18 August 2016
Did adopting a mail ballot increase presidential turnout in Oregon?
The true believers in mail ballots (absentee ballots in Montana) think so. And there’s considerable evidence that in traditionally low turnout elections — school trustee and municipal elections are good examples of the genre — voting by mail results in a higher turnout of registered voters than does voting in person at the poll on election day.
But did voter turnout in the presidential general elections in Oregon, the first state to make voting by mail mandatory, increase after Oregonians began marking their ballots at their kitchen tables instead of in the voting booth?
No. Not as measured by the voting eligible population statistic. That fact has implications for current efforts in Montana to cajole voters into voting by absentee ballot.
Oregon’s mail ballot system was adopted on 3 November 1998, when voters approved a citizens initiative by a two to one margin. Oregon’s first all mail general election was in 2000.
Figure 1 displays Oregon’s turnout rate in presidential elections for 1980–2012 using two statistics: the percentage of registered voters that cast ballots, and the percentage of the VEP that cast ballots.
For both statistics, the post mail ballot mean is higher than the vote in person at the polls mean. Rudimentary statistical tests suggest that for the registered voter statistic — the statistic that vote by mail advocates prefer — the difference is statistically significant. But for the VEP turnout statistic, the difference was not significant.
I applied two statistical tests, Student’s t (Figure 2), which Gosset developed for small samples when he worked for the Guinness brewery in Dublin, and the Wilcoxon-Mann-Whitney rank sum (Figure 3). The results provide a starting point for a discussion, but should not be embraced with the unquestioning fervor of a religious conversion. The switch to a mail ballot was not the only variable in these elections.
Figure 2
Figure 3
Although political campaigns and parties think of turnout exclusively as the percentage of registered voters casting ballots, RV turnout is not always a good statistic for assessing the health of an elections system. A quick thought experiment reveals why:
Suppose, for the moment, there is a 51st state, West Grapeland, renowned for its proletarian vineyards and world class muscatel. West Grapeland’s voting eligible population numbers one million, but because of onerous voter registration laws, and muscatel induced indifference to politics and civic responsibility, West Grapeland has only 100,000 registered voters. In the presidential election, 90,000 registered voters cast ballots. The registered voter turnout is 90 percent, which certainly sounds impressive, but the VEP turnout is only nine percent, which is enough to send a responsible citizen lunging for a Styrofoam cup brimming with West Grapeland’s best selling beverage.
West Grapeland’s predicament illustrates why a high registered voter turnout is not proof of a healthy political system.
There’s a wee bit of the West Grapeland problem in Oregon. Although RV turnout increased after the switch to mail ballots, the percentage of Oregon’s voting eligible population that has registered to vote has declined over the last 32 years. As displayed in Figure 4, the decline is slight, but it is perceptible and it’s not a welcome development.
Oregon’s experience and the West Grapeland problem should, but won’t, cause Montana’s Democrats and progressives (often, but not always, the same) to temper their enthusiasm for absentee voting.
But it won’t. As I reported last week, the Montana Conservation Voters Action Fund is spending half a million dollars to conduct a door knocking canvass in Missoula, Bozeman, and Livingstone, to persuade voters to sign-up for absentee ballots. Almost all Democratic candidates make that pitch as part of their door knocking routine. They believe that because approximately 80 percent of Montana’s mail/absentee ballots are returned, absentee ballots increase turnout.
That cannot be concluded from the return rate for mail ballots in a hybrid mail ballot-vote in person system such as Montana’s. Because absentee voting in Montana is optional, voters must be motivated to ask for a ballot, and to keep renewing their request. The return rate is high not because the ballot is being cast by mail, but because Montana’s absentee voter pool comprises people with a high motivation to vote by mail. These are Montana’s regular voters. If someday mail ballots become mandatory for all Montanans, and are sent to all registered voters, the return rate will decrease if there is an absence of the West Grapeland problem.
I always vote in person on election day. And, I always vote. But unlike many absentee voters, I’m motivated to consider the arguments and facts right up until election day, and believe that doing so is my duty as an American citizen.