Right now we are the in the midst of the midterms, the news cycle is flying at us like hypersonic shrapnel, and we are bleeding. On the one hand, one does not want to give in to false flag conspiracizing. On the other hand, many of us believe that the 2016 election was significantly affected by manufactured news cycles. The slope between implausible conspiracy (“pipe-bomber dude was planted years ago by the deep state!”) and plausible manipulation of the news cycle (the “caravan”) becomes slippery. It’s not surprising that we all end up in different places, depending on our priors. And this is “just” a midterm. After next Tuesday, the 2020 Presidential cycle begins in earnest.

It feels like election season never ends. But I want to submit that the problem is that it does end, with a very high-stakes election. The integrity of our elections has, thank goodness, become a subject of active public concern. The ways we draw district lines, the ways we form voting rolls, the susceptibility of our voting machines to malfunction or subversion, all demand scrutiny and reform. But there is another vulnerability, hiding in plain sight: the fact that our elections all take place basically on a single well-known day.[1] A short, sharp manipulation of the news cycle, if well-timed, can tilt electoral outcomes. Many people plausibly blame Hillary Clinton’s loss on a letter by James Comey released a week prior to the 2016 election. Perhaps that was not intentional manipulation, but similar events past or future might be.

It seems obvious to me that political actors of every party and creed do their best to exploit this open vulnerability in our electoral system, working to manufacture coverage or even newsworthy events likely to motivate their own electoral base in the immediate run-up to an election. To the degree electoral results turn on this contest, they seem likely to reflect the cleverness (or deviousness) of campaign operatives much more than any colorable expression of the “will of the people”. It’s hard to come up with a justification, even under very naturalistic theories, why this would be a desirable form of democratic deliberation.

Less apocalyptically, the quality of representation is undermined by a predictable election cycle. In the run-up to an election, often inaccessible public figures become suddenly more available and accessible. It is a commonplace that Senators’ willingness to take unpopular votes depends upon whether they are near the beginning or the end of their six-year terms. That might be intended in Constitutional theory, but in practice, in my view, unpopular votes are less likely to reflect dispassionate deliberation on behalf of the polity and more likely to entail privileging the interests of the elites that dominate either political party. Even in theory, however “insulated” we decide Senators should be from vicissitudes of the mob, this cyclicality in time of popular responsiveness seems undesirable. Ideally, our representative would not be “there for us” only during the short season when they are soliciting our votes. They would be there for us all the time.

For a variety of reasons, I think it a good idea that we introduce into our voting system a greater element of stochasticism, of structured, intentional randomness. This may be counterintuitive — sure, a manipulated news cycle may not express the will of the people, but how could a random number? The deep fact of randomness is that while an individual “draw” may be noise, random selection has characteristics that are well defined, widely understood, and intuitively accessible. When statisticians want to examine a population, they take a random sample and characterize that. With good randomness and a reasonably large sample size, it becomes extremely unlikely that the characteristics of the sample will fail to represent the broader population. This fact is already a part of our political process. Pollsters, who affect electoral possibilities as well as characterizing them, seek (very imperfectly) random samples of likely voters. We select juries largely by lottery, on the theory that this is a good way to get a representative sample of ones “peers”. There have been a variety of democratic experiments with sortition, simply choosing by lot, picking random names from the phone book. Perhaps overcynically, many of us might consider that an improvement over our present, professional political representation.

I do not favor sortition for the constitution of our legislatures. There is a lot to be said for choosing among representatives who express an interest in and commit to doing the work, and to some kind of voting process that ideally filters for quality. What I do favor is an idea called “lottery voting” or “random ballot“. I really encourage you to read the first link, a very readable academic “note” by Akhil Reed Amar which introduced the idea. You should also read this essay by David MacIver (ht Bill Mill). In a nutshell, everybody votes in the way they currently do for their preferred candidate. Then we throw the ballots in a big hat, and draw the winner from it like a bingo hall door prize.[2] You’d never want to use lottery voting to elect a President. Who knows who you might pick? It’d be totally random. But for a large legislature, lottery voting will predictably yield proportional representation along whatever axes or characteristics are salient to voters, not just formal political parties. Further, lottery voting is immune to gerrymandering, and every vote always has equal influence. (This is decidedly untrue of conventional “first-past-the-vote” voting, where the statistical effect of a vote — the difference in the probability of a candidate winning with and without an additional vote — depends very much on the closeness of the election.) There are lots of reasons to love lottery voting, including the conventional case for proportional representation, which I very desperately endorse. (See Matt Yglesias and Lee Drutman.) Not to let the best be the enemy of the good, I’d favor American experiments in more common forms of proportional representation (multimember districts, party lists), but lottery voting really is the gold standard. It is simple, effective, resistant to entrenchment of incumbents or capture by political parties. The US House of Representative should be selected by lottery voting today. At the very least, we should start experimenting in some state houses.

One interesting characteristic of lottery voting is there is no need that elections be simultaneous, or even take place at known predictable times. Suppose we had an electoral system that looked like this: Every month, 5% of the voting roll is randomly selected to cast a ballot for a representative. There’s no big election day: Any time during their month selected voters can come in and cast their vote. After the balloting period has passed, one ballot is randomly selected, and then a virtual coin is flipped that comes up heads only one time in 24. If the coin comes up heads, the current representative is replaced with the randomly selected ballot. If not, that month’s ballots are thrown away, and the representative’s term continues. Under this system, on average, a representative’s terms would be 24 months, but there would never be a period when a representative is more or less near an election. Whatever persuasion incumbents (or their political parties or PACs or dirty tricksters) want to engage in to see to their reelection, they’d have to do basically all the time. Challengers also could arise at any time, but would want to make their case continually. That would become a very different enterprise than existing elections, which engender an avalanche of marketing in sprints. People who wish to become representatives would want to become prominent and popular within their communities, or become endorsed by popular civic organizations (including but not just political parties), in ways that are sustainable over time.[3] Is this a good idea? One might argue that it would just make elections more expensive to contest, and so increase the influence of money. But lottery voting by its nature is much less susceptible to vote buying. Your ads can win 60% of the vote and you still have a 40% chance of losing.

There are definitely trade-offs to this idea of continuous elections. I’m much less confident of it than I am of the virtue of lottery voting in general. But stochastic, continuous elections would remedy very real flaws in our electoral system: its nonrepresentativeness of our views over time, its provocative susceptibility to manipulation of our views over short periods of time. You might fear that, rather than eliminating the hypermanipulative, sometimes fictional news cycles that divide us during election season, continuous elections would incentivize partisan provocation all the time, which would be bad. But lottery voting by its nature reduces the incentive of political parties to polarize us. If the major political parties disgust us, under lottery voting one can choose some other party to vote for without throwing away ones vote. The strategy of demonizing “the” other party just doesn’t work in multiparty systems. No one has to vote for least-worst. As Matt Yglesias put it, “It would be better to have a country where everyone is voting for a party they are genuinely enthusiastic about, and then because no such party commands majority support, the leaders need to do some bargaining.”

Also, under lottery voting in any form, turn-taking is likely. Even a very popular representative might have to sit out a term, because of some random throw of the dice. In order to protect local interests, districts would want to sustain institutional knowledge and legislative effectiveness by keeping some professional staff permanently, rather than replacing Congressional staff completely with copartisans each time a seat changes hands. Members of even a very dominant political group within a district would experience interregnums during which their interests would be entrusted to the hands of someone whose views might differ quite radically from their own. Even as proportional representation would liberate citizens to vote their very diverse genuine preferences, these institutional facts would encourage compromise and common ground. Certainty of power breeds the confidence to be cruel. Susceptibility to chance reminds us that, whatever our differences, we must all depend upon one another.


[1] Early and mail-in voting complicates this story a little bit, stretching the single voting day perhaps to a few weeks, more a short season than a day. But a short season is not so different than a day, and even in states where mail-in balloting is prominent, traditional election day ballots remain crucial to outcomes.

[2] In practice, we might not use a big hat. There are very robust, almost-impossible-to-corrupt ways of generating public random numbers, involving cryptographic commitments by multiple parties. As long as any of the parties is not corrupt, that is, as long as all the parties do not collude, we can generate a fair random number. We’d use that sort of procedure, and choose a ballot like a winning lottery ticket.

[3] You would want this to be an election among known candidates, not a procedure by which everyone nominates themselves or a friend as write-ins. To prevent that kind of thing, Amar suggests “It might…be necessary to put into the twirling basket only ballots cast for candidates receiving more than, say, one percent of the total vote.”

Update History:

  • 30-Oct-2018, 12:30 p.m. PDT: “under which why this would be a desirable form of democratic deliberation.”
  • 30-Oct-2018, 12:35 p.m. PDT: “…or become endorsed by popular civic organizations (including but not just political parties) within their communities…”