Press treatment of low voter turnout in American society has become schizophrenic. On the one hand, the press is giving strong support to voter registration reform as a way of increasing voting; on the other, press coverage suggests–incorrectly–that registration reform will not increase voting because so many registered voters don’t bother to go to the polls.
Many states have been easing the registration process by making it possible for people to register by mail, or at government agencies such as motor vehicle, unemployment and welfare offices. Congress is considering similar reforms.
The Sacramento Bee (4/14/88) likes the fact that 20 states allow people to register to vote in motor vehicle offices because it facilitates “voter registration on the quick.” And the New York Times adds (4/5/88) that people should also be allowed to register to vote when they apply “for unemployment benefits or other state services.” “If the United States really means business about wanting higher voter turnout,” the Pittsburgh Post-Gazette declares (2/28/89), “government should make it as easy as possible for people to vote.”
Some press outlets are also alert to the fact that no other Western democracy has such a “cumbersome registration process. The voter in more than a dozen nations abroad need do nothing more than manage to stay alive until he reaches the legal voting age. The government does the rest” (Dayton, Ohio, Courier-Journal, 10/9/88). Stories have been written about the way other countries make it convenient to register, such as the Boston Globe‘s article (7/17/88) on the Canadian “government’s comprehensive registration drive which includes door-to-door canvassing for voters.”
There has also been increasing attention in the press to the fact that the United States had high turnout in “the 19th century, before registration laws. In the years between 1848 and 1896, about three-quarters of the population voted” (Barre, Vermont,Times-Argus, 6/4/88).
And now the contradiction. The press has also been reporting that fewer and fewer registered voters are bothering to go to the polls, making it appear that registration reform will not increase voting. In an editorial anticipating the 1988 turnout drop, the New York Times (11/6/88) said that “the percentage of registered voters who do vote has steadily declined—from 85.3 percent in 1960 to 80.5 percent in 1968 to 72.6 percent in 1984.” And virtually every newspaper reported the day after the election that voting by registrants had indeed dropped again, to 70.5 percent.
These figures come from state officials who tally up the names on the county registration rolls, and count the election returns. In turn, a Washington-based organization called the Committee for the Study of the American Electorate summarizes all of the state statistics and issues the results to the press.
But reporters do not probe these state statistics for accuracy. Were they to, they would discover what we did: There are 20 million “deadwood” registrants on the local rolls—people who have died, or who have moved, re-registered elsewhere and are counted twice in national totals. Ernie Hawkins, the voter registrar for Sacramento County, California, told us that despite diligent efforts to identify and purge people who had died or moved, there are as many as 18 percent deadwood registrants on his rolls.
In Mississippi, where registration allegedly rose from 68 percent in 1968 to 92 percent in 1984, half of the counties now have registration levels exceeding their voting age populations. Many counties no longer purge the rolls, perhaps for fear of being charged with racial bias. Each election day, it appears that most Mississippians are registered but don’t vote.
Our estimate is that only 60 percent of the eligible voting age population is registered. But the Committee for the Study of the American Electorate reports that more than 70 percent are registered because it uses raw state statistics, uncorrected for dead-wood. Consequently, the press unwittingly treats deadwood registrants as if they were actual registrants and then reports they didn’t vote!
For example, Michigan’s registration level, which reached 90 percent in 1984, resulted from a 1975 law prohibiting county officials from purging people from the rolls until they fail to vote for ten years. If “deadwood registrants” (up to 20 percent of the rolls) were subtracted, Michigan would also show that registrants overwhelmingly go to the polls.
The U.S. Census Bureau’s studies are a far more definitive source of information about registration and voting (after correcting for people’s tendency to claim to have registered and voted when they did not), and the results of these studies tell a different story. About 86 percent of registrants voted in 1988, or almost 15 percentage points more than state statistics show.
The trouble is that the big news play goes to the registration and voting figures released at election time by state officials. By contrast. Census studies do not become available until three or four months after election day, by which time reporters don’t much care. For example, when the Census study of the 1988 election was reported by the New York Times on March 12, 1989, no mention was made of the finding that 86 percent of registrants had voted, although just a few months earlier the public had been told, based on state statistics, that only 70 percent of Americans had voted.
Critics of national registration reform claim that voting should be up because registration procedures are more liberal than ever before. True, poll taxes and literacy tests are gone. But outside of the South, the main reform in registration procedures consists of allowing people to register by mail. This reform has had little impact because provision is rarely made for the wide distribution of the postcard forms—at health, welfare, unemployment and motor vehicle agencies, for example.
Not only has the voter registration system not been liberalized sufficiently, but local structures that once helped people overcome registration barriers—such as industrial unions and urban political organizations—have decayed. Industrial unions no longer bother to register their own members, let alone reach out to non-unionized workers in the vast and growing service sector where registration levels are lowest.
Big-city parties have atrophied, replaced by candidate organizations that do little to register new voters. And national political campaigns, run almost entirely as media events, register few voters.
Without local organizations to help people sign up, or some system of government-sponsored automatic registration, registration hurdles become more telling, gradually driving turnout down. Which is why more and more Americans fail to show up at the polls.
Professors Cloward (Columbia University) and Piven (City University of New York) are the authors of Why Americans Don’t Vote, and board members of HumanSERVE/100% Vote, a national voter registration organization that advocates that people be allowed to register at all government agencies serving the public.