...Archive for April 2021

Writing as a public good

The newsletter platform Substack has grown controversial. For an overview of the controversy, see Ben Smith. To get in the trenches, read Nathan Tankus’ impassioned letter about why he’s leaving the platform.

I am mostly aloof to the Substack wars. I love trans people. I also love, read, and learn from people that some trans advocates accuse of being hateful. The humans are complicated. Love them. I have little sympathy for the big names who’ve made careers of being “canceled”. But I worry that the growth of media that blur personal and political spheres is reshaping offline norms in ways more likely to impoverish our private lives than enact useful change. These media include Twitter feeds and old-school blogs. But Substack newsletters, because of the incentives and ultimately influence that come with monetization, raise the stakes. This is most clear in the authors whom the platform woos with advances under “Substack Pro“. Substack may ostentatiously recruit trans activists to “balance” complaints that they’ve become a refuge for bigots, but arguably their and their authors’ pecuniary interest is in a lively culture war, hot on both sides, rather than in forms of deliberation that might be more constructive but less exhilarating.

As a person who likes to write but who thinks professionalization corrupts public affairs writing, these are issues I give some thought to. Jeet Heer makes a good point when he tweets “Writing is either a career or it’s an aristocratic hobby. If it’s an aristocratic hobby, it’s closed to most people.” At the same time, when writing becomes a career, the institutions and incentives beneath that career cannot help but shape the writing.

It would be good if we could finance careers in public affairs writing while largely insulating authors from financial and career incentives. Substack’s subscription model may (or may not) prove an improvement on the listicle-inspiring ad model. But the direct pecuniary stake in subscribership it provides (and gamifies) will color what authors write. People are willing to fund their clique’s warriors, so offering political “red meat” is an obvious strategy to win subscribers. As Glenn Greenwald puts it, “They’re not paying because they’re getting something in return; they’re paying because they want to support journalism that they think…needs to be heard.” Functionally Substack shares a perhaps uncomfortable kinship with ActBlue or WinRed. A subscription-based model is going to encourage writers to to flatter the interests of especially more affluent readers. Substack subscriptions are expensive compared to paywalled conventional journalism, on almost any quantitative measure of writing unlocked. Finally, in my view, public affairs writing ought to be a public good, where authors contribute to a universally accessible, intertextual commons, rather than marketing paywalled silos. It’s not writers’ responsibility to bear the weight of this ought. We need to find ways to finance the people who cultivate the commons. But it is a commons that we want, rather than a labyrinth of paywalls or (worse) a few dominant publications everyone has to subscribe to.

In writing as in many other domains, I think “high-powered incentives” — extrinsic money rather than intrinsic goods like pride in virtue or quality — are essential at low levels but destructive at high levels. We expect and want baristas and warehouse workers to be primarily in it for the money, although of course they take pride in the quality of their work too. The tax and shareholder-value revolutions of the 1960s through 1980s destroyed American society by turning the people near the top of our social hierarchies into rapacious maximizers, and we should undo that, quickly. In writing and in general, external incentivizers can easily distinguish outright incompetence from a basic capacity to do the work. Above a certain level, however, quality is difficult to observe. Incentivizing putative correlates of quality encourages gaming, with a net effect that is ambiguous at best. At high levels, people’s “skin in the game” should increasingly become attached to broad, cooperative outcomes rather than narrow measures of behavior.

In light of all this, one way the Substack model might be improved is with caps and refunds. As with Substack now, there would be paywalled content, but the paywall would be stochastic. When a browser hits a piece, if it’s an identifiable subscriber it’s allowed through. If not, a quiet lottery decides yay or nay. The odds of denial would go down as the number of subscribers go up. At very low subscriberships, this would effectively be the current model, a hard paywall. At very high subscriberships, all content would effectively be open.

On its own, this would be a prescription for free-riding and death spirals. Why pay expensively to “subscribe” when other people have already unlocked the writing for you? However, what if the net cost of subscribing declines with readership, so that if many people subscribe, the contribution requested is very small?

A simple way to do this would to impose a compensation cap. Suppose an author requests $5 per month or $60 per year with a $150K cap. If she accumulates 2500 subscribers, she’s hit the cap. Thereafter, revenue from additional subscriptions gets distributed pro rata as a refund to subscribers. If she has 5000 subscribers, each subscriber would get $30 back, making the net cost of supporting the author only $30 per year. If she has 50,000 subscribers, the cost drops to $3 per year, literally spare change, just a quarter per month, and the writing becomes part of a wide-open commons. Once an author has hit her cap, she still ought to (and I think still would) promote her work and try to get people to contribute. But her incentives would become egotistical and altruistic: Egotistical because public affairs writers want their work to be influential and widely available (as long as they are also paid). Altruistic because encouraging new subscribers would decrease the burden on writers’ already loyal subscriber base, and because more exposure really might contribute to the process by which ideas and insights make the world a better place.

A gentler approach might replace a simple cap with an asymptotic limit. Each dollar contributed would go to one of two pools, author payment or pro rata refunds. The first subscribers’ funds would go almost entirely to author payments, but as cumulative revenue increases, the share going to the author would decline, and the share going to refunds would increase, so that the author’s payment as a function of total revenue approaches a horizontal asymptote. Under this scheme, the transition from extrinsic peddle-subscriptions-to-pay-my-rent incentives to more intrinsic and altruistic incentives would occur very gradually. If one wanted to maintain some degree of financial incentive for authors to expand their contributor base and reduce subscriber burdens, the asymptotic limit could be by an upward sloping line, but with a slope much less than one, so that at the limit say 10¢ out of every new subscriber dollar would go towards the author, and 90¢ to the refund pool.

If you’ve got better ideas than these, please contribute them to the commons! We want to fund a lot of writing, of high quality and from a wide variety of points of view. But we want those voices to be independent, not just of particular institutions but also of the incentives imposed by variable remuneration. The humans are clever. Surely we can figure this out.


The “cap and refund” idea owes inspiration to buylibre.org and Clark Evans, ht Sigfried Gold. It bears some similarity in mechanics and in spirit to a suggestion by Ryan Cooper that a nationalized music streaming service pay out artists “progressively” — i.e. at gradually decreasing rates per-play, in order to encourage a musical commons conducive to a broad creative class rather than the winner-take-all status quo. The Glenn Greenwald quote is via Jemima Kelly‘s reporting, but the paragraph where I embed the quote is too wordy to include a hat-tip.

I want to add that this piece is not intended as an anti-Substack diatribe. I read and subscribe to a bunch of Substack newsletters, including some of the controversial ones. Overall, I think Substack has inspired a welcome renaissance in less institutional writing. I hope the current renaissance is a step on a journey to better things, that its concentration on a single platform decreases and that it evolves in a less winner-take-all direction. But overall I am grateful for Substack, and competitors like Ghost. If I were a better writer, I’d consider trying to make a living on these platforms myself.

Convenient, compulsory, compensated

Voting should be convenient, compulsory, and compensated.

We should all want everyone to vote. If this is a partisan issue at all, it’s partisan for like five minutes. It’s not actually clear which party would be hurt if everybody voted, but if one party were disproportionately harmed, it would just realign a bit. The US political system converges to a 50/50 electoral power divide, which under universal voting would move a bit closer to a 50/50 voter support divide than obtains now. You might argue that makes universal voting anti-Republican for the moment, since Republican power relies more strongly on disproportionalities in our electoral system. But that’s not right. The representation skew caused by the Senate, or by the vote-wasting overrepresentation of Democrats in dense places, would not be affected by universal voting. Unless the we reform redistricting, universal voting does not prevent gerrymandering. Given the 50/50 rule, so long as there remain disproportionalities of electoral power, one party or the other is going to continue to rely upon them.

If universal voting does have a near-term partisan effect, it will be by virtue of how current non-voters would vote, which absolutely nobody knows. Researchers and political professionals have no idea, because there is by definition no solid data. Opinion polling is notoriously bad even at predicting the behavior of “likely voters”, who are more reachable and easily characterized than nonvoters. Democratic partisans sometimes presume on the basis of crude stereotypes that voting expansion will always be good for them. Why are you sure Puerto Rico would be a blue state? (It should be a state if its public wants that regardless.) Any inferences you draw based on observed correlations between demography and voting behavior are confounded by a motherfucker of a selection effect. People’s choice of whether or not to vote is not random, and nonrandom in ways unlikely to be orthogonal to partisanship.

My guess is that under universal voting, both parties would abandon the fetishes of the weirdest, least popular elements of their coalitions. Our existing system gives tremendously disproportionate weight to motivated voters. Some people argue this is good thing. Why shouldn’t a democracy take into account the intensity as well as the prevalence of citizen preferences? Probably it should! But we already have way way way way too much of that good thing. In soliciting support before elections, and with every call they take after winning an election, politicians are overwhelmed by the tyranny of concentrated benefits versus diffuse costs. Our problem in politics is not underrepresentation of small, extremely motivated interests, whether industry lobbyists or righteous activists. Our problem is that the preferences and interests of the very broad public get overridden by those groups, and our polity is therefore so badly misgoverned it is in danger of collapse.

Under nonuniversal voting, turnout and suppression are inevitably dimensions of competition. But they are invidious dimensions of competition. I won’t waste words persuading you that competing to suppress the vote of your electoral adversaries is morally wrong and bad for democracy. Even avid practitioners concede the point, and mostly pretend suppression is not what they are doing. But conventional turnout competition is bad too. Uncompensated voting is a regressive tax. At best voting is a fixed cost, more burdensome to the less resourced. Affluent people with stable schedules, ready transportation, and spare attention have a built-in advantage. In practice, voting is more expensive, in terms of time and hassle, for poorer people, rendering it even more regressive. The goal of motivating turnout encourages drama and outrage rather than deliberation over tradeoffs. If the margin I am trying to overcome is “check this box, not that one”, we can talk about the benefits of the program and how the taxes will work. If the margin I am trying to overcome is “lose work hours you desperately need to stand in line, arrange childcare for that, pay for gas”, well, I’d better talk about how you are our last bulwark against authoritarianism, about how the other guy is Satan himself so failing to show up for our lone, beleaguered warrior would be a betrayal, even a sin. The recent Presidential election featured very high turnout, but I’m not sure you’d describe the foremath as a period of exemplary democratic deliberation.

Some of the biases of voluntary, uncompensated voting can be reduced by making voting more convenient. But, in a system where turnout and suppression are in fact dimensions of electoral competition, convenience reforms are non-neutral. They are likely to be supported by the side that sees electoral advantage in them and opposed by the side that thinks it stands to lose. Partisans may often be overconfident in their theories of how these reforms will (un)tilt the field, but nevertheless. You end up where we are, with one side emphasizing “voter access” (convenience reforms) and the other emphasizing “voter integrity”, which provides a not-facially-evil pretext for opposing convenience reforms.

This is not a great place to be! Because electoral integrity actually is a real concern. The integrity of the US electoral system is and has been in bipartisan doubt for some time. Before there were Trump operatives making stuff up about Dominion Voting Systems, there were people like Jenny Cohn and, well, me, worried about unaccountable touchscreens and “ballot marking devices” produced by ES&S and Diebold. In 2016, I think it’s fair to say that many Democrats were not 100% sure that Trump’s slim electoral margin might not have owed something to Russian incursions that went beyond publicly reported exfiltrations of information from state electoral systems. There is no credible evidence for what Democratic-leaning media have dubbed “The Big Lie” (or what diehard Trumpists call “The Steal”), and Trumpists’ attempts to circumvent the electoral authorities and courts whose role it is to adjudicate such evidence were despicable. But absence of evidence is not evidence of absence. If, however mistakenly, your priors were very strong that Donald Trump had it in the bag, it’s not ridiculous that with margins as slim as they were and the US electoral system as “loose” as it is, you might not have been persuaded of his loss. Again, this is no excuse for failing to accept the result — Democrats who felt the 2016 election was illegitimate didn’t invade the Capitol. To the very limited degree they promoted circumventing the outcome with electoral college machinations, that too was discreditable. We go into each election with the electoral system we have, not the one we might wish to have.

It is between elections that we get to improve the system, and improve it we should. We do want an electoral system the legitimacy of whose results are more sure, which produce more and more certain evidence in forms accessible to the public. It’s not ridiculous to be concerned about voting machines. With in-person voting, it is a bit ridiculous to be concerned about voter fraud, because the costs and risks of going to a polling place and lying about your identity are (or can be made) very high relative to the marginal effect an individual might have even on very close elections. However, with remote voting, voter fraud becomes a more serious issue, as the possibility of mass rather than one-at-a-time identity theft increases the potential effect, and the ability to perpetrate the fraud without physically turning up at a polling place reduces costs and risks. If this sounds like a Republican talking point, it’s also precisely why we don’t have internet or smartphone balloting, and shouldn’t any time soon. It’s hard to “hack” thousands of bodies showing up at polling places. It might not be so hard to steal the internet credentials of thousands of people and vote on their behalf from some perch beyond US law. Mail-in voting sits between easily securable in-person voting and clearly not-securable internet balloting. It might or might not be right to universalize it, and the raging of a deadly pandemic might legitimately affect the balance of costs and benefits. But we’re not having a meaningful conversation about those tradeoffs while what we’re really arguing about is access versus suppression.

In other words, we won’t do a good job of balancing security and convenience, we won’t even be able to discuss that balance, as long as election technique is subsumed in partisan competition over turnout and suppression. Compulsory, compensated voting would eliminate those dimensions of competition, and render electoral integrity a technical dispute within which both parties’ interests would be broadly aligned. This would make safe convenience a reachable goal. Convenient, compulsory, generously compensated voting would eliminate the structural bias towards the affluent, as the compensation would be more meaningful to the poor than to the rich. Convenient, compulsory, compensated voting would improve American governance, which is too much in thrall to very motivated parties and attends too little to the more diffuse interests of ordinary voters. Convenient, compulsory, compensated voting would quiet the crosswinds that make it impossible for the United States to competently administer elections because administrative choices and election-rigging are too difficult to distinguish.

Alex Kovner has a post I really like, “Start with the State“:

Most democratic discussions start with the people, and attempt to build structures from them on the principle of representation. This has resulted in some real gains for democracy, most notably the creation of the U.S. Constitution and the democratization of Europe, with changes such as the reform bill of 1832 in Great Britain. Nevertheless, it’s hard to point to any meaningful improvement in democratic structures since the 19th century. All the improvement has been directly tied to representation through the franchise, namely including women and minorities.

Instead, why don’t we start with the state, and ask what it needs to properly perform its broad social role of service to the people?

The reason to want universal voting isn’t (just) because all the people “deserve” representation. To be clear, I think they do. But maybe you don’t. You may think that layabouts who don’t value the franchise enough to endure its inconveniences deserve whatever government they get.

But if it is to govern effectively, the state needs to know and give weight in practice to its whole public’s values and interests. To do that, it requires some form of deliberate, secure, widely-agreed polling that enfranchises its whole public. The state has a strong interest in criteria for enfranchisement not becoming a dimension of political competition, as that is not conducive of deliberative excellence and undermines integrity of the polling.

Universal, secure, legitimate polling is a necessity of practical statecraft, and the quality of practical statecraft is the hinge upon which all of our fortunes in fact rise or fall together. Voting should be convenient, compulsory, and compensated.

A heathen’s Easter

I have always been a devoutly irreligious person. I’m not an atheist. I think that creed assumes we know far more about our circumstances than we actually do. But I found very little appealing about the religious tradition within which I was raised, and I’ve not been drawn to others. However, in recent years, I have developed some reverence for a certain aspect of Christianity’s founding tale.

“Love your murderer” is, I think, an ethical aspiration. That one ought to love in some fashion every human being does not mean that people shouldn’t be held accountable. Murderers and other doers of foul deeds should be punished to deter others, should be tasked with restorative work where such work can be done, should be segregated from society and deprived of some freedoms as long as they remain a danger to others. Failing as a society to insist on those things would be failing to love the humans more broadly. (I hope it goes without saying that we should love our murderer because they are human, not because they are our murderer.) But when we punish, deter, or restrain our fellow humans, that ought to be an occasion of sad necessity, not joy or righteous vengeance. We should be very humble in any suggestion that pain we impose constitutes “justice”. Taking pleasure in the harm or punishment of someone who has hurt us is understandable and effusively forgivable, at a personal level. At a social or political level, however, it is noxious. It deforms us. It provokes us to terrible acts, sometimes in the name of, rather than in opposition to, the law. Individually, it is understandable when we sometimes take self-righteous pleasure in other people’s harm. I do, far too often. But it is the opposite of an aspiration. It is a lapse.

Obviously, the Christian story is an example of a person loving his murderers. And for that alone, it draws from me a certain respect. But a few years ago my wife and I had a child, and the part of the Catholic Holy Trinity I identify with shifted, perhaps egoistically. It occurred to me that there is so much focus on the son, who forgave his own murderers. But what about the father, who is called upon to forgive the murderers of his child? “Love your murderer” is an ideal I can at least conceive of aspiring to. But “love your child’s murderer”? Intellectually, the case is the same. We owe love to the humans, unconditionally and universally. That is the foundation of human virtue. But, as the kids say: I. Can’t. Even. I can’t let my mind go there. It is too great a betrayal, at an emotion level, of the star around which I orbit.

My theological sophistication is about candy-wrapper level. But for whatever it’s worth, I consider this aspect of Christianity’s founding myth or event remarkable, and underemphasized. “Forgive them, Father, they know not what they do,” represents a profound plea from the lips of a man being painfully murdered. That a parent, one with fire and brimstone readily at hand and a notorious history of smiting, would forgive is perhaps even more astonishing, even more wonderful.

The history of Christianity, especially at the social and political level, imperfectly evinces this ethos which I draw from, or project onto, the tale. Nevertheless, I think the ethos offers crucial lessons for us now. All of our political factions, even the ones who coined the pejorative term, slip frequently into “othering” one another. I take that to mean a withdrawal of the love, or even the aspiration of love, from some group or class of humans, often because “they” are purported to be vicious or guilty or dangerous, to have harmed us or our values or people we hold dear. There is a lot in our social affairs that needs changing, and there will be losers as well as winners from those changes. In a broad sense, I think if we act well and wisely, there will be many fewer losers than we fear, because our misarranged society exacts terrible costs even upon most of its “winners”. We reform society out of love for humans, to create scope for greater flourishing. But when people are harmed, whether transiently or durably, that counts as a cost, regardless of how wicked we persuade ourselves are the losers. That there will be losers is no excuse for inaction, in the same way that our love for a murderer mustn’t inhibit us from sober punishment. We owe a duty to all the humans. However difficult it may be to quantify human welfare, as best we can we must find ways of improving it. But the eggs we must break are losses to be minimized, not righteous smiting of the vicious. To whomever you are shouting at, owning, canceling, legislating against, you owe a duty of love. Aspire to love even your murderer. If you are better than me (and I assure you, you are), aspire to love even your child’s.

Happy Easter, to all those who celebrate it. And to all of those who don’t.

Update History:

  • 04-Apr-2020, 4:25 p.m. EDT: “there will be losers as well as winners, from those changes”; ” But to the degree when people are harmed, whether transiently or durably, that…”; “…the eggs we must break are costs losses to be minimized…”