Kevin Drum reports on the most practical piece of political-science research I’ve ever seen: a demonstration that sending out a mailing telling people which residents in their neighborhood (including themselves) voted, and who didn’t, in recent elections, and promising to do so again after the next one, increases turnout at a fraction of the cost of other sorts of campaign activity.
This seems to have the potential to transform the political landscape.
My thoughts, in no particular order:
1. The detailed estimates (less than $2 per additional vote) seem to depend on the circumstances and on careful screening to weed out those very likely to vote anyway and those very unlikely to vote, and the choice of an election with low baseline turnout. Not clear how much this would work for a Presidential campaign. In particular, it creates no social pressure for being registered. In a Presidential race turnout among those registered is pretty high; most of the eligible voters who don’t vote aren’t registered. (Not, obviously, a problem in same-day-registration states.)
2. Not clear whether the cost estimate includes the cost of of collecting and processing the voter data.
3. Even so, it’s obviously a useful technique to some extent.
4. It seems creepy, like an invasion of privacy. But while who I voted for ought to be secret in order to protect me from pressure and everyone else from bribery, whether I voted isn’t private, and shouldn’t be. Election officials need that information to prevent me from voting twice, but that’s not the real point. If voting is a civic duty — if “voter” is a public office, and indeed the supreme office that chooses all the other officers — then whether a given voter is diligent or not in the exercise of that office is something his fellow-citizens are entitled to know, and the resulting pressure to vote (or at least to show up and cast a blank ballot) is legitimate.
5. The logic is simple: the participation rate is a public good, and (per Mancur Olsen) the only way to get rationally selfish people to provide public goods is to tie them to private goods, such as the good reputation you get as a diligent citizen.
6. Still, if I were a voter, I’d be somewhat creeped out by the mailing, and not at first blush grateful to the candidate who sent it out.
7. Boosting electoral participation is a legitimate public-interest activity that can be paid for by 501 (c) (3) non-profits, though (I think) they couldn’t legitimately select among voters in their target areas based on partisanship or indicated voting intention, or explicitly select areas to focus on for partisan reasons. Boosting turnout over all is generally good for Democrats, and since low-turnout areas tend to be Democratic-voting areas, concentrating on those areas — a legitimate choice in terms of cost-effectiveness in boosting participation &8212 would in practice create a partisan advantage.
8. Since increasing participation is a legitimate public purpose, a Secretary of State’s office or local elections board could do the same thing.
9. Nothing that I know of would prevent, e.g., the Urban League or a local Ministerial Alliance from organizing such an effort aimed at African-American neighborhoods. And of course a membership organization such as a church or a union could distribute the voting records of its members to its members, and/or post them at places where its members meet. There’s a logic to that: an organization’s political power depends in part on the participation rate of its members, so it’s in the interest of all of the members for each of the members to vote.
10. So if I were running a campaign, I’d want the job done, but I wouldn’t much want to have the campaign do it. I’d much rather rely on the Secretary of State’s office, or the League of Women Voters, or a neighborhood group. The sacrifice in targeting would be worth it to avoid the onus of Big-Brotherism.
11. Let’s get those organizations working now in Democratic-leaning areas. And let’s try to extend the pressure to registration by sending out mailings showing who is, and who isn’t, registered, complete with instructions on how to register.
12. Again, this could be done with public money in a state where the governor and the secretary of state are Democrats and the party has control of the state legislature. Of course, there are lots of Democratic officeholders who wouldn’t welcome increased participation, especially in primaries.
13. The capacity of interest groups to dominate low-turnout elections (teachers’ unions and religious fanatics in school-board elections) is a big problem. This could help change that picture.
Finally, on a different note: How did this research project get past human-subjects review? I think this sort of work ought to be allowed, but by IRB standards it looks very iffy: some subjects are damaged by being embarrassed, and none of them has given consent.
I’ve been wrong before, but this looks to me like A Very Big Deal. Liberals have been caught in a positive-feedback trap: because the poor and near-poor, and especially the unemployed) vote at much lower levels than the rest of the population, it doesn’t make electoral sense to cater to their interests. (That’s why Democrats love to talk about “the middle class.”) But the fact that Democrats do relatively little to help the poor (especially the white poor) in turn helps suppress their political participation. That effect is exacerbated because poor people don’t make big campaign contributions. Democrats have done far less than they should to protect poor people from the misdeeds of the financial-services industry because financial services is part of the donor base.
If Democrats can learn how to pay for campaigns out of small donations, and boost turnout using techniques such as this one, the partisan balance of power and the effective center of the political spectrum could shift substantially. Think of this as a non-legislative approach to campaign reform.
Time to get to work. If any reader can point me to some organization that plans to do this in a fashion that would help the Democrats, I’m willing to write a check.