Among the kids, participatory direct democracy is often taken as the ideal to which democratic polities ought aspire. But at least in theory, the case for representative democracy is strong. Political decisions really matter. They should be made well. But they are hard. Whatever interests and values you hold dear, it takes a lot of work to inform and educate yourself enough to know what political choices would in fact best serve them. This work must be performed in the face of tsunamis of misinformation propounded by those serving interests and values that diverge from yours, but whose partisans are eager to co-opt you. Democracy-skeptical public choice theorists aren’t wrong when they say that most voters are (and ought to be) “rationally ignorant“.

The genius, in theory, of representative democracy is that voters hire specialists to do the information work for them. In a representative democracy, it is not really your job as an ordinary citizen to have a strong view about the details that actually get legislated, like how many tax brackets there should be and how the rates should be structured across them. Your job is one that you (and only you) are eminently well placed to perform — to know your own interests and values, and elect a person who reflects them. That person become a specialist who gets paid to do the work of translating those interests and values into political choices that give them real effect.

This is the standard case for representative democracy you probably learned in civics class. Yet at this point, most of us roll our eyes more than a little at it. I don’t feel remotely represented by my alleged representatives, to whom I have no personal connection and little affinity of values or interest. In the US House of Representatives, I am one of about 750,000 people that my “representative” allegedly represents. In my city, I elect a “supervisor” I have never met who allegedly represents the values and interests of 80,000 of my neighbors. It’s a bit ridiculous. More than a bit. A person who “represents” a population of tens of thousands of people whose only commonality is geography effectively represents no one at all. Elections impose constraints on politician behavior, sure. If you want to keep the gig, you can’t do whatever will offend some least-common-denominator id among your constituents. You must do what it takes to raise funds for the competitive advertising campaigns that “elections” become. The first constraint is “democratic”, but it constitutes so watered down a form of representation that it counts for very little. The second constraint is often antidemocratic, since global-dollar-weighted and local-population-weighted interests and values are in conflict quite often.

Within the guard rails set by these two constraints, what “representatives” do cannot straightforwardly be described as representation. Almost any choice a politician makes would give effect to the values and interests of some of their constituents but not others. Absent the universal information work representative democracy exists to absolve us of, most of us cannot even evaluate whether the choices our representatives make are likely to further or frustrate our interests. The social architecture of contemporary representative democracy is like a how-to manual for the so-called “iron law of oligarchy“. Our institutions immerse our “representatives” among a class of electeds, bureaucratic staff, and professional courtesans. The burdens of the job keep them segregated from the publics they purport to represent. In an organic, social sense, they become bound and accountable much more to their comrades-in-arms within the governing class than they could ever be to the amorphous, conflicted group they call constituents. An economist might describe all this as an “agency problem” but that’s not quite right. There’s not a coherent enough principal whose interests the agent can be said to betray.

Modern representative democracy is simply a system whose predictable result is governance by competing coalitions of insiders, who develop deep relationships and thick connections to one another, while the electorate they notionally serve becomes an inchoate, threatening demon that must be flattered and appeased. The values and interests insiders actually serve may be corrupt and self-serving, or they may be idealistic and selfless, but they cannot accurately be described as “representing” their constituents as a body. Constituents feel unrepresented, because they are. Popular pressure builds for flawed institutions of direct democracy — ballot initiatives, referenda — under which the information problems representative democracy exists to solve run riot, but at least it isn’t always the same fuckers calling all of the shots.

Direct democracy enfranchises the citizenry to decide upon matters of whose details and ramifications they are rationally ignorant, with predictably imperfect results. Contemporary representative democracy creates a corruptible class of specialist-coworkers, who develop their own values and interests and substitute them for the those of their constituents (whose actual values and interests are too diverse and conflicted to make their own strong claim). Is there anyway we could get something that combines the enfranchisement of direct democracy with the informed participation enabled by representation? Yes.

Imagine what an online direct democracy might look like. All of us would be the legislature. Obviously, we wouldn’t meet (or mostly pretend to meet) under a neo-Roman dome in some self-important provincial city. With a legislature of only a few hundred souls, attention must be very carefully allocated. In our current Congress, there’s a whole economy of scarce floor time. (“I yield the remainder of my time to my colleague”, you’ll hear them say.) If all of our legislators were permitted to speak as much as they would, deliberation would take too much time. In fact, most legislators never weigh in at all on most issues that in a broad sense come before the Congress. Congress organizes itself into committees (by arcane means, with corrupt effect), and most matters never make it through and past committees to consideration by the broader chambers. If we had a legislature of 250 million (roughly the voting-eligible population), obviously the vast majority of citizen-legislators’ proposals and bright ideas could not be put before all their citizen colleagues. If only 1% of citizen legislators were to make a proposal each year, we’d all have millions of proposals to evaluate. That’s untenable. So we’d have to design a kind of stochastic parliament, where people’s proposals would initially go to very tiny fractions of “the legislature”. These random samples would constitute ad hoc “committees”, and each citizen would be responsible for serious deliberation on the proposals that come before them in this way, but each participant would field only a modest number of such proposals. Following deliberation and potentially modification at this stage, these ad hoc committees would vote to promote or kill the proposal. If they promote, the same procedure would recur but with a larger sample, and less scope for deliberation and modification. The number of such proposals that could be promoted to higher levels of review would be limited and so competitively rationed: only those gathering the most support would gain scarce “slots” compelling the broad polity to review them. Finally, the tournament-winning, most promoted proposals would get plenary up or down votes, like a vote on the House floor.

You can imagine this kind of thing, but it would do little to address the problems we invented representative democracy to solve. To function well, our citizenry would have to be extraordinarily engaged and informed, and it would take up all of their time. It would be like permanent jury duty.

But what if we elected representatives to participate in this kind of mass-democracy framework? Instead of electing one per 800,000 or one per 80,000, what if we self-affiliated into groups of common interest of no more than, say, 1000 souls, for whom personal, physical “town meetings” could be regularly arranged? Obviously, not everyone would wish to attend all of these meetings, but everyone could if they wished. With no more than 1000 constituents, an elected could become at least acquainted with her full constituency. She could be accessible and available to them all. She could maintain direct relationships with a substantial fraction of the people she represents, and be motivated and held to account by those relationships, by gratitude and shame experienced personally rather than by abstract shifts in what some consultant claims the polls say.

Instead of a few hundred Congresspeople, we’d have 250,000 representatives whose full-time job it would be to stay and live among and interact with their constituents, and participate in the online legislature. There would be no Congressional offices in Washington, no risk of going native among colleagues who become much closer than constituents. At a municipal level, there would be no councilmen or supervisors at City Hall. In my San Francisco, there would be roughly 800 legislators and any of us who cared to would know our representative and interact with her as much or as little as we pleased.

This proposal recognizes that the hard part of being a representative, or at least what ought to be the hard part, is not fundraising, rising through committees, learning the personalities and peccadillos of influential colleagues so that you can “legislate effectively”. The hard part of being a representative is representing. The problem we should devote ourselves to is the challenge of making one person’s voice become a capable stand-in for many others’ necessarily absent. The legitimacy of our entire system of government depends upon this thin reed, the quality of the bond between elected and constituency. When that bond becomes as attenuated and deflected as it has under current institutions, “democracy” fails to confer very much legitimacy at all, or to be effective at serving the interests of the people on whose behalf it claims to rule.

This proposal recognizes also that human beings are best motivated and held accountable by direct relationships to other human beings. Pecuniary incentives are course-grained, and always susceptible to corruption. Career incentives — they’ll serve the people because they want to be reelected! — are wildly insufficient. The range of things a politician may do and still get reelected is wide, and can deviate a great deal from their constituents’ interests. Career incentives are easy for outsiders to game. If you serve constituencies that may not entirely be those you supposedly represent, there may be a gig on K Street or a place in the party bureaucracy for you. The foundational error of the neoliberal period was the conceit that aligning financial incentives to social goods was easy, so market success and contribution to social welfare could, to a first approximation, be equated. Market success, alas, in the market for political careers as much as cigarette sales, can be welfare destructive rather than socially valuable. The best work results from intrinsic devotion to excellence plus human relationships that help steer a person’s excellence towards ends that a community values. People need to be well paid not so financial incentives direct their work, but so that financial anxieties and ambitions don’t misdirect, eclipse, distort, occlude the fragile foundations of real human achievement. Our representatives should be paid well, and should serve and live among tangible human communities whose interests they know and experience through organic personal relationships.

So “expand the House” from 435 to, um, 250,000, and put it online. Obviously, this is an idea that can’t be put into immediate practice at a national level. We have a lot to learn before we’ll trust large-scale stochastic deliberative assemblies to resolve political questions with extraordinarily high stakes. However, it is a vision that we should be working towards. Whatever you think of “crypto”, one thing that proposals like this highlight is the need for extremely trustworthy networked computation infrastructures that are credibly neutral, that are not subject to the discretion of some party that owns or operates the machines. If you want to run a legislature over a network, there can’t be a company that manages the database that might potentially manipulate it. You need the system to produce very persuasive, public evidence of its integrity at all times. I don’t think public blockchains in anything like their contemporary forms will get us there, but they are working prototypes of this sort of trustworthy computation. They are also sites of experimentation in (rudimentary, badly flawed) online deliberative assemblies such as “DAOs“. There is plenty to hate about contemporary crypto, but in the midst of all the scam and speculation there are emerging fascinating “petrie dishes” for experimental democracy, to which it is worth paying some attention, and cheering useful innovation. Most of cryptoland is understandably but unfortunately cynical of representative models of democracy. But the usual alternative — “governance tokens” directly voted, like shares of stock in a traditional corporation — performs poorly. Token-voting is plutocratic by design, and outcomes tend to be dominated by insiders and activists while much larger “rationally ignorant” groups just “HODL” (hold) their tokens for speculative purposes without voting them. The interests of stakeholders who are not tokenholders get ignored entirey. (See Vitalik Buterin’s lengthy critique of “coin voting”.)

Legislation is, in computer lingo, a very stateful application. Online deliberative assemblies will need to keep precise track of large numbers of lengthy documents and particular revisions thereof, which cannot be done on contemporary blockchains at a reasonable cost and speed. But it would not be so difficult to repurpose some of the technologies that underlie contemporary crypto to build bespoke, city-scale legislatures that could be operated affordably and generate compelling evidence of their integrity. Cities should give “mass representative democracy” a try, soon. If you live in a city of any size, do you feel, today, like you are adequately represented in city government? If not, what hope do we have to make representative democracy work at a state or national scale? We are collectively, and correctly, coming to understand that we’ve never really had the kind of Our Democracy that talking heads on MSNBC are constantly telling us we must save. It’s time to roll up our sleeves and build institutions we’ll have reason to be less cynical about.

Friend-of-the-blog David Shor gets into trouble for two distinct reasons. One is “popularism”, which provokes arguments about how poll-driven and message-disciplined Democratic electoral politics ought to be. The other is simply prediction. Shor is a Cassandra. Here’s Ezra Klein characterizing his views:

Democrats are sleepwalking into catastrophe. Since 2019, [Shor has] been building something he calls “the power simulator.” It’s a model that predicts every House and Senate and presidential race between now and 2032 to try to map out the likeliest future for American politics. He’s been obsessively running and refining these simulations over the past two years. And they keep telling him the same thing.

We’re screwed in the Senate, he said. Only he didn’t say “screwed.”

In 2022, if Senate Democrats buck history and beat Republicans by four percentage points in the midterms, which would be a startling performance, they have about a 50-50 chance of holding the majority. If they win only 51 percent of the vote, they’ll likely lose a seat — and the Senate.

But it’s 2024 when Shor’s projected Senate Götterdämmerung really strikes. To see how bad the map is for Democrats, think back to 2018, when anti-Trump fury drove record turnout and handed the House gavel back to Nancy Pelosi. Senate Democrats saw the same huge surge of voters. Nationally, they won about 18 million more votes than Senate Republicans — and they still lost two seats. If 2024 is simply a normal year, in which Democrats win 51 percent of the two-party vote, Shor’s model projects a seven-seat loss, compared with where they are now.

Sit with that. Senate Democrats could win 51 percent of the two-party vote in the next two elections and end up with only 43 seats in the Senate.

I’m not a fan of “popularism”. I agree with Shor’s view that Democratic Party activists, particularly on social issues, constitute a weird, vanguardist community that often fails by placing its own concerns and unpopular remedies before serving the actual preferences of the demos. But public opinion polling is a bad tool, both because it measures whatever it purports to measure poorly, and because serving the demos requires a richer understanding of the public’s predicament than answers tossed off in response to decontextualized multiple choice questions. I think in practice polling is as likely to mislead as to help. Our political parties require sociological change. They cannot remain platoons of ideologues supported by plutocratic philanthropy, joined at the hip to canny lobbyists and dealmakers, and serve the public well. There is no technocratic quick fix to that. The parties have to change. Democrats are no worse than Republicans in this regard, but that won’t save them.

Though I often take issue with Shor’s prescriptions, there is no one in US politics I trust more on description. If David Shor thinks Democrats are “screwed… [o]nly he didn’t say screwed” in the Senate, I believe him. And I’m not alone. It’s pretty much a commonplace, when I talk to people involved in Democratic politics, that from 2023 on, for the forseeable future, Democrats will have little means of exercising political power at the national level. Of course Democrats should be careful not to let that be a self-fulfilling prophecy. Human affairs are unpredictable and pessimism can be hubris as much as optimism. But the possibility that the United States will be governed by Republicans or else entirely gridlocked for the next decade seems like one we should be thinking about and taking seriously.

Apocalypticism doesn’t constitute taking it seriously. A loud, small group of politically active Democrats may think that Trump is basically Hitler and the contemporary Republican Party is basically Trump, ergo Republican political power is the holocaust, what’s the point talking about rearranging deck chairs in a gas chamber? But if we are trying to describe the actual world, holocaust is a tail risk, not the modal scenario. Under Republican control as much as Democratic, the range of possible outcomes is large. A politicosocial formation that prides itself on being adult, mature, serious, and devoted to reason owes the world more care than “après nous le déluge”. It may or may not be true that government at the Federal level will be dominated by the Republican Party for the foreseeable future. We oughtn’t concede that, but can we plan for the contingency? If it happens, what would make the world a better place?

The Republican Party, like the Democratic Party, is a big tent, an awkward coalition. Not everyone is Marjorie Taylor Greene. Mitt Romney is a fucking plutocrat, but he also proposed a better child allowance than any Democrat did. The fortunes of MSNBC and Fox News may depend upon salacious culture wars, but the actual welfare of most human beings depends much more on the material choices our government will make. Eclipsed by all the circus there is a great deal of heterogeneity within both coalitions on those questions.

I don’t have the answer, but a clear point of leverage in our system is the primary process. There should be no jurisdiction in the country where there is not a basically decent person with good views on material questions on the Republican primary ballot. To be credible at all, to not be a “RINO”, that person will be disagreeable on a variety of issues. In most jurisdictions they will be some shade of pro-life. They will be LGBT quietist at best, advocates of “content of their character” race-blindness at best. But a person can be unusually supportive of labor without being a RINO. They can be a devoted antimonopolist without being a RINO. They can take climate change seriously. They can, like Romney, agree that families require and deserve material support since every child is and ought to be a mouth without a job.

We make idiots of ourselves if the finest distinction we can draw is between Democrat and Republican. If Shor’s analysis is right, we’ll need a lot of the better Republicans. We should be thinking about how to encourage them to join primary contests, and how to help them win.

Then, if you are a vote-blue-no-matter-who Democrat in a red or purple jurisdiction, you should register as a Republican so you can support the better candidate in the primary. Democrats will produce a candidate. (Please, Democrats. Produce a candidate.) There will be some-who-blue you’ll still vote for in the general. Regardless of what party ID you’ve registered under, you can enthusiastically support the candidate you prefer. Ticket splitting and cross-party voting are glorious traditions in American politics that are worth reviving.

But if you are a person of conscience, and if it is in fact probable that the US will be structurally tilted towards Republican rule for the foreseeable future, you are not proving your virtue by absenting yourself from the forums that will shape what Republican rule actually means. 2025 will come, not the apocalypse. A world not in fact ended will require strategic, constructive engagement. If we are serious and not merely partisan, we should be building effective ways to provide it under plausible foreseeable futures.