Opt-Out Voting

In the area of election law, I’ve written plenty of articles on the subjects of voting rights and election administration aimed at influencing the courts, legislators, and executive branch officials. Occasionally, though, I’ll take academic license to write a more quirky, hopefully thought-provoking piece. For instance, I have published an article recommending that close elections be resolved by a coin flip rather than litigation.

Another of my quirky articles, titled Opt-Out Voting, will soon be appearing in an issue of the Hofstra Law Review. In Opt-Out Voting, I envision a different manner of casting ballots in elections. The current paradigm of elections is basically for a registered voter to go to the polling place (or get an absentee ballot) and then select the candidate he or she desires. In contrast, opt-out voting would randomly pre-select a candidate for the registered voter and then mail a ballot to the registered voter. The registered voter would then have the option of: (1) doing nothing and staying with his or her currently selected candidate; (2) switching his or her vote to a different candidate; or (3) switching his or her ballot to a vote for “None of the Above.”

The reason for switching to a system of opt-out voting would be to increase the number of persons who cast ballots at elections. Currently, the United States has a low rate of voter turnout. And this low voter turnout rate is especially stark at contests below the federal general presidential election (e.g., in local elections).

The concept of opt-out voting and the rationale behind the concept raise at least two important questions. First, is increasing the amount of voter turnout at elections a worthy goal? Second, would opt-out voting actually lead to an increase in turnout? In my view, the answer to the first question is a definite “yes.” The answer to the second question is theoretically “yes,” but more experimentation would be necessary to figure this out.

In general, most folks instinctively think that high voter turnout makes for a healthier democracy than low voter turnout. Perhaps the biggest reason why high voter turnout improves democracy is that higher turnout increases the representativeness of the electorate. Studies have shown that, particularly on the local level, those who turn out to vote are not necessarily reflective of society as a whole (i.e., lower participation by the poor and less educated). If turnout increases, government should become more reflective of and responsive to all of society and not just certain segments of society.

The question of whether providing a pre-selected candidate to registered voters will spur more registered voters to cast a ballot is a bit iffier. However, there are theoretical reasons to think that such a system might spur increased participation. For starters, the default option for elections has been switched from a baseline of non-participation to participation. In some sense, this is like a system that defaults folks into a 401(k) plan rather than defaults folks out of such a plan.

Of course, the problem with a default of participation is that a registered voter might just pay no attention to his or her ballot. What might motivate a registered voter to do something with his or her ballot? Here, it would seem that the “penalty” of voting for the wrong candidate might spur folks to vote. To take one example, Colorado had sixteen candidates for president on the ballot in November of 2008. With ballots being randomly assigned, that would mean many folks would end up with a fringe candidate from, say, the Heartquake Party. Moreover, if provided with a pre-selected candidate, registered voters might be more likely to seek out information about candidates in order to avoid voting for the “wrong” candidate. In some sense, this proposal takes a cue from Ian Ayres’ theory of penalty defaults for contracts.

Admittedly, opt-out voting amounts to an unusual proposal that is more haute-couture than ready-to-wear. Yet my hope is that it might spur more dialogue about how we can structure our system of elections to nudge more citizens to participate in elections at all levels of government.

And obviously, this is a blog post that omits a lot of the details, so check out the draft of the paper on SSRN if you are in the mood for more.

You may also like...

6 Responses

  1. Logan says:

    Well I’m not sure this is “thought provoking” but it’s definitely provoking, haha. Being that I’m at work and don’t the time to read the draft, I have two initial thoughts that may or may not be answered in it.

    First, what happens if the ballot is lost in the mail either to or from the voter? Most individuals that are actively engaged in politics would probably notice they hadn’t received their ballot by date X and would then contact their local election board but I’m not so sure the current casual voter would, let alone the roughly 40% who currently don’t vote. Similarly, how would we know if our ballot ever made it back in the mail? In either case, I’d want some way to verify ballots were received on both ends before a vote were caste in favor of the randomly selected candidate. One simple fix to this first issue would be to change your first scenario so that individuals must act to have their votes counted but this still doesn’t guarantee an individual’s ballot would make its way back to the election board.

    Similarly, you’d have an issue with recently moved voters who didn’t fill out a forwarding address form and with homeless voters. You’d probably see massive opposition to this from Democrats, especially the more progressive part of the party.

    Without some guarantees for how votes would be counted, I could see protests from groups like the Tea Party who might worry about “government bureaucrat” voting for them and arguments similar to the individual mandate with the ACA.

    Second, I think a better way to increase voter turnout would be to make election day a national holiday so that it’s easier for everyone to vote. Other ideas could involve the States, with various levels of help from the Federal Govt, agreeing to the same election laws so there is less variance among the States and thus easier for citizens to register and know their rights when they move. It also wouldn’t hurt to put more emphasis on the importance of voting in elections during K-12 education.

  2. Ken Arromdee says:

    On the average, even under this opt-out system, the impact of not voting is zero. Failing to vote does not, on the average, increase the chance of an undesirable candidate winning any more than failing to vote does under the current system. So this entire idea relies on people being unable to do math. I don’t want a system which preferentially encourages the mathematically illiterate to vote.

    Furthermore, the argument “it gives people more electoral power since they not only vote for X, they vote against Y” doesn’t work out.

    If you are going to count “ballot says Y, they vote for X” as additional electoral power, you have to count “ballot is already X, they vote for X” as zero electoral power. Needless to say, the extra electoral power exactly balances out the zero electoral power on the average.

    On the other hand, you could count “ballot is already X, they vote for X” as positive power because they get X, who they want. However, if you do that, then the original scenario (ballot says Y, they vote for X) starts out in a position of negative electoral power and the increased electoral power from being able to reject Y is exactly balanced out by the negative electoral power in having Y be the default to begin with.

    Exactly which scenarios are positive, negative, and neutral change depending on how you choose to count, but it never adds up to more electoral power than the current system.

  3. A.J. Sutter says:

    1. Plenty of folks get pink envelopes from the power company or whatever, but don’t pay until the electricity gets cut off or the guy comes to the door. When it’s a non-essential service, such as an election, the envelopes may just wind up in a stack on the vestibule table. Your article assumes that people will prioritize voting at least as highly as they do their bills. Since your concern is voter apathy, this assumption might be a little question-begging.

    2. Another issue is what you mean by randomness. I take it you mean not only that each voter has an equal probability of getting assigned to any particular candidate, but also that the actual distribution is roughly even. (Flipping ‘heads’ 100 times in a row, though improbable, can instantiate randomness in the sense of equal probability.) In any case, I think an even distribution would be necessary to make the scheme socially, if not also Constitutionally, acceptable.

    3. Assuming you intend an even distribution of passive allocations, I think there might also be some issues with your points about (i) whether voters who ignore their ballots determine elections, and (ii) the rarity of close elections:

    Of course, one can surmise that the number of elections in which registered voters who totally ignore their ballots decide the outcome of the election will be quite small. For starters, in a large electorate, the voters who totally ignore their ballots are likely to be distributed about evenly. Second, even if ballots that are totally ignored are slightly non-randomly distributed, most elections just aren’t that closely decided. [Citation omitted] Thus, even if there is a non-random distribution between the candidates of voters who totally ignore their ballot, the non-random distribution is likely not to determine the outcome. In other words, it’s not likely the difference in the random distribution among ballots between the first- and second-place candidates will exceed the margin of victory for the candidate who won among those registered voters who actively participated.

    The assertion that “most elections just aren’t that closely decided” refers to conventional elections where most voters trudge down to a booth, and others actively request absentee ballots. I.e., where all voters are active, and where the fraction of participating voters is < 100% and often < 50% . It doesn't relate to opt-in elections that allow passive voting.

    In an opt-in system 100% of voters will vote, either actively or passively. There isn't any way to distinguish between people who ignore their ballot and those who "actively" agree with the pre-selected choice. Maybe some people who ignore their ballots actually would have preferred some other candidate, but were too busy with work, etc. to vote. Philosophically, it's debatable whether you can say that they didn't do anything to "decide the outcome" — but for their inaction, the candidate might have had a reversal of fortune. Moreover, if you exclude them from a role in deciding the outcome, you'd have to say the same about the candidate's active supporters who received ballots consonant with their views.

    In that case you'd be saying that only those active voters who change their ballots decide an election. Once that feature of the system is understood, I’m not sure that it will warm peoples’ hearts to voting more than the current system does. Also, it will mean that “decisive” voters will always be voting in opposition to what the government has chosen for them. This may have spin-off consequences for the general attitude towards government and the social fabric overall.

    In addition, your article doesn’t seem to contemplate more than 2 candidates per election. But if ballots are randomly allocated in the sense of (2) above, then candidates from the Socialist Workers’ Party, Libertarian Party and Greens will each get roughly as many votes as the Republican and the Democrat, a priori. If people who ignore their ballots are randomly distributed, as you hypothesize, this should mean many more elections won by pluralities rather than majorities — i.e., many more close elections.

  4. A.J. Sutter says:

    Sorry, both your post and the article do mention minority parties; not sure why I was momentarily hit by tunnel-vision there. However, your idea that being allocated to a minority will “nudge” voters into action still presupposes that large numbers of voters won’t ignore the ballot altogtether or simply acquiesce to the choice: the question-begging I mentioned above. My points about the increased occurrence of close elections and the social consequences of voting to oppose government still remain.

  5. TJ says:

    Even putting aside whether the proposal will work (Ken’s point), the other big question is, why use this convoluted method? If you want to increase voter turnout, just give every voter a $20 bill along with the ballot.

  6. PrometheeFeu says:

    First, I disagree with the idea that getting more people to vote through that system would be a positive thing for democracy. A higher voter turnout is a “good thing” because it means more people are engaged in the political process and thinking about the issues. If you just shove people to the ballot box, you will simply increase the number of “bad” voters who do not pay much attention to the issue. After all, if they could not be bothered to vote before, why would they bother paying attention now?

    Also, this could have disastrous consequences on the legitimacy of elections. There are many elections every year. As such, it is almost inevitable that in some races, the random assignment will end up favoring one candidate over an other possibly changing the outcome. Even if that is not the case, bitter losers will surely bring up the possibility of that being the case a seriously undermine the legitimacy of the election. This is a Bad Idea.