integrity, using our home computers or smartphones. It’s only by pulling the plug on the polling place—abolishing it completely as a ballot-delivery mechanism, and starting each election period with every voter having his or her ballot already in hand—that meaningful change is going to happen.

But make no mistake: this change will be fought hard—and fiercely—by those who believe that higher electoral turnout will only hasten the day of their demise. In the last twenty years, only one Republican-controlled legislative chamber has even allowed a vote-by-mail bill to reach the floor. It happened in Montana on January 27, 2011, when a surprisingly bipartisan effort resulted in the passage of HB 130 by a 57-43 margin. Literally overnight, fifteen House Republicans had second thoughts, changing their votes from aye to nay, killing the bill. The reason? Sudden concerns about voting fraud and the security of ballot drop sites.

Today, such legislation is under serious consideration in just one state, California. The newly elected secretary of state, Democrat Alex Padilla, is pushing Assembly Bill 450 to give counties this option—a gradualist approach that took a decade to succeed in Washington State. But this July, Padilla agreed to defer action on the bill until at least January, noting concerns being expressed about it—primarily by fellow Democrats.

That leaves Michigan.

Once she formally launches her petition drive—which she’s now pushed back to mid-January—Jackie Pierce and her colleagues will have just 180 days to collect 315,654 valid signatures. Asked what she thinks it will cost, she guesses “at least $1 million,” much of it for paid signature gatherers. She says (quite correctly) that times have changed; you just can’t rely too heavily on volunteers anymore. And how much does her group, Let’s Vote Michigan, have in the bank right now? She laughs. “About $5,500—mostly in $5 and $10 contributions.” I asked Pierce what she hopes she’ll be doing for the November 2018 election. She says she hopes she’ll be voting somewhere other than at a polling place, for the first time in her life. I ask her where that will be. “I hope I’ll be right here, at my dining room table, with my ballot and the Internet close by,” she says. “And I’ll no doubt be arguing with my husband about who I think he should be voting for, too.”

Revolutions can’t succeed on vision and audacity alone. Perhaps it’s time for those who say they want more people to vote to give Pierce a call and ask a simple question: “What can I do to help?

Taylor Woods and Stephanie Hawke contributed to this story.

Phil Keisling

Phil Keisling, a Washington Monthly contributing editor, served as Oregon secretary of state (1991–99) and is currently the director of the Center for Public Service at Portland State University.