This episode features our guest who directs the Office of Research & Analysis at the National Endowment for the Arts. Sunil sits down with Henry Oliver to discuss the state of reading in America, including the long-term decline in book and literature reading, the growth of audiobooks and digital formats, what these shifts mean for children and literacy, how reading relates to civic and social life, the importance of schools, libraries, and access to books, and much more.

New episodes of this podcast season will come out every two weeks.

TRANSCRIPT

HENRY OLIVER: I am here with Sunil Iyengar, who directs the Office of Research and Analysis at the National Endowment for the Arts. This means that he can tell us all about how many people are really reading books. Sunil, hello.

SUNIL IYENGAR: Great to meet you again.

OLIVER: You’re a poet, you’ve published an anthology of narrative verse, and you direct the Office of Research and Analysis. You’re familiar with the literary and literacy scene in a number of different ways. Give us your overall impression of reading in America.

IYENGAR: Thanks. Just to be clear, I’ll speak here on behalf of my role from the NEA’s perspective, from the National Endowment for the Arts, because in that capacity, I get to direct the Office of Research and Analysis. What we do there is we try to mine data and understand what statistics, among other sources, can tell us about arts and culture and society.

Of course, reading is a huge part of what we do, understanding reading. When you ask that question, I might answer it differently from my pastime perspective. In terms of the arts and my role at the NEA, I will say through surveys we’ve done—even before I came to the NEA, 1982, in fact—the first survey, major national representative survey of reading occurred and was directed by the US Census Bureau with the NEA.

We’ve been doing that survey periodically for many, many years, really trying to understand how Americans participate in the arts so we can serve them better through our programs. In fact, reading, I will say the reading numbers have really commanded a lot of public attention, as you suggest. Right now, I will just share—you want me to share?

OLIVER: Of course.

IYENGAR: Right back then when we did the 1982 numbers, it was about 60 percent of all adults who read a book of any type, not required for work or school. You could call it leisure reading. In 2022, in our most recent survey, we’ve actually found that that number’s now at about 49 percent, so just under half. When you look into people reading imaginative literature—so, poetry, plays, novels, short stories, a lot of the stuff you like, of course—we find that those numbers are even lower. In fact, the drop has been steeper. There’s been basically a 30 percent decline in the rate of adults reading those works.

OLIVER: 30 percentage point or 30 percent?

IYENGAR: 30 percent decline. In fact, now it’s about 40 percent of adults who read any form of literature. That’s poetry, plays, short stories, or novels. That’s, again, a 30 percent decline from back in 1982.

OLIVER: Is this the same as some of the other arts, or is reading seeing a bigger decrease?

IYENGAR: Actually, I would say over time there’s a couple of things. One is just by sheer percentages, the sheer percentage of people who read anything in the US is much higher than some of the different art forms we ask about. In other words, you could say that as a way of participating in the arts, reading still commands a very high percentage of adults relative to other art forms.

That said, in terms of the declines, yes, I think not only have they been sharper, but they’ve actually occurred more persistently, in both the share of adults who read literature and the share of adults who read books of any type. I wanted to focus on the literature part because that’s where we see this decline being the steepest right now.

I just said in terms of literature specifically, but if you look at novels and short stories, that’s particularly what’s, I think, driven down those numbers. It’s 40 percent, I think, of adults who are reading any form of literature right now—sorry, I mean books.

OLIVER: I shared a survey number online the other day saying that—I don’t remember the exact number—but quite a lot of Gen Z young women are reading. I just shared the stat, no commentary. I was flooded with people saying, “Yes, but they’re reading pornographic trash. It’s all romantasy. It’s not real books.” Now, I don’t really have a problem with that. Reading is good. Reading is reading. Whatever. Is that true? You’re saying there’s some big shift that the readers who are left tend to be reading that kind of thing and it’s less—

IYENGAR: We don’t know the details of the type of literature they’re reading within, say, fiction or within poetry or these other genres. What we also have seen through other data is that it seems like the people who are reading are maybe even reading a little more, which is interesting. People who are dropping out and not reading, there’s a larger share of those people, the general public.

OLIVER: The ones who are left, it’s more hours of book time?

IYENGAR: That’s coming from another data source. There’s another federal data source that we’ve supported in a study which looked into how much time people spent reading. On average, I think over, again, a very long period, like 20 years or something, it was something like 3 percent drop per year in the share of people who reported, on a given day, reading for fun. That could have been anything. Doesn’t have to be high literature or whatever you want.

OLIVER: Magazine, newspaper?

IYENGAR: Yes. In fact, they found that the people who were reading, they were reading for a longer period than for their counterparts 20 years ago. The few who are reading, or the relatively few who are reading, were reading more intensively, it seems to be, at least in terms of time. That’s a somewhat interesting finding because we don’t monitor sales of books or book buying, but from what you can see from industry reports, it does seem that there may be a difference, for example, between unit sales and total revenue.

A lot of publishing industry may be doing really well, but the number of units, we don’t know exactly whether that’s actually keeping up with the revenue growth and whether that’s actually contributing to more people reading or, in fact, the same people reading maybe more books.

OLIVER: Are people reading on paper, digitally? What’s the breakdown?

IYENGAR: We don’t have a great handle on the differences right now, but I can share that when we ask these questions, we’re neutral to the platform, just to be very clear about that. This is reading of any type, including electronic reading. One good question, it’s related to that, is audiobooks. In fact, when we asked whether people were reading audiobooks as well or doing audiobooks and included that in the number, it grows to 53 percent, but that’s still lower than, say, 55 percent that it was a few years ago, like five years ago.

OLIVER: You’re saying if you factor in audiobooks, the decline is smaller.

IYENGAR: It’s smaller, exactly.

OLIVER: Quite a lot smaller?

IYENGAR: At this stage, we don’t know because we’ve only been doing it for two survey periods. For that one differential, it was just a relatively small decline. It’s possible that if you add audiobooks, if it continues to grow, and if we count that as a percentage of all readers, then, in fact, the decline is mitigated quite a bit.

OLIVER: Actually, it could be that reading has declined, or it could be that a certain percentage of readers have just switched to audio, and the change is not that big overall?

IYENGAR: Right. In fact, we did a report a few years ago which was precisely about that, called How Do We Read? It was all about audiobooks versus print books. From the limited data we have so far, it does seem that adding audiobooks to reading, if you include that in your definition of reading as a whole, then, in fact, it’s still a decline from the share who did that five years ago, but it’s a much smaller share. It does suggest, in the out years, it may very well be that audiobooks mitigates the decline much more, and maybe it even puts us on par with the trend line for the past or traditional reading.

OLIVER: Now, there’s obviously a huge online debate about, do audiobooks count as real reading? Do you have or have you seen any good data, or analysis, about retention rates, comprehension rates? Do they change across print and audio?

IYENGAR: That’s a great question. I haven’t seen anything about print versus audio in terms of how it improves comprehension or learning. I do know though, that—and I think you’d agree—there are almost different complementary skills that are fostered through those different ways of engaging with literature. On the one hand, it’s listening, it’s attentiveness with audiobooks, it’s imagination to a large degree.

With reading, you’re dealing with lexical patterns of recognition and different cognitive processes, so I would be surprised if it was equivalent, but clearly, I don’t know enough to say whether one is superior to the other in terms of overall comprehension.

Digital reading is a little different, though. With digital reading, we do have some data, particularly from other studies that have been done internationally, suggesting that digital reading, especially for early readers, it may have a negative effect on overall comprehension. That is to say, if you come in reading through digital reading only, it’s not clear that that’s really improving your overall comprehension, whereas it is clear with print reading in early stages, especially of learning how to read, childhood development, that it does benefit total reading comprehension.

OLIVER: Is that because of some weird magical factor of looking at a screen is bad, or is it because you’re possibly just likely to be reading less challenging material?

IYENGAR: It’s a little of both, actually. The studies I’ve seen talk about perhaps less challenging words and vocabulary that’s maybe accessed through purely digital means, but that doesn’t seem to be the primary factor. It also seems to be that through digital reading, there’s a lot of skipping around. Maybe there’s hyperlinks, or you’re led to another place, the jumping around.

There’s a whole theory about shallow processing, the idea that perhaps people cognitively are not imbibing as much when they’re reading that way, especially in early stages when their domain knowledge might not be as big as it is later in life. There’s a tendency, at least in some studies, to show that early reading, especially through digital formats where you’re jumping around a little more, you’re being invited maybe to go to another webpage and come back to where you were, that might be great in some settings, maybe for some kinds of informational processing, but not necessarily for reading long text.

OLIVER: It’s a question of developing a different sort of skill, a different type of reading skill.

IYENGAR: Yes. I think this is truthfully something that needs further research. The verdict is not all in for purely print all the time. One thing, again, I think it’s like what we were talking about with those people who are still reading today—and there are a lot of them, of course—who are reading print and digital. I think that’s what you’re going to see, I think, is much more of this complementary function.

Some of the studies I’ve seen show that if people already have a good grasp of reading and print, then they’re much more likely to gravitate to digital and maybe even choose that as their primary reading method. Nothing seems to hold up their comprehension in that regard. It’s more people who go straight to digital without having had that print experience.

OLIVER: If you start with digital aged five to 10, you end up as a very different sort of reader than if you are a print-based reading learner, and switch in your teens or 20s.

IYENGAR: Even the way we assess comprehension, maybe that’s going to change in the future because the very things that we think valuable retaining may change depending on what the formats are and what the types of texts are. Right now, it does seem, at least from the studies I was looking at—and they’re largely international studies, like meta-analyses of hundreds of thousands of people included in these studies—where they seem to be consistent on digital reading is at best modest benefits for people’s comprehension, but at worst, a full-on negative effect on particularly an early-stage reader.

OLIVER: For the young children?

IYENGAR: Yes.

OLIVER: There’s this constant thing in these debates where people are interested in two different things, I think. The first thing is, are people reading novels? What’s the state of literature? How are the arts doing? The second thing is more just to do with literacy, the level of reading skill, comprehension, and the downstream effects of that socially, politically, in the workforce, whatever. It’s very hard to keep those two things separate, isn’t it?

Can you give us some sense, from all the data you have commissioned and looked at and worked on, are these two things going in a trend together, or is it that a lot of people like us, who like literature, are getting very scared about the decline of reading, but in real money, it’s not making that much difference in day-to-day society? What’s the balance here?

IYENGAR: I think it’s very important to keep that in mind because people who read anyway, and agencies that promote reading, are doing it because they know it’s intrinsically valuable, when you do talk about the why, for people who have no interest or idea of why this is something important, you do have to tie it to measurable outcomes.

There are data, and we’ve produced some of it, showing that, in fact, readers tend to be more civically engaged, more socially engaged, and they tend to read better too. There’s a bidirectionality here, of course, because people who read well to begin with are going to gravitate, a lot of times, to reading and reading more. It also works the other way, that people who go to reading and want to read, get better as readers. It is reciprocal.

In terms of other instrumental kinds of outcomes, we know that readers are, like they said, more likely to be engaged in their communities than people who don’t read. A lot of this is also overlapped with the variable of education, though. What kind of education they have, there’s all kinds of socioeconomic graphic variables in this. When we’ve controlled for some of that, it still seems that reading is very positively associated with some of those kinds of civic benefits of volunteering, taking part in school, community events.

In school settings, we’ve seen that with out-of-school activities correlated with things like that. I do think, though, that just knowing how to read and the literacy component is obviously etched into the way we’re supposed to conduct a democracy and civic life and being a member of the democracy and contributing in all other ways. I think it’s very much tied with the health of the nation in a lot of ways that are writ large, even if sometimes we have to keep pointing to those connections through research.

OLIVER: There is one thesis that goes around is that as people read less or as fewer people read, people generally get worse at media literacy, at being able to evaluate the truth of claims. This leads to the general instability and craziness of the world that is being experienced in multiple ways. There doesn’t, to me, seem to be much positive evidence to link these two things. It seems to be a sort of these two things are happening, therefore. Do you know of any such evidence?

IYENGAR: Linking the people’s failure to read with?

OLIVER: It could just be that smart, conscientious people will always be more civically engaged, better at evaluating the things that politicians say, better at knowing that this news channel is biased in this way or whatever, and that it’s not because they read. The reading decline is like a graph that gets shared online, but it isn’t necessarily—

IYENGAR: Tied to these things?

OLIVER: Yes. If reading was going up 3 percent a year instead of down 3 percent a year, is it that in five years, do we have research that really suggests that, yes, society would be very different.

IYENGAR: I think you’re right to point out that this is all correlational, there’s no causal effect. We’re not showing that if people do read more, automatically their circumstances are going to improve. It’s just that over time, even when you control for other factors, you often see that people who read more, and read literature even more, tend to be correlated with outcomes that objectively seem better than outcomes for the people who don’t, according to certain outcomes.

We look at economic factors, outcomes, their own earnings and civic contributions, things like that. There’s a lot that we’re not capturing. It would be too much to say that if the chart’s going down, you’re automatically going to see everything go to hell in a hand basket or whatever. That’s not really what we’re trying to articulate either. I think it’s just these are indicators of how engaged people are with the act of imagination.

I think sometimes the word empathy is overused in this context, but it’s the ability to inhabit another perspective or viewpoint, or character when you read. I think those kinds of functions, and even the quietening of the mind to have a space where one is focused and has attention, those are skillsets, if you want to call it that, or attributes that I think can easily get eroded.

The other thing I’ll just share is—you know this—Maryanne Wolf and researchers who are neuroscientists have talked about, reading is not a natural activity. It’s really something that humans do that takes a lot of work, cognitively, neurobiologically, and semantically. It’s something that we can’t take for granted. Sometimes you do need some markers or some trend lines to monitor, to know whether we’re on the right track.

OLIVER: This is that wonderful book, Proust and the Squid.

IYENGAR: Yes.

OLIVER: That’s largely about children. I want to go back to that point about children reading on screens. I looked at some of the studies and the data about this, and it’s not clear to me what’s really going on. If a child is, let’s say they’re eight, they’re using WhatsApp, they’re clicking hyperlinks, they’re doing that kind of digital reading. Obviously that’s probably not sensible after a certain threshold, but if they’re given an iPad and they’re going to read an encyclopedia article on the iPad, is that really any different to if they’re just given—we had a paper encyclopedia.

IYENGAR: See, this is where it gets tricky because when you censure digital reading like that, you’re saying that every single category is just useless.

OLIVER: The research isn’t picking apart—

IYENGAR: No, you’re right. It’s not quite as nuanced as that. A lot of even these great meta-analyses, I think, are categorically talking about digital reading. Some of it is only focused on nonfiction, information-type stuff. I’m sure there’s subtleties like that, but if you want to paint it with a broad brush, I think it’s fair to say that some of the characteristics of digital reading that involve skipping around, there’s been studies of eye-tracking movements and how shallow or the skimming patterns that go on with digital reading versus print reading.

I think the jury’s out on long-term effects on those, but even anecdotally or even just understanding some of the research that we’ve seen, as limited as it is, suggests that people are not engaging as deeply with that particular text. Maybe that’s okay with just an encyclopedia entry or a timetable for a metro train or something, but when you’re getting into something that maybe is more substantive, we think—we, the society thinks—it deserves greater scrutiny or greater attention, like maybe a newspaper article that’s about a candidate; or if you’re talking literature, a work that’s going to move you to a new height of inspiration or understanding.

I think those kinds of works probably deserve much more attention, and it’s not clear yet that some of the digital formats enable that. Now, I do think e-books and digital reading, there’s nothing to suggest that it’s not doing as well as print books in terms of comprehension, but I think when you talk about digital reading as a class versus just print reading, I think it’s easier to point at potential gaps in digital reading.

OLIVER: If we want to take the research we have and make as informed a decision as we can about how children should read—books are good, a Kindle that only has books and doesn’t have any other stuff on it, that’s basically probably as good, an internet connection that they use for encyclopedias and other things is probably a lot better than social media, WhatsApp, whatever.

IYENGAR: I think that’s very well put. I think some of those are really good for things. It’s possible that I would recommend maybe a kid to—I’m not speaking for a policy thing—but I would say it may be that in some cases, the encyclopedia online is going to be way better than going and trying to find the book that bears that particular article in it. It depends on what it’s being used for, what kind of information is being retrieved. I think information retrieval is, obviously, in a different category in a way than reading, not just for information.

Again, the surveys we do are about books and also what I’m calling imaginative literature. Literature that’s novel, short stories, poems, and plays where, yes, there’s a different kind of information in it. Like William Carlos Williams said, “You can get the news out of poems.” There is news there, but it’s not really information retrieval as we’re talking about it.

OLIVER: Do we know what predicts whether someone will become a reader?

IYENGAR: Shortly, I don’t know exactly what the answer is to that, but I know studies have shown that clearly high levels of parental education, socioeconomic affluence doesn’t hurt, but I would say, clearly access to raw materials. In fact, even things like the numbers of books in a home, I don’t know if they’re dwindling now as a society, but that has been highly correlated with, in fact, reading ability and propensity to read as well.

OLIVER: Does it matter if there are books in the home versus the child is regularly taken to a library, has a library at school?

IYENGAR: I don’t know enough to know those differences. I don’t know if there’s a big difference there.

OLIVER: Access to books?

IYENGAR: Having them around or even having the ability to access them, I think, is a big piece of all this. I don’t know if you’ve ever tried to trick a young child into reading. Part of it is just having a book around maybe in a place where they wouldn’t expect it.

OLIVER: I have lots of tricks, yes.

IYENGAR: There are many like that, I’m sure. I think that’s another thing, is having the proximity to these points of engagement.

OLIVER: Are libraries more or less important than they used to be?

IYENGAR: We actually don’t conduct surveys directly with libraries, but we do look at some of the library data, there’s a public library survey. Yes, they seem to still be doing really well. It’s possible that maybe a lot of what’s attracting people to libraries for a long time was internet access. I don’t know if that’s the case anymore, but a lot of that is also in the mix like engaging with technology in these libraries.

I know that, of course, as we were just talking about, going there and engaging with books and checking out books, it remains, I hope, a major part of what they do. I think a lot of libraries are also changing a lot of what their functions are in their communities. They’re no longer just repositories of books. They’re definitely engaging through community activities, education. They’ve become artistic venues as well. I think they’re definitely very dynamic. I don’t know enough about the finances or where they’re headed right now, but I will say that from all the signs I’ve seen, they’re still going strong.

OLIVER: One of the things in some of the historical data I found is that there was an increase in reading literature between 2002 and 2008. It was this funny bump in the graph. A lot of that was brought about by young people reading more literature. Do we know why that bump happened?

IYENGAR: That’s a great question. The questions I would get when that came out is, “Was it the Harry Potter effect and all that stuff?” I really don’t know. It was a really interesting bump. I will share that, soon before I came into the NEA, the chair there, Dana Gioia at the time, had commissioned a report called Reading at Risk, because he basically found that the survey data that I was just describing, in 1982, that number I gave—

OLIVER: 60 percent.

IYENGAR: —it slipped to 57 percent.

OLIVER: Yes, the graph is quite steep.

IYENGAR: Yes. That was a pretty sharp decline from ’82 for reading literature; it was 57 percent to 47 percent of adults. That was between 1982 and 2002. He said, “Let’s just do this report. Let’s get people to understand this problem,” which he saw as a national crisis. The NEA teamed up with publishers, bookstores, libraries, and really tried to get ahead of this, and did things like national book reading campaigns. The NEA founded something called The Big Read, which still goes on. It’s very successful. I think it’s engaged 6 million people through its programs, 2,000 or so programs all around the country.

That was something that started, and there were a lot of such programs around that time, between 2002 and 2008, when that study came out. I’m not saying it all can be laid at the door of NEA’s programs, but I know there was a wholehearted attempt by a lot of people in the literacy community and in the literature community to try to get more people to read. I don’t say it’s attributable to that one thing, but I think it was part of a movement that happened over that brief period.

OLIVER: Now, the other thing that surprised me is that the adult audience for poetry seems to have grown in recent years, or at least remained pretty steady. It’s gone up a bit more for young people than for adults. That’s quite shocking.

IYENGAR: It’s interesting. I think it was 12 percent 20 years ago or so, and then it dipped, and then it came back up. Not quite to the original level of 12 percent of adults who read poetry, but it’s still 9 percent.

OLIVER: We’re between 9 percent and 12 percent.

IYENGAR: Yes. It’s about 9 percent or so who, in the last survey, read any poem in their last year. Something I should tell people is when we ask these questions, we’re asking them to think back to the last year, 12 months, and did they do that activity? Did they read a book? We don’t say they have to have finished it. Did they read a book at all? Did they read a novel, short story, play, whatever?

OLIVER: You know my favorite Samuel Johnson quote, it’s in Life of Boswell, and he’s talking to someone about how he read this book, and he didn’t think it was very good, and he didn’t finish it. His companion says, “What do you mean you didn’t finish it?” Johnson goes, “Sir, do you read books through?” Appalled that anyone would finish a book.

IYENGAR: Exactly. Poetry reading, it’s interesting because at the time—and there were a wave of these news articles that came out when we came out with these findings suggesting it was the—I don’t know if they’re still around, the Instapoets or the Instapoetry.

OLIVER: My follow-up question is, was it Instagram?

IYENGAR: Yes. That was a big theory that that may have been some of it. Again, I don’t know, but I think there were a lot of people at that time who were very happy to see that up. Again, it tells us these aren’t irreversible declines. That gives some hope. I think again, just to say novels of short stories right now, it’s, I think I said 38 percent of adults who read one of those, and that’s pretty low. The thing is, I didn’t share this, but when you look at men versus women, it’s just a starker decline. It’s like 20 percent.

OLIVER: I really want to ask you about that because the 20 percent gap, the gap has remained the same while the other numbers have gone down. What is that?

IYENGAR: The percentage of men and women reading are reading much less than they did before in all these forms. However, women are still reading more than men at a higher rate, and at 20 percent point split. That’s persisted. I guess we should be worried, especially about the men here, because I said it was 38 percent of adults who read a novel, short story. It’s more like 28 percent for men. It’s going lower and lower.

One thing I will share with a lot of our arts attendance questions, even for arts participation, you often see women doing a lot more of these activities than men.

OLIVER: By participation, you mean going to a concert?

IYENGAR: Largely, attendance. Some forms of art creating, we see men actually having a higher rate, but a lot of times the women are leading in terms of the sheer number of people.

OLIVER: Where do men lead in art creation?

IYENGAR: That’s for certain things. I know that occurred when I was looking at something like music or jazz, particularly. You can see some specific forms where you see that happening.

OLIVER: Dad’s in a band.

IYENGAR: Maybe, yes.

OLIVER: Interesting. Then you mentioned The Big Read. You also have the Poetry Out Loud scheme. Both of these schemes seem to have done a lot. What is the success? What’s the factor?

IYENGAR: I think there’s a real rich community element in all this. The NEA gives grants, and through its partners, supports these events, where it’s not just one lone nonprofit organization who’s trying to make a difference. They’re partnering with others to make something happen. For Poetry Out Loud, for example, it’s schools and the state arts agencies.

Every state has a state arts agency. They are, in turn, our partners with this. It’s like a pyramid structure of a competition, like the Spelling Bee or something, for the Poetry Out Loud recitation contest. There’s a whole infrastructure there that’s mobilized with our funding and with our partners. Therefore, it becomes almost like a competitive event.

If you’ve ever been to a Poetry Out Loud event, I’d highly recommend it if you’re in any state in the country, and then the national ones here in DC. There’s a cheering section. People are really excited about hearing a poem being recited. It’s done with such flourish. There’s poise, there’s things you look for in an arts performance.

There’s some great outcomes for the students. They often will report afterward how much it made them love reading and love literature, but also feel engaged in their communities and in life. There’s a lot of that. I think there’s a spillover benefit of bringing the literature out into the communities and not so much assuming that it’s happening all in solitude, although, of course, a lot of reading is happening in solitude. There’s ways to bring that social element to literature that I think we can uniquely do through arts programming.

OLIVER: For these two programs, how well do we understand whether the young people involved were already readers, or are being brought to become readers, or are becoming more serious readers? Do we have a good grip on that?

IYENGAR: With The Big Read particularly, we know that all kinds of people get engaged in these programs from nonreaders to readers who don’t read that much, to people who are really passionate about books. It’s usually like they’ll choose a book as a community to celebrate and to do arts programming around. I think with America 250 this year, there’s a lot of American literature that’s really in the forefront.

I think that that is very variable. There’s talking about books. There’s maybe book clubs around it. There’s social events. There are people who aren’t necessarily in it for the literature, frankly, but are just there to have a good time. In the process, they rub shoulders with the work because it’s a theme of an event, or maybe there’s a speaker who comes to give a little lecture on that book.

They participate in a way that’s not quite directly reading the literature. I think these are other ways that people can get at reading. Then hopefully down the road, and we don’t know yet, the follow-up is, are they likely now to read if they didn’t read before?

OLIVER: I think you said 6 million children have done The Big Read?

IYENGAR: Not children, people.

OLIVER: Do we know how many of them were low-volume readers of any sort before they joined the program?

IYENGAR: We don’t know that from that data. That’s a great question. In fact, that’s something I know that, when he established this program years ago, Dana Gioia, was really big on understanding. We did some surveys at the time to understand the initial years of The Big Read, whether people were actually changing their reading practices.

What we learned is a lot of people were changing their attitudes toward reading. We don’t know if they continued reading because we didn’t continue the study. What we saw was they were much more likely to say they were going to go to a book club, or they were going to, next time go to the library even more often, and things like that.

OLIVER: How much do we know about reading habits among people for whom English is a second language?

IYENGAR: Oh, great question. Unfortunately, we don’t know a lot at the national level. As I said, with our survey data, when we break it out, for example, by race and ethnicity, we can see that certain demographic subgroups have much lower rates than the national level rate that I gave. Some of those may, in fact, be people whose first language was not English.

OLIVER: We don’t know if they are reading in their first language?

IYENGAR: We don’t know that. The survey does not specify a language. The Census Bureau has ways of getting at households where the first language isn’t English. I don’t think we have enough of a sample to know what, in general—that’s a great question—people whose first language isn’t English, how they’re doing, whether they’re reading their own literature, whether they’re reading English literature. We don’t know.

OLIVER: We don’t know if among people for whom Spanish is a first language or a co-language, we don’t know if they’re doing a lot of Spanish reading?

IYENGAR: It’s an analysis we could do with these data. It’s very interesting that you asked that. I think it’s worth pursuing.

OLIVER: One thing that really stayed with me was when I learned that sales of poetry written in Irish are actually quite a lot more robust. There’s a strong community of people who still read that, which, I don’t know, it felt unexpected to me.

IYENGAR: Was that in the UK?

OLIVER: I think it’s in Ireland itself.

IYENGAR: Oh, in Ireland. That’s great.

OLIVER: It’s very interesting.

IYENGAR: That is very interesting. I’d be very interested, as a reader, to know a lot more about the sub-genres and what kinds of groups are gravitating to what kinds of reading. With these surveys, as you can appreciate, they’re an instrument that can sometimes be a blunt instrument because it’s a quick survey, and you have to ask it of thousands and thousands of people.

Tens of thousands of people are getting these questions. You only have limited time with them to ask these questions, especially since we work with the Census Bureau, which has its own survey, so we tack onto their survey. Not to bore you, but I’m just saying that there’s a lot more questions if we had the real estate we would love to ask on the survey.

OLIVER: I wanted to get to this because there are a lot of surveys being pushed in the media about the decline of reading, whether they’re your data, other people are doing surveys. There’s so much of this. There are so many graphs showing, “Oh my goodness, it’s going down.” One thing I’m getting from you in this conversation is that we don’t know everything we would need to know, and that we can make some tentative conclusions, but we should be open to the idea that the real picture might not be quite what it looks like.

IYENGAR: I think you’re right. The only thing I would caution about is we’ve been asking the question, in some ways, the same way for years and years and years.

OLIVER: That, do you read a book?

IYENGAR: Yes. Do you read a book of any type? It could have been any kind, et cetera. Now, of course, the ways people read have expanded enormously since 40 years ago or 30 years ago or 20 years ago. I think there is something to the trend line. There is a marked decline in the general population. Again, the thing to remind people is, it’s the general population.

People often say, “People are still reading. I’m seeing people on the metro reading Middlemarch.” I’m like, “That’s great. Where do you live. What’s your zip code?” I think you have to talk about the entire country, and it’s a pretty large country. If you see the rates eroding over a period of many years and somewhat—there was a blip here and there, but it’s pretty much a particular direction—I think it’s worth maybe sitting up a little more, then say, a poll, which is a snapshot survey, and it says that things seem to be fine because 80 percent of people are reading. I do think there’s a lot of nuance, though, that we have to unpack. There’s a lot more we’d like to know as researchers, and I think there’s a lot the policymakers could know. I would say there’s a preponderant amount of these data showing that fewer and fewer people are reading works of literature, as we broadly define it, and also books in general.

OLIVER: Do we know from time use surveys what they’re doing instead?

IYENGAR: We know some of the things you would imagine, TV, going online, all those things take up much more time in an average given day.

OLIVER: What’s the biggest? To me, it always seems that TV is the enemy of reading, not really the internet.

IYENGAR: It was TV. I think the last I looked, it was like close to two hours.

OLIVER: It’s more like three hours, I think.

IYENGAR: I think it’s like that. I don’t know what the latest was. For reading, it’s something like 16 minutes on average. This is not that everybody’s spending 16 minutes. They’re taking average numbers. Sixteen minutes a day out of a day, leisure time, I have to stress that, so they could have been reading for work. Sixteen minutes versus two or three hours for TV.

That’s staring you in the face right there. Part of the reason we do these surveys is because we want the public to be aware of how the country is doing on certain indicators because it’s discretionary time. People have limited leisure time and different people from different backgrounds have even less leisure time than other people. It’s not an equal attribute. It’s scarce.

We want people to know that there are things you could be doing with your leisure time and there’s opportunity costs to that. Reading is probably one of those things that’s high on the list of something people could spend more of their time budget doing and maybe reap some rewards.

OLIVER: This general idea that it’s social media and phones that have taken away time from reading, that’s not necessarily reflected in the time use surveys? It’s more likely to be TV or it’s a bit of both?

IYENGAR: I think it’s a bit of both. I don’t know that the time use survey actually has a clear read on that social media stuff. I’ll have to go back and look. That’s not our survey. It’s the Bureau of Labor Statistics survey that has that. I wouldn’t be surprised if social media is now a bigger share of what they do.

OLIVER: We don’t know if social media is taking away from time that would have been TV, right?

IYENGAR: That I don’t know right now.

OLIVER: It always seems to me that if you’re watching short-form videos on Instagram, you would otherwise have been watching Netflix or watching HBO.

IYENGAR: That’s what I want to go back and look at, is when they ask about TV, I can’t imagine it’s sitting in front of a TV set.

OLIVER: I think a lot of is streaming.

IYENGAR: Streaming or it could be in the middle of a social media post. It could be very much intermingled with social media activity. I don’t know how they ask that question. It’s easily discoverable though because it’s on their website and everything.

OLIVER: The basic takeaway from the research you’ve done and that you’ve read is that it’s not so much the internet that killed reading as television and radio?

IYENGAR: I don’t know. You look at the time use survey and you keep seeing this discrepancy of, on the one hand, TV and reading, maybe social media in there. You might conclude that that’s the big bear, TV. I think the reason I’m hesitating a little is because, again, this is all correlational, but when you look at the co-occurrence of this with the rise of social media, and particularly some of these declines happening and being accelerated during a period when social media is even more prolific, I guess we just question whether there’s any kind of relationship there because so many educators and others have attended to perhaps excessive social media usage, especially among the young, eroding certain patterns of cognition. If that’s the case, then we would assume that some of that is bearing out in these reading numbers as well. I don’t have a hard answer for that.

OLIVER: On this point about patterns of cognition, there was an email posted on Marginal Revolution last year from a teacher. He said his students are not good at linear reading anymore, which is, he gives them Aristotle and they can’t just read the book and understand it. They’re much better at finding connections between texts and spotting these patterns across, which he said is maybe because that’s what they do online. They constantly move between things. To what extent do you think we’re seeing reading is changing? It’s not dying; it’s becoming something different?

IYENGAR: Maybe we’re underselling some of the benefits of digital reading. It’s just that I think it’s important really to have both. I guess I don’t know the context, like if the person—assuming this is higher education, these are probably people—

OLIVER: I think he’s a high school [teacher].

IYENGAR: Good. I think that there’s clearly benefits to that kind of reading in knowledge retrieval, again, as we were saying, but also to make new discoveries en route to reading a text. I guess the question remains if that substitutes for what he’s calling linear reading, if the benefits are going to be less weakened. And what are the benefits we’re talking about? Retention, attention, the ability to form one’s own idea or relationship to the material, and true engagement with the text.

I don’t know if it’s being compromised or it’s being expanded. I suspect that if people already enjoy reading, as you said or suggested, that it is circular, people enjoy reading and they go to reading anyway, then they’re probably getting some benefits out of that type of reading. It’s just that if that supplants what that person called linear reading, I guess I would be a little concerned.

OLIVER: Do you think there’s a problem that children are made to read books in school that they hate, and that that kills the love of reading?

IYENGAR: That could be it. I don’t know. I think everything has to be questioned. I don’t know that it’s texts are not the ones they like or whatever. When I was looking at the men and women split, part of that, I was wondering, is it that they’re not directed to the right things they may like.

OLIVER: Does that split start in childhood?

IYENGAR: Yes, it does start in childhood. Actually, what I was thinking of just now is just getting back to kids, and this isn’t about gender, but I’m just saying that among nine-year-olds, it turns out like 40 percent of them now say they’ll read for fun almost every day. That’s pretty good. That used to be 53 percent 10 years ago. Similarly, for 13-year-olds, I think it’s half of the share of people who said they would read for fun every day has gone away. It used to be 27 percent, now it’s 14 percent.

I think there’s definitely this decline that I’ve been talking about in adults, the survey we’re doing is an adult survey. You can look at other surveys that have been done, again, by the government through the Department of Education, and you see a similar pattern. Getting back to this issue of, is it just a blip and is it irrelevant? I don’t know if you can say that, like what if society’s great and everything’s all right?

You see so many of these things through pretty legitimate federally—not everything in federal is necessarily legitimate—but I’m saying these nationally representative surveys. The time use survey, and the Department of Education surveys—

OLIVER: —they’re all pointing in the same direction.

IYENGAR: They’re all pointing in the same direction, and they’re tending to show that split between men and women, boys and girls. Again, both of them declining, but with that split in terms of girls reading more than boys.

OLIVER: How important is it for parents to read to their children?

IYENGAR: Oh, it’s extremely important, we know. When you said predictor, I think that’s another one.

OLIVER: That’s a big one?

IYENGAR: That’s a big one. I think the reading to children piece, if you look at that time use study, it was a very small percentage of parents who did that. That was, I think, one of the things the authors of the study noted, is they were concerned about the fact that so few adults—I don’t have the percentage, but it was in the single digits, very low—who said they read to their children.

OLIVER: Do we know that it’s gone down? Because I know that anecdotally it feels like it’s gone down, but I say this to a lot of parents and they say, “Who are these parents who’ve got time to read to their kids? No one ever read to me when I was a child.”

IYENGAR: That’s true. Actually, this time use study, didn’t find a change in the share of people reading to their kids, but it was a low share to begin with. That’s what their point is.

OLIVER: It might be that it’s always been low and we overrate it because we love reading to our kids, but actually it’s having the books around and they do it at school, they’ll find a way?

IYENGAR: Right. I think we can overplay the social benefit or the instrumental benefit of doing this stuff. Another angle or piece to think about is our artistic heritage, if you will. Like just having access to the great things that have been done in the past. You could probably quote some really good philosophers on this. I guess what I’m saying is, are we forfeiting access or precluding access to great ways of participating in the arts? That’s what we’re trying to do is bolster and give people more direct engagement with arts product, if you will.

OLIVER: That was Adam Smith’s justification for—he didn’t call it the NEA—but when he said the government should fund something like the NEA, he didn’t think it would improve society or be wonderful. He thought education was instrumentally useful, but he basically said it was for the benefit of the citizens and it would be good for them individually.

IYENGAR: I think it’s a very humane way to think about it.

OLIVER: What about the international comparison? Everything we’ve said has been about America. Are we an outlier in this country? Is it the same in Europe? What’s the play?

IYENGAR: I do remember years ago—and again, I’m sorry I don’t have this in the top of my head—but there’s something called the PISA, which is an acronym for an international assessment of students across the country. I don’t believe they look at propensity to read or whether people read for fun, but they do look at reading scores. If I remember, we were in the middle of the pack, the US was. I’ll have to go back and look, but Department of Education, I think, used to be involved in that. I do know that internationally, I know in the UK, they’ve been having similar issues with reading rates. It’s been widely publicized. I haven’t looked at too many other countries at this point.

OLIVER: If you could implement one new policy to get the number to go up, what would it be?

IYENGAR: I can’t, as a mere government official right now, say what policy I would implement. That wouldn’t be my place.

OLIVER: Does the research suggest what such a policy might beneficially—

IYENGAR: It takes an infrastructure of support, so you need to have—

OLIVER: There’s no magic wand.

IYENGAR: There’s no magic model, but you do need to have some basic supports. I think you need to have community reading, like libraries, bookstores, book fairs. You need that kind of stuff. You need schools that value literature and ideally rich breadth of literature, reading of all types. You need to have places where kids can sneak away and read the nonofficial books. You need to have all that kind of stuff.

Of course, digitally, you need that access through digital media. It is there. There’s a lot of great places to go. It’s just that, as we all know, you need a guide, you need a Virgil or something to take you through the internet, to show, especially young learners, where to go for this kind of material. Some schools and a lot of educators are really good at directing them, but I’m just suggesting that at this early stage, especially given what we’re seeing with the numbers, it probably behooves us not to close the door on digital reading, but definitely don’t close the door on print reading at that stage, I would say.

OLIVER: If the takeaways are something like, the younger they are, the more print they need—

IYENGAR: According to some studies, yes.

OLIVER: This is all research-based. The younger they are, the more print they need, social infrastructure is very important, institutions are very important, but we should bear in mind that the research is not definitive about digital audiobooks, print, the whole balance, and we need to try and get the best of all of these aspects?

IYENGAR: I think so. I think it’s a more and more kind of thing. Cultural researchers refer to the omnivore theory, the idea that when people get hooked and get into one particular art form, a lot of times they’ll hop around and they get engaged with some other art form because they’ve had the bug.

I think that’s something we want in readers because it’ll expose people to many, many more ideas and portals of imagination than just one route alone. As you said, there are also measurable outcomes that we’ve seen over time. We can’t say it’s causal, but in terms of those people who tend to read more, also seem to have, in aggregate, certain economic outcomes, civic outcomes, and social outcomes that are more favorable than those who don’t.

IYENGAR: There’s a lot that we still don’t know, and I think it’s really important that we don’t be complacent.

OLIVER: Are we going to find it out, or are we just going to have to live with the mystery?

IYENGAR: Surveys are getting better. We’re hoping to redesign the survey, so maybe in the next few years we’ll have something better. The thing is with government surveys, there’s always a bit of a lag because we do the survey, and then the data comes out. We have to process it and do the write-ups. Maybe AI will help. We don’t know.

Then at some point, the report comes out, and it’s maybe a couple of years after the actual survey was conducted. I do think there’s a lot of smart ways now people are getting data in more organic ways, through online transactions, through all kinds of other means that I think would greatly enhance what we know about reading in the future. I’m hopeful that we can tap into some of that stuff in the future.

OLIVER: It’s good to have the lag because everything is so driven by the current scare story, the current thing.

IYENGAR: That’s true. I think what’s great about these surveys, even if I sometimes critique our own work, or I often do, it’s large. These are large shares of the population, and it’s truly nationally representative. They do everything they can to knock on the doors and get the right people into the sample frame. The fact it’s been done periodically or historically, there’s a great trend line there. I think we can never be complacent about it. I think it’s great that you’re asking about these questions because it suggests that maybe there’s more people who care about this than just us lone social science researchers.

OLIVER: I think a lot of people care about the general question of the reading decline, and they want more in-depth numbers.

IYENGAR: More information. Definitely.

OLIVER: It can look like we really know what’s going on, but what I find interesting about this is, well, up to a point.

IYENGAR: One of the things we are trying to get better at with these surveys is, not only the people who read or receive stuff, but put stuff out into the world. A lot of our work on creating art, we’re now getting to things like doing podcasts, what all kinds of formats people engage with. There’s definitely some connection there. We, in fact, ask about writing creatively, and we get those numbers too. That, I think, has declined a little bit, but not so steeply.

OLIVER: You don’t think people are switching from consuming art to creating it?

IYENGAR: That’s what I want to see, actually. I want to see more of that. I think it’s going to take some time as well as the actual changes to the survey.

OLIVER: Now, I can’t let you go without bringing up the fact that, as a child you had a party trick, which was that you could date any book by its smell.

IYENGAR: Apparently, yes. I was living in rural Mississippi at the time. I don’t know why, but we just moved there. I guess maybe I was trying to be popular. I don’t think this is something I’d recommend. We were in the school libraries and stuff. I don’t know how this started, but I’d find these books, and they’re often very old, sitting on the shelves. Who would have read them in a school library? I’d pick them up and I would smell the pages. In doing so, for some reason, I could usually date the book.

OLIVER: Within?

IYENGAR: Within a couple of years, usually. I got pretty good at it. I didn’t do this 24 hours, but it was something I would do as a party trick in the school. I remember once, a teacher getting irritated because I was making people laugh or doing something like that. Then that teacher got interested and was giving me books to look at and read and smell.

OLIVER: The teacher had you sniff the books?

IYENGAR: Yes. I remember this. I would always get it within a year or two, and I think this is something to do with a vintage or a fine wine. You could just tell what year it was brewed or whatever. It was something like that. It’s long lost; I’ve moved to reading them or—

OLIVER: You’ve moved to reading the books.

IYENGAR: —talking about them or doing studies about them rather than actually sniffing book glue.

OLIVER: Very good. Sunil Iyengar, this was excellent. Thank you very much.

IYENGAR: I hope so. Thanks a lot. Great to be here.

Well, it’s a fantastic and incredible honor to be here. I’ve gotten to know a number of you in the last hours, which has been a great pleasure. I can report from the hinterlands that you all are changing the world, and no pressure. I’m very conscious that what you’re doing now and what you’re going to be doing in the near future will promote the cause of liberty, not as a slogan but as lived reality with concrete ideas that instantiate it, and that makes it amazing to be on this floor. This set of remarks is an ongoing project that’s a constellation of three things, which I guess are in the form of stories which converge. The first is, as an advanced teenager early in college, I read The Road to Serfdom.

As happened for so many people, including I know some in this room, it almost literally knocked my socks off. The sunlight that burst in the sky as I read chapter after another was incandescent. The framework that Hayek introduced stayed with yours truly, as with so many others, and it is ingrained in the DNA of all those, I think, who read the book. The second tale is a somewhat older, young law professor at the University of Chicago, Hayek’s own institution for a long time. I was surrounded by people who were 7’4”. Gary Becker must have been 7’4”. George Stigler, who, according to the height charts, was 6’5”, was actually, in person, 9’8”.

They were all giants, and they were, in some sense, Hayek’s younger siblings. What they added, among other things, was something that Hayek didn’t stress and maybe didn’t believe, which is a commitment to human rationality. For the Chicagoans, the loosely Hayekian giant colleagues, it seemed, to tiny me, there was a theory of human behavior, which was imperfectly matched to the literature that I had studied in college, literal literature like novels and poems, and also imperfectly matched to their own tennis games, which seems to suggest overconfidence, unrealistic optimism, a use of heuristics about what shots to hit when.

This is a polite way of saying they weren’t very rational tennis players, and some of them were getting divorced, meaning they’d made choices that had gone wrong. This led me, Hayekian-inspired though I was, to think that departures from perfect rationality were all around and causing problems. This needed to be analyzed. I started to write a paper on departures from perfect rationality and pre-commitment strategies and the endogeneity of some preferences.

The paper was well underway in the midst of a squash game with a law and economics follower of Hayek. I told him, actually, after the game that I was working on this paper, and he said, “This is a terrible paper. You shouldn’t publish it. It’s doomed to failure. The central idea is all wrong, and not just wrong but destructive, but there’s someone else working on this topic who’s actually published a paper almost as bad as yours.”

“His name is Richard Thaler.” I thought it was T-H-A-Y-L-O-R, and it took me a long time to find the paper. When I did, it was toward a positive theory of consumer choice. The paper helped get Thaler the Nobel Prize many years later. At the time, it was an obscure paper. I got a sunburst that wasn’t as violent as the Hayek sunburst, meaning it wasn’t as explosive, but it was, nonetheless, a sunburst in which Thaler cataloged departures from human rationality in a way that overlapped greatly with my primitive effort to systematize what I had learned from Shakespeare and James Joyce and Samuel Beckett as departures from perfect rationality.

There’s a conflict here, isn’t there, between the devil and the angel, or the two angels? Hayek and Thaler. What I’ve been trying to work out in the last three or four years is whether it’s possible to create a Hayekian behavioral economics, that is, a form of economics that is self-consciously and proudly and insistently Hayekian, but is alert to departures from perfect rationality. What would that look like?

This is going to be a cut at it and to presage the heart of it. If human beings often lack information as individuals, and if that replicates itself in the price system as it sometimes does, if human beings show unrealistic optimism like tennis players and occasional investors, if people show present bias in the sense that the short-term matters a lot, the long-term not so much, and if there’s a catalog of these things complementing informational deficits, then what are regulators to do?

We have a problem. The solution is not to ask planners just to figure it out. The solution must take another Hayekian form. I’m not going to spoil the surprise quite yet. I’m going to say that’s the enterprise. Let’s just notice the stakes are large in the sense that there are people who die too young, get sick when they ought not to. 125,000 Americans die prematurely because they don’t take prescription drugs, which if they took them, they wouldn’t die prematurely.

We have tens of thousands of deaths on the highway. Those aren’t as easily preventable as those that come from a failure to take medicines that are prescribed. Surely, they are preventable. What are we going to do about this kind of thing? This is a project for freedom lovers everywhere to take super seriously. Okay. You all are sufficiently familiar with Hayek to know that his description of the price system as a marvel meant to shock the reader out of complacency for taking the price system for granted was based on the sense that the market is a system of telecommunications that no one sees until Hayek quite as such.

What the price system does better, even than the great Habermas thought democracy did, is to register collective intelligence. It’s the day after the 34th, I think, anniversary of Hayek’s death. It’s just a few days after Habermas’s literal death, and these are the great theorists of collective intelligence, with Hayek emphasizing the price system, Habermas emphasizing democracy, and I think Hayek wins that argument.

In light of modern behavioral findings, we might object that the price system is not always so marvelous. If consumers show limited attention, they might be subject to manipulation on the part of those who hide features of a transaction. If people are subject to unrealistic optimism, it might be that savvy sellers can exploit that and create an equilibrium that embeds unrealistic optimism, and then the price system is going to be giving the wrong messages.

We might agree with Hayek’s arguments about planning and prices while thinking that certain forms of regulation, actually pretty aggressive ones, aren’t out of bounds. Hayek said, and you can feel the irritation on the page, “Probably nothing has done as much harm to the liberal cause as the wooden insistence of some liberals on some rough rules of thumb above all the principle of laissez-faire.”

Hayek didn’t choose his words carelessly, and it’s worth pausing over that claim. Probably nothing has done as much harm to the liberal cause, or more specifically, these words, to prohibit the use of certain poisonous substances, notice the P word, “prohibit,” or to require special precautions in their use, to limit working hours, or to require certain sanitary arrangements, is fully compatible with the preservation of competition.

The only question here is whether, in the particular instance, the advantages gained are greater than the social costs that they impose. Those of you who know some of my scribbling will know that my affective reaction on reading that sentence was a bit over the top. While dancing in response to a sentence like that is a little excessive, something in my legs really wanted to dance in reading those sentences.

Maybe, in view of the endorsement of occupational safety and health restrictions, limitations on working hours, maybe a mandatory seatbelt law, a ban on trans fats, amply justified by benefit-cost analysis, the benefits crush the costs, would be unobjectionable. Maybe Hayek’s arguments create a bridge towards cigarette taxes or taxes on sugar-sweetened beverages. I say “maybe” because the answer isn’t entirely clear about the right Hayekian approach to those issues.

Focused on coercion, Hayek didn’t much love paternalism. While he didn’t discuss paternalism as such, his most stunning words on liberty stand against it. This is an un-Hayekian sentence written by the man himself. Coercion is evil precisely because it eliminates an individual as a thinking and valuing person, and makes him a bare tool in the achievement of the ends of another.

Pause over that and maybe engrave it somewhere. What makes it not a characteristically Hayekian sentence is it’s strikingly Kantian. It’s not welfarist. Nothing there about costs and benefits. It doesn’t speak about welfare at all. Notice the use of the word “evil” and the objection to treating people as means rather than as ends. Indeed, Hayek seemed to embrace something like a Kantian non-welfarist foundation for freedom in his less famous passages.

In the introduction to The Constitution of Liberty, he wrote, “Some readers will perhaps be disturbed by the impression that I do not take the value of indisputable liberty as an indisputable ethical presupposition, and that in trying to demonstrate its value, I am possibly making the argument in its support a matter of expediency. That would be a misunderstanding,” Hayek says, though he never spelled this out.

At pivotal points, Hayek’s argument was epistemic, not ethical. Thus, he urged the argument for liberty, its chief basis, his words, is the awareness of our irremediable ignorance. If it appears that the market mechanism leads to an effectuation and utilization of more knowledge, that is the chief foundation of the case for economic liberty. Notice the dual-mindedness of this. There’s the Kantian beating heart, but the epistemic, maybe conceptual framework.

“The case for individual freedom,” Hayek writes in The Constitution of Liberty, “rests chiefly on the recognition of the inevitable ignorance of all of us, considering a great many of the factors on which the achievement of our welfare depends. Thus,” he says, in his least Kantian moment, “if there were omniscient men, if we could know not only all that affects the attainment of our present wishes, but also our future wants and desires, there would be little case for liberty.”

I think he couldn’t have meant that entirely. It contradicts the early sentences, but he said it. Hayek, there, roots his claim for liberty, his most cherished ideal, in the absence of omniscient men. “If there were such men, we would be able to offer little case for liberty.” I don’t agree with that for many reasons, and I doubt that Hayek believed it either. Let’s not let that detain us here.

Emphasizing human fallibility and the propensity to blunder, some people have questioned the marvelousness of the price system. I’m thinking of Akerlof and Shiller, Nobel Prize winners both, who write on this topic. Other people have been asking whether new institutional arrangements involving mandates and bans have fresh justification. This seems anti-Hayek on the Kantian and the welfarist ground, but the motivation for this enthusiasm for paternalism isn’t obscure.

If we know that people’s choices lead them to poverty and ill health, why should we insist on freedom of choice? Ought not institutional design to take behavioral biases into account? Hayek, as always, was a bit of the head of the game here. In speaking of the dispersed nature of knowledge, he also defended nudges before the word came to have its current meaning. “Where,” wrote Hayek, “most individuals do not even know that there is useful knowledge available and worth paying for, it will often be an advantageous investment for the community to bear the cost of spreading the knowledge.”

We all have an interest in our fellow citizens being put in the position to choose wisely. If some have not yet awakened to the possibilities which developments offer, a comparatively small outlay may often be sufficient to induce the individuals to take advantage of new opportunities, and then to advance further on their own initiative. What Hayek is defending here is educative nudges in the form of information provision that overcomes both a lack of knowledge and a behavioral market failure in the form of identifiable behavioral biases by which people’s choices make their lives go less well by their own lights.

“When this is so,” Hayek urges, “a corrective response might be put on the table, at least in the form of information that helps people not to blunder.” We have to emphasize the word “might.” Planners might claim to find identifiable behavioral bias when there is no such thing, the knowledge problem on the part of the choice architect. Planners might have bad incentives of their own. You’ve heard about that, yes? They might be subject to the influence of well-organized private groups.

Can I tell you a story? In my White House experience, there was a big movement to try to get labeling of genetically modified organisms. This was an effort to nudge consumers away from same. The problem, of course, was the effort came from guess who? The organic industry, which was trying to get a competitive advantage in the face of repeated disclaimers from people whose business it is to know these things, suggesting that genetically modified organisms don’t cause health or environmental problems.

It was purely an interest group-driven effort at a nudge. The Obama administration resisted the effort. Trump won, went forward. Trump forward didn’t go forward because he loved it. He went forward because Vermont had endorsed the labeling requirement. It looked like we were going to get a national mess of labeling requirements. The labeled industry went to Washington and said, “Please get us a label that won’t cause inconsistency or mass terror.” If you look at the label that emerged, it’s the most cheerful, sunny label. It nudges exactly no one not to go for genetically modified food.

Here’s the large question. Suppose we’re in search of a behavioral economics for policy purposes that is taking on board everything Hayek said about liberty and knowledge. If we’re in search of that, we would firmly reject the idea that planners should be content to identify behavioral biases and declare victory. Instead, we would want to engage in something like comparative analysis. [coughs] How costly would the errors that the planners introduced be compared to the errors that would be introduced by an alternative approach?

A Hayekian initiative. [coughs] I read a book about public speaking that said, if you’re getting into really complicated territory, you should pretend to cough [laughter] and drink some water so you can assemble your thoughts. Is it working? You don’t know yet. Okay, so the question is whether we can adopt an approach, which is Hayek in his Kantian phase and Hayek in his welfarist phase, that tries to reduce the knowledge problem faced both by the planners and by the people whom the planners are trying to help by asking a single question, which is, what do individual choosers do under epistemically favorable conditions?

That’s going to be the rabbit out of this hat where the planner does not ask an unanchored cost-benefit question and does not ask, “Do I have a superior assessment about what consumers should do than consumers?” Does not even ask, “Do consumers suffer from a behavioral bias?” but asks, “What do actual choosers do under conditions that are epistemically favorable?”

What does epistemically favorable mean? It means that people know things, and it means that people don’t suffer from an identifiable bias, such as, for example, availability bias, which can lead people to exaggerate the likelihood that a salient risk is going to come to fruition, and to downplay the likelihood that a risk that isn’t salient will come to fruition. What we’re trying to do is isolate, not by imagination, but by data, what people actually do under epistemically favorable conditions.

The beauty of this is that we are in the midst of maybe year six of a Hayekian research project that tries to answer exactly these questions, and to develop policy on the basis of what we know. Year six is early, but here are some examples of the kinds of questions that a Hayekian behavioral economics would put front and center. First, what do informed choosers choose as opposed to people who don’t know a thing about the context in which they are making choices? Do you have any data on that?

Second, what do consistent choosers choose, unaffected by clearly irrelevant factors or frames? Suppose it’s the case that if you frame a question saying 90% of people are fine after a certain operation, do you want to get the operation? People say, “Absolutely.” Then, if you ask a similar population, 10% of people aren’t fine after the relevant period. Do you want to have the operation? They give a different answer. That’s a problem.

Maybe there’s a class of people who give a consistent answer who aren’t affected by the frame, and they would have at least a degree of authority that the inconsistent choosers lack. What do active choosers choose? In some cases, people end up in a situation just because there’s a default with which they stick. It might be something that says they’ll give $10 a month to someone whom they barely know because they didn’t unclick a box that was clicked.

I think that just happened to me last week, but I was too busy to pay attention. I’ll tell you when my credit card bills arrive. If active choosers who actually make choices rather than passively accepting things make a certain stream of choices, then they have authority. We know what the relevant population thinks. In circumstances in which people’s view screen is broad and they’re seeing a full set of ingredients of a transaction, what do they select as opposed to people whose view screen is small and they’re focusing on three of seven factors?

We might be able to know from context whether people’s viewpoint is broad or narrow by making salient seven things where a manipulative, let’s say, seller is making salient three things. That might be a profit-maximizing strategy. Suppose we know or can work to know what people choose when they’re free of present bias and unrealistic optimism. Maybe people who lack present bias and don’t suffer from unrealistic optimism think, “I don’t want to be in that darn savings plan.” Maybe the reason they don’t want to be in the darn savings plan is, A, they have plenty of savings.

They have no need for it, or B, that they have an urgent current need. It’s not a matter of unrealistic optimism. It’s epistemically sensible. They know what their situation is. Those are, let’s say, Hayekian subjects of the best kind, and they have authority. The claim is that policymakers ought, in cases in which there’s reason to think human distress is occurring and a product of informational deficits or behavioral biases, to ask these questions. If public institutions can learn what consistent and informed and active choosers, less influenced by present bias or limited attention, choose, they might have real guidance.

The submission is these questions can be answered empirically. That’s not just a promissory note. We know some things about this. There’s data suggesting, for example, that if consumers are flooded with information about the fuel economy of vehicles, they don’t make different choices from the choices they make if they’re not so flooded, which is strong evidence that informed choices are choosing the vehicle mix we basically observed, and that the fact that we observe less in the way of electric cars and less in the way of hybrid sales than a planner’s cost-benefit analysis would seem to suggest isn’t a product of a lack of information because information provision doesn’t alter choices.

In other words, that’s not a problem. We have data consistent with that. Should employers offer opt-in savings plan or opt-out savings plan? We have data suggesting that many employees are affected by the frame. If you’re automatically enrolled, people stick. If they have to sign up, at least in the first few years, they don’t sign up. We also have data suggesting lots of people aren’t affected by the frame.

The people who aren’t affected by the frame typically end up in a savings program. Not everybody. We don’t want a mandate here, in my view. We want freedom to be preserved. Opt-out has advantages over opt-in insofar as we know that people who are consistent generally opt-in. That’s the right default. If the consistent choosers, there’s a qualification, aren’t different from the inconsistent ones, except that they’re affected by the frame, we have a reason to think that the choices of the consistent choosers are the correct ones.

If the consistent choosers are a different population from the inconsistent choosers, then this just isn’t going to work, so it has to be randomized. Suppose we know that consumers make an active choice to enroll in certain insurance programs when those programs are designed so as to promote active choosing. If so, we have some reason to suppose that a default in favor of the insurance program makes some sense because we have some reason to think that if people aren’t choosing to enroll, it’s not because they don’t want to, but because of inertia or inattention.

That’s the central point. If, under circumstances of active choosing, people go for it, then we have reason to think that’s the thing that people want. If they are actively engaged and if they don’t enroll, it’s not because they don’t want to. It’s just because they are inattentive or suffering from inertia. Okay. We know that from experiments about energy-saving appliances that if you not only inform people but you make it highly salient to them, what the savings are? We don’t observe a change in choices, not much of one.

If the question is what kind of light bulb to get, one that is environmentally a little better or one that is environmentally a little worse, and let’s stipulate what’s true, then the environmentally better one is less expensive in the long run, but it’s not as bright. People choose when informed of the economic benefit of the environmentally better one, they choose the one that’s brighter.

Even when they are given clear information that should overcome present bias or inattention, still they get the brighter one. That’s highly suggestive that a plan or response to the supposed behavioral bias would be ill-advised, because we know from the data that if you counteract the supposed behavioral bias, people do the same thing, which is strongly suggestive that there is no behavioral bias. People just want the product that has the brighter light.

Now, you might be thinking about externalities. That’s completely fair. Hayekian style and externality might justify the environmentally preferred light, but it would be because of the externality, not because people are making the wrong choice. Okay, there might be heterogeneity in the relevant population. If there weren’t, that would be shocking, making it challenging to generalize from what part of the population does.

At least we now have a framework by which to embark on empirical tests and on policy initiatives in which planners are building not from their own convictions, but from the observed behavior of informed and behaviorally unbiased choosers. The reason is that planners are building on the choices of the right choosers. If we look at fuel economy and energy efficiency, which is we can see what a Hayekian research program and what a Hayekian set of policy initiatives would look like, and what is a thing of, to me, surpassing beauty, is both the Biden administration and the Trump administration have really earned their keep on these issues.

The technical people have done splendid work. We have a lot of research on fuel economy choices and on whether people are suffering from a behavioral bias, inattention, maybe a form of myopic loss aversion when they buy cars that aren’t fuel efficient and that aren’t electric. There is data consistent with the idea that people don’t perfectly internalize the economic consequences of a low fuel efficiency car. The data is suggestive that people aren’t taking on board 100% of the economic cost.

That’s the data that the Biden administration has emphasized. The Biden administration was impressively cautious about the data, knowing that the informed consumers don’t make radically different choices from the uninformed ones. Knowing that once you try to counteract the behavioral bias by throwing gas prices in people’s faces, as the market frequently does, people start buying different cars, which is testimony to the less-than-very bounded rationality, meaning quite excellent rationality of consumers in this domain.

The Biden administration noticed that when there is a correction of a miles-per-gallon error so that the overstated fuel efficiency of a set of cars is fixed, consumers don’t respond a whole lot, which is consistent with the people are making mistake choices. The Trump administration has some plausible responses to that evidence and has been very cautious about accepting the claim that consumers are making systematically wrong choices, by reference to data of the sort I’ve been describing, while noticing and emphasizing that the jury is still out on this question.

Okay, time to summarize and wrap up. The price system, even with the behavioral findings, remains an extraordinary system of telecommunications. In some areas, including occupational safety, food, as Secretary Kennedy is emphasizing, and other areas, the marvel is not unerring. If people are suffering from limited attention, a problem of self-control, unrealistic optimism, or a focus on the short-term, prices might not capture important factors.

If the consequence is a serious welfare loss for people, Hayek’s ears prick up in his welfarist incarnation, there’s an argument for some kind of public response. If people are running risks of mortality or otherwise ruining their lives, then the Hayekian machinery is, in play, consistent with the premises I’m trying to endorse. In the end, as Hayek appeared to know, occupational safety and health regulation, which he was okay with, has to be justified on behavioral grounds.

They forbid workers from facing certain risks. That’s what they do. On Hayekian grounds, and according to Hayekian behavioral economics, the best response might well be for public institutions to provide information in a way that is attuned to rather than unaware of behavioral biases. It’s true, yes, that the best approach might be to do nothing on the theory that the cure might be worse than the disease.

Remember, if you would, the questions that I wanted to isolate as answerable in principle questions and appealing in theory questions to orient a research program and a program of behaviorally driven policy that is responsive to the limits of planners and that founds itself on what people in epistemically favorable conditions actually do. In the first instance and probably in the last, behaviorally informed policy has to be based not on the values of social planners, but on learning from the actual choices of informed and unbiased choosers. We’re increasingly able to identify what those choices are.

If so, we’re on a good and starting to be paved road, which is toward identifying appropriate interventions, probably focusing on freedom-preserving nudges, but conceivably involving economic incentives, too. “It might be extravagant to claim,” he concedes, “that those interventions defended by references to people’s choices, informationally full enough and free from behavioral biases, it might be extravagant to claim that those interventions are Hayekian.” I’m going to deny that and say that it’s not extravagant to insist that they are fully in Hayek’s Kantian as well as Benthamite incarnations and respectful of his fundamental concerns. Thanks.

Only a philosopher could argue that despite the fact he was the only one to condemn slavery, Adam Smith is not a great philosopher, as did recently.1 This is why Rebecca is wrong to dismiss The Theory of Moral Sentiments — it contains some of Smith’s most powerful writing about slavery. In a discussion about “savages”, Smith says that savages experience “extremities of hunger” and are “habituated … to every sort of distress” and is unable to give in to the passions excited by this distress. Discipline is necessary for survival. This is part of the difference between the civilized and the savage: “Before we can feel much for others, we must in some measure be at ease ourselves.”

This self-control is seen most severely in preparation for death. “Every savage is said to prepare himself from his earliest youth for this dreadful end.” Smith recounts an idea that savages learn a song of death, which is to be sung after a savage has been captured and tortured—it is full of insults to the tormentors. In his account of this “contempt of death”, Smith makes a powerful statement about slaves and their masters.

The same contempt of death and torture prevails among all other savage nations. There is not a negro from the coast of Africa who does not, in this respect, possess a degree of magnanimity which the soul of his sordid master is too often scarce capable of conceiving. Fortune never exerted more cruelly her empire over mankind, than when she subjected those nations of heroes to the refuse of the jails of Europe, to wretches who possess the virtues neither of the countries which they come from, nor of those which they go to, and whose levity, brutality, and baseness, so justly expose them to the contempt of the vanquished. (V.ii)

Here we see Smith showing the supposedly civilized slave masters as lacking sympathy and the African slaves—who were so freely traded in Smith’s day—as the ones full of magnanimity. Smith is praising the slaves for their self-command, a virtue he prizes above all others. He goes on to contrast this with the emotional nature of modern European civilization, which is much more weepy than ancient Rome.

Shortly afterwards, in one of the most unforgettable sections of The Theory of Moral Sentiments, Smith argues that it is custom which accommodates men to such evils as slavery. In ancient Greece, infanticide was allowed. Smith says this must have begun in barbaric times and survived into the era of civilization because

Uninterrupted custom had by this time so thoroughly authorized the practice, that not only the loose maxims of the world tolerated this barbarous prerogative, but even the doctrine of philosophers, which ought to have been more just and accurate, was led away by the established custom… (V.ii)

It is characteristic of Smith to make these barbed remarks about philosophers. In The Wealth of Nations, he quotes Cicero to the effect that no idea is so stupid that some philosopher will not entertain it. It is sometimes said that The Wealth of Nations condemns slavery on economic grounds, but makes no mention of injustice. This is not quite right.

The pride of man makes him love to domineer, and nothing mortifies him so much as to be obliged to condescend to persuade his inferiors. Wherever the law allows it, and the nature of the work can afford it, therefore, he will generally prefer the service of slaves to that of freemen. The planting of sugar and tobacco can afford the expense of slave cultivation. The raising of corn, it seems, in the present times, cannot. (III.ii.10)

The pride of man makes him love to domineer… perhaps this is not an outright condemnation of slavery on the grounds of abstract justice, but it is a clear statement of the moral corruption which slavery involves. Dan Klein documents how Smith’s writing in Theory of Moral Sentiments was the inspiration to later abolitionists. (He also notes that Smith was not the only Glasgow philosopher to condemn slavery.) Smith’s paragraph about the magnanimity of the Africans inspired the 1764 pamphlet An Essay In Vindication Of The Continental Colonies Of America, from A Censure of Mr. Adam Smith, in his Theory of Moral Sentiments, later quoted by Clarkson. (Wilberforce quoted Smith, too.)2

Smith may not have had a great influence on the discipline of philosophy, and Theory is more of a work of observation than philosophy, but he had a not inconsiderable part in the argument for abolition. Philosophers are fussy about the question of how arguments are made, and rightly so, but perhaps Smith is a standing caution to their discipline. He is dismissive in Book V of The Wealth of Nations about the inadequate teaching of philosophy in the English universities of his day, writing “if subtleties and sophisms composed the greater part of the metaphysics or pneumatics of the schools, they composed the whole of this cobweb science of ontology, which was likewise sometimes called metaphysics.” Maybe he was up to something that philosophy could learn from…

Smith argued against the poor philosophy professors of the time, not the discipline itself, but his criticism is important. Too much inwardness can make philosophy a self-defeating subject. Smith preferred to work from a “foundation in nature”, criticizing thinkers like Wollstonecraft and Mandeville for not accounting properly for people’s full range of feelings. The morally formative nature of social life is his subject. In his understanding of economics as being institutional, social, depending on “rivalship and emulation”, he was not a scientific economist but a humanistic one.

The Wealth of Nations is not merely a work of the division of labour: it is about how a nation organises itself, the question of spontaneous order, the rights and duties of the sovereign, standing armies, national defence, churches, public education, and the inner life of the citizens. Smith is no mere technician, no mere describer of supply and demand equilibria: he sees how status, incentives, desires, and rivalries organise the whole of a society as if guided by an invisible hand. That metaphor is perhaps his single most important idea. Rather than being creatures led by gods, souls, or daimons, we are “led by an invisible hand to promote an end which was no part of his intention.” The invisible hand, of course, originates in The Theory of Moral Sentiments. Smith’s metaphor explains more about how life works than whole books of philosophy.

Tyler Cowen has argued that “Smith’s account of how a modern commercial society can hold together and overcome some of the most basic deficiencies of human nature” in Book V is what makes The Wealth of Nations cohere as a treatise of society, competing with Plato. The invisible hand is the answer to many old philosophical problems. Tyler writes:

We now can see a new way that Wealth of Nations and Smith’s earlier Theory of Moral Sentiments hold together. Both are concerned with individuals being excessively narrow, short-sighted, and obsessed with local information at the expense of global information. Smith, by putting his alienation discussion into his treatment of education, showed he understood that all of life and all of work is an education of some sort, just as school and religion are. The real social problem is about the fundamental shaping of individual character.3

The mere existence of The Wealth of Nations ought to give philosophers pause. Why is it after so many centuries, treatises, arguments, and disputes that it took a man no longer considered to be a “proper” philosopher—despite the title of his chair at Glasgow university—to explain something so fundamental about what makes society work? Rather than arguing that Smith is no major part of their profession, I would think the philosophers might want to learn from his accomplishments.

Like I said, I don’t want to have to pick philosopher or economist, but if I have to choose, I’ll defend Smith as a philosopher. He taught us to see the way society isn’t just divided by division of labour, but uses that as a means of co-operation. Philosophy and economics might not be such distinct disciplines as they appear…

One of things I enjoy the most about my job at Mercatus is moderating in-house philosophical debates. I organise these occasional debates as special sessions of the Philosophy Working Group I run every Wednesday.

Today, in honour of the recent 250th birthday of Adam Smith’s The Wealth of Nations, I moderated a debate on the topic ‘Adam Smith: Economist or Philosopher?’. We also ate a delicious cake featuring a picture of a (not really) invisible hand.

Since I’ve been meaning to write about Adam Smith this month, here follows a non-verbatim version of my opening remarks. You may notice a little hyperbole.

Adam Smith: Economist or Philosopher?

I’m going to break with the tradition of these debates, and give a quick answer to the question at hand(!), myself. I’m doing this partly because I think it’s funny that, as a philosopher, I’m firmly on the side of ‘Adam Smith Economist’.

But I also wanted to kick off the debate in this way because ‘Adam Smith Economist’ is clearly the obvious and only correct answer. So I thought I should take one for the team by admitting my norminess, to let everyone else make clever but wrong arguments.

My main reason for being on the ‘Adam Smith Economist’ side comes down to the stark difference in quality between The Wealth of Nations (WoN), which is Adam Smith’s great economics book, and The Theory of Moral Sentiments (TMS), his supposedly great philosophy book.

WoN is one of the greatest economics books of all time. Probably the greatest. It covers everything from land to labour to money to metals to interest to imports to teaching to taxes. It does all this coherently, and mostly convincingly — even on some things that weren’t formally worked out until over a hundred years later!

WoN is also incredibly readable. I started rereading the last section of it at 3am this morning, and I couldn’t stop. Why did I start at 3am, you wonder? Because I’d been reading The Theory of Moral Sentiments, and I needed a break! I needed to read something good! This is because TMS is really a very annoying, quite subpar, torturously written philosophy book.

Stylistically, TMS reminds me of the worst of Kierkegaard. This is not a good thing! If you’re ever tempted to read his Works of Love, then don’t say I didn’t warn you. All of these overly complicated little anecdotes with their clever wordy sub-headings.

If you’ll allow me an English person digression, they remind me a little of the mid-20th-century English children’s novelist Enid Blyton. Blyton breaks her stories up with silly stylised chapter titles, like ‘After a lovely day rambling in the sunny hills, the five children return home to the caravan and a nasty surprise!’.

This is exactly what Smith does in TMS. And it points up the way this book is full of bad narrative continental-philosophy-type storybook rambling, rather than the brilliant ice-cool analytic-philosophy-type argumentation I love!

Then, substance-wise, 90 per cent of TMS is Smith finding yet another way to say ‘if you feel a bit bad when something bad happens to someone else, then that’s normal, so it’s good!’.

And of course almost all of it is just mediocre takes on David Hume. I should admit that I don’t like Hume much, for his writing or for his positions. But he and many of his arguments — when you can work out what they’re about! — are undeniably great. It’s thanks to Hume that I don’t spend every minute of my waking life lost in a morass of skepticism. Only about half of them, which isn’t bad.

All that said, the one thing we philosophers should be extremely grateful to Smith for is that he’s pretty much the only philosopher before about 1900 who condemned slavery. Which is a seriously dreadful reflection of our discipline — more dreadful than I have time to discuss today.

Indeed, by far Smith’s best philosophical position, on many grounds, is his radical egalitarianism about the universality of the capacity for judgement. Smith believes that whether you’re rich or poor, and no matter which country you’re from, you have the capacity for great judgment, given the right education and conditions. He even thinks this about English people! Even though he was perhaps the original Scottish hater of the English...

Smith was a proper egalitarian, relative to his time, and much more generally. And I will always love him for that.

However, it’s worth noting that the relevant arguments he makes — arguments about the badness of slavery, and about the shared human capacity for judgement — are mostly much more explicitly and clearly made in WoN and in the Lectures on Jurisprudence, than in TMS!

This reminds us that the supposed WoN/TMS split is vastly overstated. The two books overlap substantively, and they both contain a lot of philosophy. The difference of course, however, is that WoN contains a lot of economics. A lot of really, really great economics! And it helps, again, that it’s also such a good read — no wonder people find it hard to believe they’re by the same guy! It’s one of the best-written books of any academic discipline.

If you were only going to read one economics book ever, then surely it would be WoN. And my guess is that it will take a long, long time for this to change.

Adam Smith is one of the great economists of all time. Likely the greatest. So far, and my bet is, for way into the future. Whereas, as a philosopher, he doesn’t make it into the top 40. Or 50. Maybe even 100.

You’ve got to hand(!) it to him, however, one out of the two is pretty great.

Have you seen the video of Trump with Seinfeld music in the background? It works pretty well. He’s a bit more like George’s father, Frank Costanza, than Jerry, but his little riff about paper straws really is like a Seinfeld bit.

Trump really does have the rhetoric of a stand-up comedian. (Although this whole idea means we really are in Bizarro land.) You might not find him funny yourself, but he speaks from an alternative reality where the things he says are acceptable because some people think he is funny. A lot of his rhetorical credibility comes from the tones, patterns, and habits of a stand-up comedian.

What once was said for laughs, Trump now says for votes.

Trump’s Comedic Rhetoric

Trump ought to be a stand-up comedian’s dream. Telling minor anecdotes about his visit to a factory can get a good laugh. Even Pete Hegseth does a Trump impression. But you can rarely do better than to impersonate Trump using his own words. Often, rather than watching a comedian impersonate Trump, it is funnier simply to watch him, such as in the widely shared video that shows him and Obama announcing the death of terrorists.

Obama is serious and stately. Trump is ridiculous. But the internet chuckles.

Trump’s comedic-rhetoric is why his tweets get so much attention.

I would like to wish everyone, including all haters and losers (of which, sadly, there are many) a truly happy and enjoyable Memorial Day!

Trump’s ability to react to events can make him harmlessly funny, such as when his podium wobbles and he says, “It’s drifting left, like too many other things.” But very often, the humorous quality of Trump’s affect is that it is both offensive and compelling, vulgar and amusing. This is why he is so polarizing. While some people are outraged, others are making YouTube compilations of his funny moments.

Trump’s style comes out of the sort of offensive humour we associate with the comics of the 1960s and 1970s. Of those comedians, the one Trump most resembles is Don Rickles, who was famous for his Hey Dummy! style of comedy.

Three resemblances: monologue, offense, control

The first resemblance is the rambling monologue. It was Don Rickles’ standard approach when appearing on a talk show to dominate the whole proceeding. The host—be they Johnny Carson or Dick Cavett or Frank Sinatra or Joan Rivers—hardly got a word in, while Rickles proceeded on his unstoppable, not-quite-incoherent medley. Incoherent, but funny. Just like Trump, Rickles would bounce from one thing to the next. It seems like a bunch of non-sequiturs but it all made sense as a matter of mood and temperament. Even the moment when Jack Nicholson called Rickles a “very dignified maniac” has a Trumpian tone to it—the very stable genius.

The second resemblance is offense. Rickles was funny by being offensive. His jokes were largely based on race, class, and sex. It is no coincidence that these are the core concerns of Trump’s opponents. He made the calculation that forceful humour, of the sort no longer deemed acceptable, would be a more effective response to their arguments. It was standard for Rickles, even late in the twentieth century, to not only make racially insensitive jokes, but to perform racist stereotypes. He had a standard range of jokes about his wife, things like: “they are always in heat when you are not ready”. Rickles was also vulgar. When Carson corrected his pronunciation of someone’s name, in front of a silent audience, Rickles drawled: “Wonderful. Look at how the crowd got excited. Look at that: two guys in that row dropped their pants and fired a rocket.” Trump does the same. He’s happy to stereotype.

The third resemblance is their ability to retain control of the monologue. Trump never quite seems to be answering questions. He treats reporters almost like his audience, giving him set-ups for his act. When a reporter began a question with “As you know…” and started talking about his son, Trump jumped in: “How would I know? How would I know that?” Then he turned to some people on the other side of the room, gestured to the reporter and said, “I don’t even know who this guy is. That’s all right.”

Rickles was a master of this domination. If a talk show host asked him a personal detail—about his war service, his wife’s name—he snapped back: “What are you, a detective!?” Given half-a-second, he would jump in and say, “Admit it. It’s over.” During anecdotes, any mention of someone’s wife or family would prompt an aside, about how he has two wonderful sons, plus one in the Philippines, or about how on his wedding night, his wife went: “NO!” Whatever it took to prolong the monologue.

Compare Trump responding to reporters by saying “What a stupid question.” When a reporter from Yahoo! news asked Trump about comparative rates of vaccination between the USA and South Korea, Trump had someone read out the figures and then said, “Are you going to apologize Yahoo? That’s why you’re Yahoo. Nobody knows who you are, including me.” This is a pure Rickles play. As it was when a reporter mentioned Elizabeth Warren and Trump said, “Who, Pocahontas?” (referring to her claims to Native American ancestry.) When a reporter called back, “That’s very offensive,” Trump eyeballed her and said, “Oh really? I’m sorry about that.” It was reminiscent of Rickles, who, after he told a joke so offensive it made the audience groan, would snap, “For the money you people are paying that’s a funny joke!” When he was asked if he regretted calling Warren Pocahontas, Trump said, “I do regret calling her Pocahontas because it’s a tremendous insult to Pocahontas.” (All of those can be seen on this compilation.)

On the Charlie Rose show in 1992, in a rather dull interview about his comeback, Trump had a flash of the comedian “I love getting even with people.” Rose cracked up. “Slow up. You love getting even with people?” “Oh absolutely,” Trump replies, “You don’t believe in the eye-for-an-eye? You do—I know you well enough.” It’s a moment when Trump realizes his ability to play the guy across the table. The delivery wasn’t quite right, but it was a little touch of Rickles early in Trump’s career.

Bravado

For Rickles, though, it was all an act. The conceit was that Rickles doesn’t really mean it. (And, indeed, his friends attested that he was a lovely man off-stage, unrecognizable on-stage.) On the Dick Cavett show, the whole thing is obviously a game. Cavett introduces him as Dan Rickles. When he apologizes, Rickles says it doesn’t matter, “no-one sees this show.” Rickles got laughs by saying, in response to Cavett saying he didn’t picture Rickles as the sort of person who got laryngitis, “I really don’t care what you picture, Dick.” And so on. It’s a game of bravado.

Compare this Trump last year introducing J.D. Vance at a ceremony honoring US Navy sailors. Vance had recently said on live television “we borrow money from Chinese peasants to buy the things those Chinese peasants manufacture”, and Trump roasted him by saying:

We’re thrilled to be joined by a proud Marine Corps Vice President J.D. Vance… JD? Where is JD? What the hell happened to JD? He was just here. He must’ve gotten a call from … China

Listen to Trump’s delivery and you can hear the comic inflections more familiar to stand-up than the stump. It’s like Rickles making jokes about Sinatra. But the joke really matters now: it’s part of how Trump remains dominant.

It’s a roast, but he means it

Rickles’ never-ending medley was part of a larger culture of offensive humour. These comics often did their best work at a roast. In one sense, all of their work was a roast. They were good at picking on each other. “Dean, I say this from my heart, really. I’ve never liked you.” Bob Newhart, Rickles’ best friend, wasn’t much of a roaster. It wasn’t in his character. But for the Johnny Carson culture, almost all comedy was a form of roasting. When Rickles was the one hosting the Tonight Show, Lee Marvin Rickled Rickles, refusing to answer questions, turning to talk to the other guests, and undermining the whole attempt to interview him.

Trump took that game into politics and he made it serious. When he roasts a reporter it does sometimes raise a chuckle from the aides in the room (and many of the viewers at home, no doubt) but it is brutal. He is using patterns of comedic speech in a bullying manner. It’s Rickles’ style stripped of its intention to make the recipient smile.

In an appearance on Letterman late in life, Rickles spent the whole interview mocking Letterman. When Letterman uh’d or ah’d, Rickles imitated him and said, “That’s Alzheimers Dave!” If Letterman pointed his finger, Rickles snapped, “Why are you pointing at me!? I’m right here!”

Trump does this much more harshly to reporters all the time. “CNN should be ashamed of itself having you work for them.” If Rickles said that on a late-night show, it would be funny. When Trump says it to a CNN reporter, the humor has a much sharper edge. It makes the punch line a lot punchier.

So many of Trump’s most famous lines could have been spoken by Rickles, such as when it was put to Trump that he called women “fat pigs and dogs,” and he deadpanned, “Only Rosie O’Donnell.”

Yesterday’s man?

Some people believe that this sort of comedy doesn’t happen now. A discussion between John Stamos and Bill Maher about Rickles earlier this year focussed on how Rickles couldn’t make those sorts of jokes today. Maher saw him live in 1995 and thought he was already out of kilter with the times. By then, a new generation of comedians had taken over.

After Rickles, Robin Williams’s endless monologues involved impersonations of gay people, routine use of foreign accents, and elaborate routines about sex all seemed friendlier than Rickles, but were a continuation of the old tradition.

Williams really exemplifies the Rickles manner—Hispanic voices, thanking a woman for “getting into that dress”, sign language about sexual matters, and an Irish joke, all in two minutes. It’s the sort of comedy that Ricky Gervais uses, most famously when he hosts the Golden Globes.1

But the bluntness, the blatant offensiveness, the racism and the sexism, that does seem to have died away. It is notable that Williams’s 2009 tour was much more political, and had far fewer impersonations of minorities.

Until it was reborn in Donald Trump’s rhetorical style.

Boo this.

When Jeb Bush tried to get a word in during a debate, Trump said, “More energy tonight, I like that.” That is classic Rickles, who was forever turning to the other guests on talk shows and saying, “Are you still awake Charlie? You’re staying up late tonight. Very good.” Rickles even made that joke when he performed for Reagan’s second inauguration. After a series of cracks about Charlton Heston and Elizabeth Taylor, he turned round and went: “Is this too fast Ronnie?”

There is a continuity between this and the more demeaning approach Trump takes with reporters. Telling Jeb Bush “Oh you’re a tough guy Jeb, I know” was exactly what Rickles would have said.

When Trump called Joe Biden Sleepy Joe, that was a Rickles moment. So it was during the debate when he responded to Biden’s comments by saying, “I really don’t know what he said at the end of that and I don’t think he knows either.” When he told reporters about people saying he closed his eyes during a three-hour Cabinet Meeting, “Look, it got pretty boring,” that was a Rickles move.

Trump, like Rickles, loves to roast people. “I mentioned food stamps and that guy who is seriously overweight went crazy.”

In the 2024 election he got laughs by imitating Biden looking confused, just like Rickles joking about Letterman having Alzheimer’s. When he told a woman reporter, “I know you’re not thinking: you never do,” that was pure Rickles. This is why there’s a whole genre of YouTube videos where people compile Trump roasting and insulting people.

Sometimes he gets booed, like when he told Jeb Bush to be quiet. And then he snaps back like Rickles. “That’s all his donors out there. Boo this.” That is a classic Rickles mode. His jokes often crossed the line, and when they did, and the audience hissed a little, he actually said, “Boo this.”

Trump knows what Rickles knew—when you are playing the game, you have to snap back. That’s what keeps people entertained. But if Rickles’ act was all a game, Trump’s is not.

Reagan had quite an opposite style. He knew how to pause for effect. His jokes were about ideas, not people. He was ideologically consistent. His delivery was even-tempered and charming. Reagan was a friendly uncle. Trump is a nasty uncle.

Something all of us would like to do if we had no class

In response to the idea that he was offensive, Rickles said “all I do is laugh at ourselves.” He thought the “offbeat words” people disliked in his act were “in the eyes of the beholder.” In an entirely serious moment on the Cavett show he compared his father’s ability to hug a woman without it being dirty to many other men, who, if they hugged a woman, would make them uncomfortable. “It’s the way you do something.” This is a typical analogy for Rickles. But it cuts to the heart of the issue. Of course, he concluded this moment by saying, “And I tell you Dick, from the bottom of my heart: I never liked you.”

“You see,” Cavett laughed, “you can’t be serious.” Cavett then continued,

I think people don’t admit that deep down inside that you do something on the stage that all of us would like to do if we had no class.

This is what Trump represents. For some people, it is always offensive to say these things. For others, Rickles was a sort of pretense that acted like a valve. Donald Trump is now that valve. He says things people want to be able to say, however unacceptable or offensive they might be.

Libidinal liberalism?

It’s a longstanding point of progressive liberal politics that speech can be violence, or that nasty words can have real world consequences. In a world where micro-aggressions are problematic, Trump is macro-aggressive. Another variation of liberalism, as argued by Oliver Traldi, believes that these speech vices might act as a sort-of safety valve.

Traldi posits that “libidinal liberalism”

is the idea that many of the behaviors governments try to restrict are driven by basic, urgent, perhaps irrepressible human drives. The specific behaviors that are restricted, on this view, are merely the outlets for these fundamental drives, and the restriction is said to “bottle them up.” The idea of “bottling up” is that the drive is just made more intense by being contained, and that its inevitable expression will be more extreme, dangerous, or violent than it would otherwise have been.

Libidinal liberalism argues that pornography might reduce violence towards women “by providing a safe outlet for male sexual desire” or that legalised abortion allows those abortions that would already happen to happen safely.

Some liberals believe in the perfectibility of human society. They think we can overcome the offensive humour of Don Rickles. The libidinal liberalism thesis would say it’s little wonder that so much offensive speech appears on Twitter or in the President’s rhetoric when we took away the safety valve of offensive stand-up comedy. Without people like Don Rickles, the human urge to say offensive things—or to hear offensive things being said—is bottled-up until it cannot be contained.

And, in this case, it re-emerged in the political rhetoric of the President. Classical liberals and libertarians are often guilty of valuing freedom so highly that they don’t take enough account of the real costs and harms of things like drugs, pornography, and online radicalization. The rejoinder, which Traldi discusses, is the idea that pornography is not merely a safety valve but something that creates new, darker desires. On this view, Trump’s rhetoric is not merely a re-emergence of Rickles’ humour, but a dangerous replacement of it. What is funny in the comedy hall becomes scary at the political podium.

Either way, what Traldi says about Trump and his removal from social media platforms might in turn apply to the emergence of political correctness that saw Rickles’ style of humour become unacceptable.

…the libidinal line of reasoning says that Trump and his fans are bad for America and the world, that they’re probably genuinely evil or even white supremacists, but says further: If you ban them, they’ll have to go somewhere; they can’t be extinguished; they are something we have to control and channel.

Rickles was not a white supremacist, but his offensive remarks made people laugh for a reason, and they did go somewhere. The White House.

The imperfectability of man

I think of this not as libidinal but tragic liberalism. Tragic liberalism accepts that while human flourishing, progress, and improvement are central to the liberal project, humans are not perfectible. Human society cannot attain utopian goals. Some darkness always remains.

There is a scene in Coriolanus when the great war hero comes to dinner with his former enemies. Having been exiled from Rome, Coriolanus has come to Rome’s nemesis to discuss an invasion. Shakespeare’s genius is to not show us the discussion between the war-like generals, but instead to have it relayed second-hand by the servingmen. One of them comes rushing in to the kitchen, full of the exciting news that war is being discussed. Rather than reacting with despair, the servingmen are excited. One of them says,

Why, then we shall have a stirring world again. This peace is nothing, but to rust iron, increase tailors, and breed ballad-makers.

And another,

Let me have war, say I; it exceeds peace as far as day does night; it’s spritely, waking, audible, and full of vent. Peace is a very apoplexy, lethargy; mulled, deaf, sleepy, insensible; a getter of more bastard children than war’s a destroyer of men.

Aren’t these the sort of sentiments we heard from Rickles and Trump? Can’t you just hear Rickles saying to the audience, “You sir, what do you do? Ballad maker. Of course you are.” And isn’t that phrase “it’s spritely, waking, audible, and full of vent” exactly what people want from Trump? He really is audible and full of vent.

In retrospect, while Rickles was distasteful and unkind, we might wonder if his approach to comedy was more liberal than what we have now with Trump’s rhetoric. In the setting of a stage, offensive humour can be regulated—Rickles went out of date, he was booed, people had to pay—or not pay—to hear him. Society moved away, on the whole, from that sort of material. But now, Trump is audible and full of vent and we all have to listen. It is no longer merely a joke in bad taste, it is a mode of governance. Rickles was a vulgar ballad-maker; Trump is a maker of wars.

Rickles said, “I make fun of the President. I make fun of everyone. That’s America.” Well, now it is the President making fun of everyone—and the joke is on us.

I gave up on liberal interventionism the day the Chilcot Report was published. I was working at a think tank in Westminster, just round the corner from the QEII — a concrete hangar of an events centre, where, that afternoon, Tony Blair was questioned by the press for two hours about the Chilcot Inquiry’s analysis of the UK’s part in the Iraq War. At the think tank, we watched it live on a big TV.

Blair gave a masterful performance: serious, fluent, substantive. It’s hard to think of another contemporary UK politician who could’ve responded in such a manner. The problem, of course, was that the strength of Blair’s performance served to emphasise the report’s conclusion that the case for WMD had been overstated, and that the UK had been pushed into a badly planned and legally unsatisfactory military intervention. He’s such a showman.

But I’ll avoid getting into many details here about Tony Blair. Rather, I want to think about why I changed my mind that day — not only about that particular war, but about liberal interventionism more generally.

After all, many other members of the UK public had been convinced for years that they’d been misled over the invasion and subsequent war in Iraq. Many of them had argued against the war, on both moral and legal grounds, since before it began. And what’s more, most members of the UK public are nowhere near as generally sceptical about state power as I am. Yet I had considered the UK’s part in the war to be justified — beforehand, during, and for years after.

In fact, my support for the war — or for the initial military action, at least — was overdetermined. I supported it mainly because I thought that if we could free the Iraqis from Saddam Hussein’s violent oppression, then we should. But I also supported it because I assumed there must be good evidence that Hussein was close to attaining nuclear capacity. And a nuclear Iraq was something I thought should be prevented, for many reasons.

To generalise a little, my reasoning reflected three key beliefs I held strongly back then: 1) that it can be justifiable to take military action in order to overthrow a violent authoritarian regime; 2) that it can be justifiable to take military action in order to prevent such regimes attaining nuclear weapons; and 3) that the leaders of war-going democratic nations typically have, and act on, crucial information about these matters, which they are unable to share publicly.

On the day of the Chilcot Report, I changed my mind about the relevance of the third belief — about the relevance, that is, of secrets to the justification of military action. I also concluded that while freeing people from authoritarian rule is a good aim, I should accept that military intervention isn’t a feasible method of bringing it about. Since then, I’ve changed my thinking about the second of these matters, but I’ll get to that in a minute. First, I want to talk about the relevance of secrets.

1) Don’t depend on secrecy

I can’t remember a time when I didn’t believe that there must be secret information, known by senior politicians and military leaders, which could strengthen the case for pretty much any instance of military action taken by a democratic nation. I assume this is a relatively standard belief. But on the day of the Chilcot Report, I realised that I should never again rely upon this belief when evaluating whether going to war was justified or not.

Of course, perhaps you’ll tell me that just because the Chilcot Report ‘showed’ that the security services didn’t have the kind of information about WMD that Parliament and the public were led to believe, this doesn’t mean that such information didn’t exist. Perhaps you’ll tell me that this information was so secret that the members of the Chilcot committee deemed it should be kept from the public — or that perhaps it was kept from the committee. Indeed, these are the kinds of things I’m still tempted to say, myself!

But aside from the fact that I found the report generally convincing, I realised that day, while listening to Tony Blair, that the biggest problem wasn’t that he’d led us to believe in secret information that likely never existed. Rather, it was that something as non-substantive as beliefs about secrets had been allowed to play a decisive role in something as morally serious as going to war.

I’ve argued elsewhere that state actors have serious obligations to be transparent about their actions and their reasoning — obligations that correlate with moral rights held by members of the public. Beyond exceptions around relevance, I believe that state actors are justified in withholding such information from the public only when doing so is in line with democratically deliberated and determined public rules. And even then only on a temporary basis.

On my account, therefore, if state actors deem secret information necessary to the case for military action, then that information must be revealed to the public. This isn’t just because I think a necessary condition of legitimate military action is that it is ‘backed by the nation’ through the process of its decision-makers having gained the support of legislators. It’s also because the information that such decision-makers depend upon requires extremely thorough testing. Openly deliberating about these matters brings epistemic advantages, therefore, as well as legitimacy.1

Now, I’m happy to accept that in certain urgent situations — like unfolding terrorist attacks — immediate state action is required, and the temporary withholding of relevant information can be justified on the grounds of the avoidance of harm. But going to war is not a matter for quick decisions! Indeed, instances in which leaders feel the need to take immediate military action — to push the nuclear button before the enemy does, for example — are instances we should be protected against. Justified public rules governing such instances should be set in place during times of peace, to guard against carelessness and against the dangerous idea that legitimacy can be thrown aside in times of seeming urgency.

Haste and secret information have no place, I’ve come to realise, in justifications for going to war. Neither do unevaluated social norms. I mean, just because — as I put it above — “the leaders of war-going democratic nations typically have, and act on, crucial information about these matters, which they are unable to share publicly”, does not automatically mean that this is okay!

2) The goals must be explicit

A further conclusion I’ve reached is that for military action to be justified, its goals must be made explicit in advance. This relates to my claim above that state actors have rights-correlative obligations to be transparent not only about their actions, but also their reasoning. Leaders should be open about their reasons for going to war; the strength of these reasons, alone, cannot suffice.

This points to another failing in my previous position. As I said above, my main reason for supporting the Iraq War was that I thought that if we could free the Iraqis from Saddam Hussein’s violent oppression, then we should. I believed that the people of Iraq had the right to non-authoritarian rule. And I believed they had the right not to suffer violent oppression and torture. I still believe these things; I’m fully convinced they are true things. But I’ve come to accept that the truth of these things does not mean that the people of the UK necessarily held the obligation — or even were morally permitted — to intervene. Justification was still required. It was required for any nation to intervene, and it was required for the UK, in particular, to intervene.

So yes, rights entail correlative obligations, but we must always ask: what are these obligations, and who holds them?

Beyond that, I’ve come to accept that even if the Iraq War had been short, and even if there had been minimal casualties, and even if Iraq were now a flourishing democracy, then this wouldn’t retrospectively justify the failure of UK decision-makers to be open about their reasons for taking military action. If freeing the Iraqi people was the goal — or one goal among several — of taking this action, then that should’ve been made explicit. And as with my point about secret information, openness about this would have brought epistemic benefits as well as a possible route to legitimacy. Whereas, if freeing the Iraqi people was not a goal of taking military action, then neither the goodness of that goal, nor achievement towards it, can be depended upon as a post-hoc justification.

Of course, it is difficult — even with the 2.6 million words of the Chilcot Report — to know enough about the reasons for the UK’s involvement in the war, because what we do know is that the government was not sufficiently transparent. But for my purposes, what matters here is that freeing the Iraqi people was not a publicly stated goal of the UK’s involvement in the war. And it wasn’t treated as such within the formal public deliberation process. Indeed, Blair stated in Parliament that regime change was “not the purpose of our action; our purpose is to disarm Iraq of weapons of mass destruction”. What matters also, however, is whether goals like freeing the Iraqi people can serve as justifiable goals of taking military action. These things matter regardless of whether Hussein had or was likely to obtain WMD, and regardless of what UK decision-makers knew about that.

This brings us full-on to the question of liberal interventionism. That is, these conclusions about explicit reasoning and goals are crucial to evaluating liberal interventionism, because liberal interventionism pertains to particular reasons for going to war. I won’t provide a run-down of alternative conceptions, but when I say ‘liberal interventionism’, I’m referring to military action taken by a democratic nation with the goal of freeing the people of a foreign nation from authoritarian rule.

The arguments I’ve presented so far don’t dismiss liberal interventionism, therefore. Rather, they simply set some limits on it, and on all instances of military action. So I’m now going to discuss the realisation that turned me against liberal interventionism, per se, on the day of the Chilcot Report. This realisation takes the form of an argument about feasibility. Then, I’ll end this piece by telling you about the better argument that convinces me today.

3) Is it feasible?

I’ve never been much into Kant, but on the day of the Chilcot Report, I thought a lot about the Kantian idea that ‘ought implies can’. This is the idea that you can’t have moral obligations that you can’t fulfil. And I came round, that day, to concluding that Iraq was one time too many: that its failings exemplified the way in which military intervention isn’t a feasible route to bringing about liberalisation. Or, at least, not in anywhere near a sufficiently reliable manner.

Indeed, I beat myself up about why it had taken me so long to accept this. Why had it taken the exposure of UK mendacity for me to accept the truths of UK overreach?

The post-WW2-war period has not been kind to liberal interventionism. Think, in particular, of Iraq and Afghanistan. But also go back to Greece and Vietnam and Laos, and all those Latin American countries in between. As above, don’t think about instances where freeing the people from an authoritarian regime was not a central goal — stated or secret — of military action, or instances where different improvements to an intervened-upon nation have taken place.2 Liberal interventionism requires some assurance that military action has the possibility, and ideally the likelihood, of bringing about positive regime change. Yet recent history stands in the way. It seems that even when a liberal regime is put in place during such situations — often at extremely high human cost — it’s hard to make it durable.

Many people who still believe in the goal of liberal interventionism have come to accept this feasibility constraint on its enaction, therefore. And I assume that, for most of these people, the strongest explanation for this feasibility constraint is that lasting liberalisation must be bottom-up: that the people of the nation must lead the charge; that skin in the game, and belief in its rules, is necessary to its long-term success.

Two features of this kind of reasoning have come to concern me, however. My first concern relates to the dismissal of liberal interventionism through reference to a historical pattern. I have various problems with the idea of making historical occurrences into normative grounds, which I won’t go into here. But one problem I have with the ‘historical pattern’ argument, in particular, relates to its specific focus on a collected set of outcomes: on its dependence on things that have happened in the past, on the relations of these things to each other, and on what these relations can tell us about the future.

Now, if you’ve read my writing previously, you might assume that I’m about to start criticising consequentialism. Sometimes, people who know how strongly I oppose consequentialism make fun of me when I point out the moral relevance of a consequence, or of a set of consequences. I usually respond to them by saying that of course consequences are important — they’re just not the only thing that’s important! That is, that when you’re evaluating the moral value or disvalue of an action, or a state of affairs, or a rule, or a law, or a principle — or anything that can be assessed for goodness or rectitude — then one of many relevant considerations is the consequence.

And surely the track record of liberal interventionism involves much more than consequences. I mean, it’s easy to think of instances of bad behaviour within recent such interventions — bad behaviour ranging from the lack of transparency to war crimes — as well as instances of bad decision-making, and so on. Nonetheless, the problem remains that if we’re pushing all of these wars together, as linked elements within a pattern, in order to make assessments about what should be done in the future, then we’re at risk of falling into similar traps to the consequentialists.

A second concern relates to my general uneasiness about depending on predictions within justificatory arguments. Most of this uneasiness pertains to my appreciation for epistemic humility. And in particular, to the fact that some things are much less predictable than others. Certain reliable predictions — like ‘this guy is about to stab me with a massive knife, so if I don’t do something to protect myself, then I’ll probably die’ — seem strongly relevant, for instance, to determining whether someone acted in self defence. And general rules about everyday practices such as financial planning — like ‘it makes sense to have enough liquid wealth to rely upon in moments of illness or other trouble’ — seem reliable general rules of thumb, particularly during what seem like stable times.

But it’s very hard to see how something like regime change could fit into these kinds of buckets, almost by definition. I mean, upheaval clearly breaks reliable patterns and increases the chance of unpredictable unintended consequences. No war is the same as any previous war. How could it be, when wars involve almost uncountable numbers of actions and moments and events? And when wars are generally treated as exceptions, not least within moral theory.

So I find myself increasingly tending towards the conclusion that we cannot put much weight on the track record of liberal interventionism, either to support it or to oppose it. Of course, this isn’t to deny that we can pick out particular things that were good or bad or right or wrong in any particular instance of war — and seek to guard against the recurrence of the bad and wrong ones. And it isn’t to deny that we can come to some general conclusions like ‘attempting regime change in faraway countries comes with particularly serious epistemic and practical difficulties’.

Indeed, one thing that the military interventions of the past eighty years have taught me is that physical distance retains importance. We live in a much more interconnected world than the first of those interventions; countries like Iraq and Iran seem so much closer to us today, in so many ways. But war remains a physical matter — a matter of bombs and shrapnel. Perhaps one day, maybe soon, war will take the form of computer programs fighting computer programs to gain assets. But even then, the outcomes that hit the hardest, in moral terms at least, will relate to our human physicality: reduced access to basic goods like food and shelter; reduced access to urgent care. War is a physical matter, a local matter, so it’s hard to see how distance doesn’t matter.

Nonetheless, dismissing liberal interventionism on the grounds of feasibility — through a dependence on its recent track record, or through reference to particular constraints like distance — seems open to defeat by a single future example in which liberal interventionism works. Or a convincing example from the past. Except, of course, that ‘working’ cannot suffice, at least on my account. What I mean by this is that the method by which we reach the state of ‘working’ must itself also be permissible; the ends cannot justify the means! This is where we come to my current position.

4) Only in defence

One of the biggest changes of opinion I’ve undergone since the day of the Chilcot Report is that I’ve come to believe that physical violence can only be justified for defensive purposes. This conclusion has forced me to rethink my views on various matters that are central to moral and political philosophy, including punishment. I recently wrote here, for instance, about how “my guess and hope is that our descendants will look back in horror, and struggle to believe that we really imprisoned all these non-violent people”. The distinction between locking people up for defensive reasons, and locking people up for punitive reasons, is a distinction that has come to mean a lot to me.

My opposition to non-defensive physical violence has also recently helped me come to the conclusion that it is wrong to treat wartime as a time of moral exception. This piece has been mostly about ‘going to war’, however, rather than ‘times of war’, so I’ll leave my argument against that part of Just War Theory, for another day. But my new position means that when assessing any ‘successful’ outcomes of liberal interventionism, I have additional reason to take into account every violent act that took place along the way.

In other words, I can’t dismiss the harms of war on the grounds that they furthered any ends, and I also can’t dismiss them on the grounds that different moral standards obtain during such times. The suffering of every person who is harmed as a result of military action must be taken seriously, therefore, regardless of any harm they were suffering beforehand, and regardless of whether they’re a civilian or not. And that’s before we turn to the wrongs of war that aren’t covered by discussion of its harms.

Now, perhaps you want to tell me that my approach to determining permissibility during wartime is extremely burdensome — well, good. Bombs and shrapnel are serious matters. Bombs and shrapnel should stand in serious conflict with liberal commitments.

My opposition to liberal interventionism now hinges, therefore, on my belief in the crucial distinction between the following two premises: 1) that it can be justifiable to take military action in order to defend the inhabitants of a foreign nation against the regime that violently oppresses them; and 2) that it can be justifiable to take military action in order to overthrow an oppressive foreign regime and replace it with a liberal alternative. I’ve come to the conclusion that I can — and should — retain my conditional support for the former, while opposing the latter.

Of course, complicated questions remain about what counts here as defence and offence. About whether, for instance, terrorist attacks against Country A that are sanctioned by the government of Country B count as the kind of activity that justifies ‘defensive’ military action. This brings us back to the complicated nature of moral evaluation. It helps me to accept that first-mover war advantages cannot be held by liberals; that the lives of people in foreign nations should never be traded away for future tactical military advantage, no matter the cost to your own forces. And it helps me to accept that the distinction between offence and defence is a starting point, rather than an end.

It’s not easy being a liberal. But it is good.

What I find surprising, as always, is how un-literary, how un-liberal was the response to the news that the Post closed Book World. For a group of people who prize independent thought, I saw an awful lot of people who believed the same thing everyone else believed, and for the same ideological reasons.

Can we not regret the way (and the fact) that individuals lost their jobs but retain a disinterested view about the overall effect of the decision?1 Can we see it as part of the long decline of an industry insufficiently adapted to the modern world? I am wondering why there isn’t more negative capability in the way people think about these issues.

I don’t know if what happened at the Post is good, bad, or null effect as far as the overall position of literature and criticism goes. I find other people’s certainty about it (literally) unbelievable.

This is because I think about markets very differently. The essence of the Hayekian view of the market is that it is a process of discovery, always in a state of disequilibrium. This is not an unknown way of thinking, or ought not to be, to the literary mind.

My politics are not the politics of this discussion: maybe the wrong decisions were taken, and maybe that was detrimental to the book section—or maybe the large amounts of money Bezos spent subsidizing the entire paper for several years still weren’t enough to save it.

The overall situation, though, is hardly amenable to the (well-expressed) idea that literature is important and that we still need critics. One former member of the Washington Post Book Review has acquired tens of thousands of subscribers on here, thousands of them paying. The market has shifted a little. In the long death of newspapers, this is a small funeral.

My classical liberalism means I am interested in the readers and the decentralised system they create. Every click, every purchase, every book blogged about, every enthusiastic tweet, is a twitch on the thread that draws us to the new future. We don’t know yet how things will turn out: we do know that the institutional writers are often in competition with the internet in ways that make it harder for them to prosper.

One of the best critics at the Post, Becca Rothfeld, wrote in the New Yorker, that

The books section of a newspaper plays an altogether different role. It does not cater to aficionados; it seeks new recruits… A newspaper is—or ought to be—the opposite of an algorithm, a bastion of enlightened generalism in an era of hyperspecialization and personalized marketing.

Maybe newspapers did that, once upon a time. But human action has moved on. I can tell you from a decade working in advertising that the people in charge of “hyperspecialization and personalized marketing” are constantly frustrated with just how impersonal marketing remains. You are followed by the toaster advert because they are unable to target you any more specifically than that. The difficulty of knowing how to spend your advertising dollars is still very real, which is why advertising proliferates and you ignore most of what you (don’t) see.

As for the algorithm problem, I have people reading my literary blog The Common Reader who work in Silicon Valley, just as Naomi Kanakia has many readers in the rationalist community. These are two groups who are suspected of being “unliterary”, but who seem to be benefitting from the ability of the internet to find new recruits. The whole internet is your newspaper now!

Shortly before this news, Becca Rothfeld wrote in the Point, (which is also on Substack) about the problems of modern liberalism—the paucity of the abundance movement and the associated “Parks and Rec” aesthetic.

The sort of art and argument that could make its audience want to be liberal would have to begin by regarding its audience as agents. It would have to enlist them as equals instead of demoting them to the role of pupils; it would have to demonstrate just what form—or, more appropriately to the liberal sensibility, forms—the beautiful abrasions of communal self-determination might take.

So then why must we talk about “cultural production” and be ideologically aligned about the problem of billionaires? Why does the unending negotiation with the difficult and intransigent adventure of humanity always lead the literati to have the same left-liberal suspicions of capitalism? At what point do we have to ask not about the lack of aesthetics of our politics but the insistent politics of our aesthetics?

If the abundance bros are claimed to have the same disinterested Obama-era Parks and Rec aesthetics, then perhaps we can admit there is something bien pensant about the political vision of many critics. Maybe I am heartless because I do not know the people who lost their jobs, but the question of criticism’s place in our culture is not the same as the question of who works for the Washington Post.

Human action is as ceaseless as the tide, as everlasting as the wind. We all make decisions, and in those decisions some information is revealed that becomes part of how a price is formed. In every choice we make—the coffee not the scone, the organic fruit, no alcohol this month, more Substack subscriptions, a different sort of car—we are part of the everlasting fluctuation of human activity that creates an economy. From these decisions, prices emerge, incentives are set, and people make further decisions.

This is not to diminish the role of human agency—everything is human action! But it is to note that whatever decisions were badly made in the recent shuttering of Book World, those decisions were responsive to all those other choices that were also being made by a lot of other people, many of them the readers on whose behalf critics are making their laments.

Competing with other newspapers is a different task to competing with the whole internet. But ought we not to be at least neutral about the ability of literary (and liberal) criticism to be part of this new era of common reading just as the journals of the eighteenth and nineteenth centuries once were? After all, the internet brings us small poetry presses too. And the existence of The Point itself—a successful online literary journal—makes the argument in the best way of all.

The world is moving on and it is of very little use to ask it to stop.

There’s this guy, Alex Honnold, who just climbed a skyscraper in Taipei, the Taipei 101 — so named because it’s 101 stories tall. Honnold’s virtuosic climbs are in the “free solo” style, meaning that he uses no equipment for protection or assistance. He has some sort of special shoes, I think, and in the videos you can see him chalking his hands, which must help somehow. Other than that, it’s just him and the nearly flat surface he’s taken on — a skyscraper, the nearly-bare face of a mountain, whatever. Apparently when the 2018 documentary Free Solo about Honnold was filmed, most of the crew looked away repeatedly during his climbs.

Rebecca and Henry have asked me to write something for this blog about liberal education. What is a liberal education, they ask? Why does the “Great Books” stuff I’ve been doing as a teacher for the past few years, they ask, fit the model? My former Tulsa Honors colleague Dan Walden just came out with an essay about these topics in The Point, too. I largely agree with Dan, but with some differences of emphasis.

I think an appropriate liberal education is one that teaches us to “free solo,” intellectually speaking. The deepest, most difficult questions that arise in our lives stare us down like barren rock walls, unmarked and unscalable, pristine in a precipitous terror of sand. To live thoughtfully, reasonably, and morally is no small task, let alone to do so freely and authentically. Yet this is what our very ability to think and act seems to demand of us – at least to some of us, the way a mountain might demand to a free soloist to be climbed. Through a liberal education, free people grow in their capabilities, their powers, until few things accessible to their reason and will remain foreign to them.

“Liberal education” is so-called not because it emerges from the philosophy of liberalism but because it was, in the ancient world, the sort of education thought to be worthy of, or perhaps required of, a free person. This conception, of course, is somewhat out of step with our more modern, more egalitarian understanding of freedom as something that falls, like rain in the Bible, on the worthy and the unworthy alike. Indeed, education can be a real toughie for liberal philosophers. Why think that some “expert” teacher at the front of the room knows better than the “marketplace of ideas” in the student audience? Why let parents and teachers order children around and discipline them, given the whole freedom-is-like-rain thing?

At the same time, there is obviously something very liberal about emphasizing education. A system that sees the growing power of people’s minds not as a threat to present exclusive rule but as a boon to future mutual rule, which aims to provide tools like reason and rhetoric to a citizenry rather than hoarding them, is exemplary of the liberal-democratic approach to society and governance. Even in contemporary times, the liberal arts are often defended by reference to some goal of being a good or informed citizen — especially in new “civic education” centers and projects like the one I teach at now in Toledo, Ohio.

It’s hard to say just what our civic responsibilities are such that a liberal education would enable us to fulfill them. Because of the specific nature of my research into group epistemology, I tend personally to focus on the ability to maintain one’s independent judgment and in doing so add to an aggregate epistemic enterprise — the so-called marketplace of ideas on a Millian picture, the popular vote on a Condorcet-type picture, and so on. And because I’m a philosopher, I tend to think that the best training of independent judgment develops first-principles reasoning.

It’s great if liberal education has these effects, but I want to contrast this way of thinking with my way of thinking about liberal education. To me, education should be fundamentally concerned with developing student capacities. Of course, these capacities could eventually be put to some sort of public use, but really they are of the most use to the learners themselves. Thus I don’t categorically separate liberal education from training in “skills.” Instead, liberal education develops a wide variety of specifically intellectual (or “mental,” or “cognitive”) capacities, which we can use to take on a wide variety of questions and problems in life. When we come up against life’s big questions, we should have a dizzying array of personal powers at our disposal: art and literature, science and philosophy, history and theology.

I don’t have a great account of just which intellectual skills should be developed in a liberal education. Mathematics and logic, for instance, seem like they fall well inside the ambit. Chess and poker are a bit more questionable, although they might be useful applications or tests of different types of reasoning. Physical fitness seems outside the scope, although I’m not necessarily opposed to the ancient idea of training the mind alongside the body.

What’s crucial and crucially missing in much of modern schooling is the challenge inherent in a genuine liberal education. This challenge shouldn’t be a matter of “viewpoint diversity” or encountering perspectives that might offend one’s sensibilities, but rather a matter of raw difficulty. Our abilities can only be developed through very hard work. This work can occur in many different disciplines, but a liberal education is not an “interdisciplinary” one so much as one in which concepts like subfield and profession take a back seat.

One way in which I think my analogy is truly apt here is that these challenges must be “free” in the sense that free soloing is: free of artificial aids to ascent. In particular, liberal education is not the sort of education that can be assisted by the autocomplete function of a large language model, or through reference to an encyclopedia, or through deference to an expert consensus, or through inspiration from a demagogue. A line of thought in the academic literature on the social epistemology of ethics suggests that moral deference – treating someone else as an expert on right and wrong and doing what they say – is somehow “fishy.” I don’t know if I completely buy this, but deference on big questions does not suffice for a liberal education. That said, familiarity with major theories and historic and religious traditions’ answers to those questions does seem to be a core aspect of liberal learning.

Though our attempts must be “solo,” our learning needn’t. In fact, I think some of the best liberal learning occurs in a certain kind of group, like a good seminar. In my opinion, the best seminars don’t involve anything like group cognition or any sort of consensus-building. That is what happens in a political group, not an intellectual one. Rather, a seminar in the mode of liberal education allows participants to offer their own interpretations of readings, their own understandings of puzzles and problems, and their own answers to questions, training their free solo mindset and skills but while being “spotted,” so to speak, by their co-learners. More than anything, this setting demands of students that they think for themselves.

One specific thing Rebecca and Henry asked me to address is how the Great Books fit into all this. I don’t buy a lot of standard justifications for studying the Great Books. I don’t care that they’re “our” tradition, and I don’t know if it’s important to have a kind of “story of thought” where we figure out where our own ideas came from; actually, I hate those kinds of genealogies. I don’t care about “the ancients and the moderns” or whatever. I don’t buy that they’re things that “every educated person should know” – just try to talk to an average educated person about them. I also don’t share the visceral sense of “greatness” others who do this stuff have. For instance, a friend recently told me that Heidegger is palpably “greater” than, for instance, Rawls or Parfit; I think this is silly. But the Great Books are great didactic tools for a few other reasons.

First, many Great Books model the free solo approach to the life of the mind for students. The Greats include the first philosophers, the first historians, the first poets. They look at the world and look at their minds and think of ways to connect the two together, without having a list of jobs or disciplines to refer to. This opens up questions for us like: What is philosophy, or history, or poetry for, to begin with? Why did someone even start to think of these as activities that were worth doing and that were open to people like us? Non-first Greats have some sort of feeling of firstness about them, of thinking in a fresh or new way. Descartes, for instance, wanted to prove things like the existence of the external world and the existence of God from truly first principles.

The Great Books are models in other ways. Many authors of Great Books found themselves in difficult personal circumstances, which led to unique insights, strategies, or obsessions. The death of Plato’s teacher, the imprisonment of Boethius, the torture of Machiavelli, the civil war facing Hobbes, the assassination of Martin Luther King, Jr.: all of these make it obvious that these writers were not just playing intellectual games with language or logic, as we analytic philosophers are often accused of doing, but figuring out how to live in light of the activity of their intellects.

The Great Books are also challenging. This is part of the advertising of many Great Books programs. Education overall is in a period of horribly declining standards. Teachers tell me their students can’t read more than a paragraph or sit still for more than ten minutes. Everyone uses artificial intelligence to write everything. Great Books programs, by contrast, pile long, difficult, foreign texts in front of their students. This trains them to be powerful thinkers and acquaints them with a wide variety of modes and schools of thought. They do not do so uniquely, and in my opinion should be paired with other difficult academic pursuits, including math, physics, logic, and foreign languages. But they are a rare spot of intentional challenge in the contemporary humanities.

Finally, the diversity of a Great Books syllabus tends to explode student notions that a liberal education will involve giving them final answers to big questions. This can be very unsettling or even disappointing for students trying to navigate the world. But when they see that so many Greats disagree so radically with one another, even when sharing a cultural, political, or religious context, students begin to understand that they will have to make their own assessments and decisions. The inescapability of this challenge is a tough lesson in itself, but one which makes even sweeter the realization that the challenge is, in fact, manageable.

So there you have it: my theory of liberal education. True liberal education provides people with the ability to “free solo” through the sorts of difficult questions and challenges that life presents. It trains us to be free because it demands that we do our own thinking and builds up our abilities to do so through challenge and difficulty and through models of others who have. What we find as we age is that many of the questions and situations raised in Great Books in particular, which seemed so abstract or antiquated or absurd or melodramatic or mythopoetic when we first encountered them, are actually the stuff of life. And life is, though many undergraduates have not reached the point of realizing this, very hard. Liberal education gives us the tools to face it head-on, equipped with the strongest possible form of our native reason and maintaining our integrity, our dignity, and our self-respect. We will almost all fall many times in the attempt.

The question was: aren’t political rights more important than economic? We need the dignity of representation, not the opportunity to make the rich richer. Isn’t the history of freedom the history of political freedom?

At this point the table split between an abstract notion of politics, and an empirical, realistic one. It is little use talking of justice if one doesn’t have a good account of the facts of life. We can talk all day in abstract terms about whether my rights of speech, free association, and personal liberty ought to entail the right to economic freedom—and the basic fact of their inseparability. But the world’s best theory of just redistribution will face an implacable object in the face of a society that cannot generate wealth.

It is instructive to look at history. What use were the sort of political rights we prize before the Great Enrichment? When a man labors for subsistence in a field, democracy can do little for him.

The women’s rights movement at the end of the nineteenth century was divided about the importance of the vote. What women wanted was a whole raft of rights—divorce, property ownership, labor market access, professional status, education, and so on. Campaigning for the vote was the best route to get those rights. Once women were half of the electorate, the rights they wanted would follow.

Political rights are not just inherently valuable as a means of dignity: they are the means by which an individual’s freedom to act in society are secured. When you have few rights, like women in the 1870s, it is obvious that political and economic rights are unified.

A free individual is free to make economic arrangements as well as political ones. The grubby presence of money doesn’t undo that. If you want equality, the right to divorce is not essentially different than the right to own property, the right to speak is not essentially different than the right to work.

What good would it have been to give women the vote but then deny them the right to work or exercise their talents? Giving people the freedom to choose for themselves means accepting that their political and economic rights cannot be easily divided.

Everywhere I look, I see people referring to AI doing things like reasoning. I see this in casual chat. I see it when I use AI products. And I see it in formal writing like this Nvidia glossary entry:

“What Is AI Reasoning? AI reasoning is how AI systems analyze and solve problems by evaluating various outcomes and selecting the best solution, similar to human decision-making.”

I’ve written before about how I think the word ‘zombie’ should be placed in front of all such verb uses. AI is not reasoning, on my view; AI can only do ‘zombie reasoning’! Neither is AI evaluating or selecting.

I think the word ‘zombie’ should be added here because I do not believe that AI has the kind of interiority that is necessary to doing these things. That is, I don’t believe that AI is reasoning or evaluating — or even selecting — in anywhere near the sense we ordinarily talk about such things. Rather, I believe that AI is simply producing outputs of the kinds that a human doing such things would produce.

Don’t get me wrong, I think it’s amazing that AI can produce such outputs! And that these outputs are getting better, in so many ways, extremely quickly.

But if you are into philosophy, then you will see that I’m making an implicit distinction here between conscious activity understood as a phenomenological matter, and conscious activity understood as a functional matter. You will also see that I am implicitly claiming that activities like reasoning and evaluating, and even selecting — on an ordinary use of such terms — can only be done by the kinds of things that have phenomenological consciousness.

You don’t need to know about all that meta stuff to get my point, however. Rather, you just need to read the following, which I wrote elsewhere on Substack a while back:

“to use a standard philosopher analogy, AI reasoning is ‘zombie reasoning’. That is, in the same way you can imagine a ‘zombie you’, who looks just like you, and goes about your life doing the things you do, but has no interiority, you can think of ‘zombie reasoning’ as reflecting the outputs, and even some of the processes, of the reasoning of a free agent — but also without recourse to interiority. It’s the same with AI ‘deciding’ and ‘thinking’ and ‘knowing’, and some of the less introspective concepts like ‘acting’. The reason it makes sense to put these words in inverted commas, that is, is because the AI versions of these activities are shadow versions of what we humans are capable of. Of course, it’s hard for us to conceive what it would be like to do these things on the shadow level, but AI has it much harder, because it cannot conceive at all.”

I should note at this point that there are, of course, important long-running philosophical debates about the extent to which non-human animals can do things like reasoning — and indeed, less mentally-complex activities like thinking, and even feeling.

But I simply don’t accept that most of the people who talk about AI doing any of these things — reasoning, selecting, thinking, feeling, whatever — truly believe that AI is doing them in any more than a ‘zombie’ way. Whereas I do accept that many people think that at least some non-human animals can do some of these things in a non-zombie way. I even count myself as one of those people, these days!

The deepest reason I believe that most people do not think that AI can do any of these things in a non-zombie way relates to my assumption that most people do not think that AI is alive. Sure, there are some crazy people out there who do believe that AI is alive. And sure, there are some even crazier people who believe that AI doesn’t have to be alive in order to do things like reasoning in non-zombie ways! Astonishingly, none of the people I’ve met who believe this second thing — including, most recently, a pretty serious philosopher — have been able to explain to me how on earth this could be. I’m still waiting!

But I also see this phenomenon — this dropping in of these self-awareness-requiring terms — in pieces of writing where the writer is also keen to stress that they are not one of the crazy people.

I saw something like that, for instance, in the highly enjoyable piece Scott Alexander published yesterday about Moltbook:

“We can debate forever - we may very well be debating forever - whether AI really means anything it says in any deep sense. But regardless of whether it’s meaningful, it’s fascinating, the work of a bizarre and beautiful new lifeform. I’m not making any claims about their consciousness or moral worth. Butterflies probably don’t have much consciousness or moral worth, but are bizarre and beautiful lifeforms nonetheless. Maybe Moltbook will help people who previously only encountered LinkedInslop see AIs from a new perspective.

And if not, at least it makes the Moltbots happy:”

What can it possibly mean for something to make the Moltbots happy, if the Moltbots are not the kind of thing that has internal awareness? Okay, I’m reading extensive implicit claims into what Alexander is saying here, to come to this conclusion about his position! And again, there are important long-running philosophical debates about whether, for instance, dogs can be happy in the ordinary sense of the term.

But how could you be happy without being alive? Does Alexander really mean that the Moltbots are alive, when he describes them as “lifeforms”? That they are living things, in the sense that we ordinarily understand the term ‘living’? And even if he does believe this astonishing thing (!), then how could the Moltbots be happy without having any of the interiority that only phenomenologically conscious kinds of living things can have?

In other words, it’s hard not to come away from the Alexander extract thinking that Alexander is saying something like: ‘Hey, even if the Moltbots have no inner life, Moltbook makes them happy!’. Or less strongly: ‘Hey, I don’t need to get into discussing the “consciousness or moral worth” of the Moltbots, or whether or not they are able to mean anything they write, or anything like that, to be able to conclude that something can make them happy’.1

This is bizarre!

But why does it matter? Why am I here getting so exercised about language use, when I could be reading Moltbook, or looking at paintings in a nearby gallery?

Well, one of the things that is particularly special about being human is that we have this capacity for self-awareness. Again, I’m not claiming that no non-human animals have this capacity — I don’t need to get into that here. Rather, I’m simply claiming that the capacity for self-awareness is a central, obvious, and incredibly significant feature of being human.

If you are reading this piece — really reading it, rather than somehow just having my words ‘inputted’ into your brain — then you are self-aware. And, a fortiori, this self-awareness is necessary for us humans to be able to do more mentally complex things like deliberating and reasoning.

What would it mean to reason if you were not self-aware? If an evil scientist took away your self-awareness for five minutes, and somehow during that period they manipulated your brain into receiving and ordering some signals, then would you count that as reasoning? I mean, would you count the thing that had been done to you as ‘you reasoning’, even if it had led to a neat little ‘reasoned’ set of conclusion statements, produced by your manipulated brain?

And what if we were to turn to an example that featured no mysterious evil scientist manipulating your brain in the kind of way that’s reserved for philosophical thought experiments? I mean, what if we considered an instance in which your brain seemingly ‘came to some conclusion’ without any kind of external manipulation or self-aware input? Think here of those ‘automatic’ behaviours you sometimes carry out — whether it’s your leg jumping when the doctor hits your knee with a tiny mallet, or your hands turning the steering wheel even though you’re not really focusing on the road. Would you count those behaviours as the outputs of reasoning, ordinarily understood?

Without any self-aware deliberative element, reasoning becomes a zombie matter. Again, this isn’t to degrade the value of zombie reasoning, including its interest to those of us obsessed by the many philosophical questions arising! Rather, it is to emphasise the special nature of reasoning as we ordinarily use the term.

The specialness of reasoning doesn’t simply track an astonishing feature of being human — this way in which we can reflect on things, and weigh them in our minds! It also has important moral implications. If you have reasoned on whether to act in some way, for instance, then this can have important moral implications for your ensuing action. If you reason about whether or not to kill me, and then you kill me acting on that reasoning, then it becomes possible to count you as guilty in a way that a non-reasoning thing could never be counted. You become, as we philosophers like to say, a candidate for blameworthiness.

Reasoning is one of our most significant capacities as humans. Our capacity for reasoning sets normative constraints on how we live together. It has crucial implications for how we should treat the other things around us. For how we should treat ourselves. For how we should treat AI! This provides just one reason why we should refrain from loosely using the term ‘reasoning’ in ways that imply that non-reasoning things can reason. And it applies to how we should speak of AI activity, much more generally.

Okay, anthropomorphising the ‘actions’ of non-human things is hardly anything new. People talk about cars in such ways. They talk about toys in such ways. They certainly talk about robots in such ways. But AI is much more slippery than any of these things. Even young children can grasp that the car does not choose to drive along the road. Even young children can grasp that cutting their toys with scissors is different from cutting their siblings. And when children cannot grasp such things, it is urgent that we teach them!

Talking about the ‘actions’ of AI in loose ways comes with serious epistemic risk, therefore. Doing so will deaden our awareness to truths of the revolutionary moment in which we live. It will leave us open to manipulation by people with an interest in covering up the ways in which AI is developing. It will lead us to miss out on the rich and exciting intellectual opportunities that the development of AI offers us all — including opportunities to refine our thinking about important matters such as consciousness.

Now, I’m not so pessimistic and paternalistic as to propose that we must all start using my 'zombie’ language in order to save humankind! And I’m not so rigid as to argue that the meanings of words can never change. But I firmly believe that giving up on the ordinary way in which we humans have used crucial terms like ‘reasoning’, for thousands of years, does not come cheap, on any level.

If we want, as a species, to discuss the role that AI is playing — and should play — in our lives, then we need to get much better at talking about it.

Thanks to GPT for the zombie reasoning picture!

A lot of commentary has focussed on the “joined” people being like an LLM. Although the writer, Vince Gilligan, denies this, there are so many moments when the joined do very LLM-like things, and things that are discussed a lot in LLM discourse. (Read Hollis if you are interested in this.)

What I found more interesting about the joined, though, is the way that individualism becomes a more acceptable belief in the face of the collective.

The plot involves the main character, Carol, learning to take the joined seriously (and even sleeping with one of them, but that section doesn’t always make a lot of sense), and a large part of the show is about how to cross partisan divides. Ultimately, though, that doesn’t work. The joined are just too collective to respect Carol as a person. What does work is Carol’s stubborn sense of independence.

The opening sequence when the “joining virus” first infects humans shows the newly joined going around and forcing others to join them, either by kissing them, or by licking a communal box of donuts, or incubating the virus in petri dishes…

By the time everyone other than eight people have joined, though, the new collective is super nice and happy and literally wouldn’t harm a fly or even pluck an apple. It’s dramatically interesting to show a disagreeable individual refusing to join a group of smiley smiley people. O one can smile and smile and still be a villain! (The way they joined are presented and characterized is one of the main strengths of the show.)

But the dark core remains—they say they love her, but they don’t want her to exist separately from them. If they could lick her food and force her to become one of them, they would.

Community has to be voluntary, and no amount of telling Carol it is simply better to be a joiner can make her give up her sense of individual self. One man’s blissful community is another woman’s loss of autonomy and identity.

Individualism is easily maligned as “atomised individualism” or “pernicious individualism”. But the question of “would you join” at the heart of Pluribus refocuses on the real idea of individualism: you are a unique and autonomous person of equal worth with all others, who should not be forced into a state of living against your will.

Collectivist politics so often involve a certain amount of “with us or against us” and this mood is taking over many parts of the political spectrum today. Much in the way that Trump and the tariffs persuaded Noah Smith to re-think the value of libertarian ideas, I suspect that when confronted with the “community” of the joined, a lot of people would choose individualism the way Pluribus does.

Some of us have a greater appetite for community than others. One of the unjoined does in fact choose to join (it is significant that she is an impressionable young person, not a fully developed adult; it is quite a poignant scene) but the show clearly doesn’t think that the professed bliss of the joined is any sort of compensation for the loss of individuality.

At the end, we discover that the joined have got hold of Carol’s frozen eggs. They don’t need her permission (under their weird and slightly inconsistent morality of non-interference) to use these eggs to get her stem cells. At which point they can create a virus that will “join” her to them. However, they are also obliged (again, under their own morals) to give her whatever she asks for. So she gets an atom bomb, to prevent them from screwing with her eggs.

Carol has been friendly with them up to this point. Now she rejects the joined and agrees to work with one of the other non-joined characters, Manousos, who wishes not only to return the world to normal, but to destroy the joined, who he thinks are evil. I expect the show will prefer Carol’s approach, but either way after the appalling destruction of the teenage girl’s individuality, the show ends in a state of classic American resistance.

It would not be inconsistent to have the words “Don’t Tread on Me” appear at the end, perhaps with the libertarian porcupine rather than Gadsden’s original rattlesnake. Carol is something of a porcupine—prickly but vulnerable, only dangerous when she’s attacked—and however much I felt the last few episodes began to drag, I never tired of her insistence on her own inviolable identity.

A politician can say “rent freeze” and make every economist across the nation reach for a stress ball. The same politician can also do something that would make housing cheaper and small businesses easier to start. Sadly, as I have recently learned, in my circles only the first fact seems to count.

That mismatch is the puzzle I can’t shake. If you say you care about growth, supply, and competence, why does the policy stop mattering the moment it comes from a political label you oppose?

I’m a fairly standard DC think tank type. I do labor economics research, write papers, attend conferences, and sit through panels where people argue constructively about incentives and trade-offs. I spend my professional life in policy rather than politics, which is a comfortable place to be if you believe evidence can move institutions at the margins.

If I had to label my views, I’d call myself a small l liberal: committed to markets and liberal institutions, skeptical of mass politics, and deeply wary of moral frameworks that treat political disagreement as evidence of moral failure rather than reasonable pluralism. That worldview is not currently trending. Populism is rough terrain for policy wonks and for classical liberals alike. Which is how I’ve ended up here, writing on my philosopher and literary colleagues’ blog, trying to explain not just what I think, but how it feels to watch the ground shift under a set of assumptions I once took for granted.

Here’s one shift I’ve noticed: many Big‑L Liberals—the coalition, the party ecosystem, the people who used to treat “markets” as a suspicious word—have moved toward the middle on a set of practical questions. The “abundance” argument is the cleanest example. Build more housing. Streamline permitting. Let energy projects happen. Fix choke points that keep supply from expanding. A decade ago, some of these ideas were coded as right-of-center. Today, a lot of them are being pushed by people who would never have called themselves free-market advocates.

That should be good news for anyone who has been arguing, for years, that scarcity is often a political choice. It also raises an uncomfortable question for my small-c conservative friends—the people I usually agree with on markets and growth: when parts of the left move toward the center on an abundance agenda, are they meeting that shift with the same openness? Or are they treating politics as a package deal—accept the label or reject the person?

It’s hard to dodge that question in the reaction to Zohran Mamdani, New York City’s newly elected leader. Mamdani ran with a social-democratic identity, and much of the early commentary treated that identity as an existential threat to the city’s economic future. I get the instinct. The first campaign video I remember seeing from him was about a rent freeze. My economist brain did what it always does with price controls: it panicked.

New York is the canonical case study for why rent control backfires. It is taught in undergraduate microeconomics courses precisely because it produced such visible harm, from housing shortages to abandonment and arson in the Bronx. Spatial and temporal myopia are real, and there will never be too many articles or lectures explaining why price controls reliably backfire.

But a second thing also happened when I paid attention. Mixed in with the socialist branding and the genuinely bad ideas were policies that were plainly pro-growth and market-friendly. And my instinct wasn’t to defend the bad ideas. It was to separate them from the good ones—publicly—because reflexive opposition teaches the wrong lesson.

Politics is a repeated interaction. If you want politicians to keep doing the useful things, you have to show that useful things earn support. Constant vetoes don’t build trust. Selective agreement does. It tells people on the other side: “When you move toward better policy, we’ll meet you there.”

I assumed that instinct would be broadly shared among people who also say they want a bigger pro-growth coalition. Instead, I found myself oddly alone—at least publicly—in pointing to the constructive parts of Mamdani’s agenda. In my network, the socialist label ended the conversation before the policy details even got a hearing.

That bundling reflex is exactly the habit we claim to oppose. We complain—often correctly—that progressive politics treats “markets” as a moral stain rather than a tool. But if we respond by treating “socialist” as a conversation-stopper, we’re doing the same thing in reverse. And if this is a moment that calls for coalition-building, it can’t be a coalition where only one side is allowed to cross the bridge.

To be fair, the fear driving this reaction is not imaginary. I recognize it in myself. I grew up in a developing country and studied economics as an undergraduate at George Mason University, then went to graduate school in Germany. In that combination of life experience, the failures of collectivism were not abstract.

So when I hear elite Western students speak fondly about the “warmth of collectivism,” I flinch. Not because they are callous—many are deeply empathetic—but because they have little exposure to how these systems actually functioned when put into practice. The ignorance is both spatial and temporal. Recently, in a café in Bonn, I overheard students say, “Capitalism is the only thing we know. We haven’t really tried anything else.” Germany tried something else within living memory. So did Russia, China, Cuba, and Venezuela. If those examples feel overused, we can talk about Bangladesh in the years leading up to the 1974 famine, or Syria under Ba’athist socialism. The costs of collectivism and central planning recur across contexts, differing in form but not in consequence.

From that perspective, the warnings about rent control and state overreach are not wrong. They’re necessary. Historical amnesia is real, and part of the job is saying, over and over, “We have seen this movie.”

But it’s not the whole job.

Because I also recognize a different reaction in myself: genuine excitement when I see a politician quite literally cutting red tape—scissors in hand—while explaining how they’re reducing regulation to speed up housing construction. And in the weeks after Mamdani’s win his office announced several moves in that vein. They weren’t typical for someone running on socialist vibes. But they also weren’t hidden. He talked about them during the campaign.

What surprised me wasn’t that he followed through. It was how little credit that follow-through earned from people who claim to share the same policy goals.

Below is a simplified list of the kinds of changes his team has highlighted. I’m not doing full urban economics here. I’m making a narrower point about how we respond.

The red items: policies that work like central planning.
A rent freeze is the classic case. Set the price below market and you don’t get a gentler housing market; you get less maintenance, less investment, and a slow-motion decline in the quality and quantity of housing.

Near-free childcare has a similar shape. Set the price at (or near) zero and demand spikes. If supply can’t expand quickly—more slots, more staff, more facilities—scarcity reappears as waitlists, rationing, or quality problems. The bill doesn’t disappear. It moves to taxpayers, and the politics harden because reversing “free” is harder than never promising it.

Fare-free buses follow the same pattern. Eliminating fares creates a large budget hole while increasing demand. Without new revenue, the system absorbs that gap through taxes, service cuts elsewhere, or crowding. Again: the price signal goes away, but scarcity doesn’t.

None of this is obscure. You can critique these ideas loudly and still be right.

The green items: policies that utilize market mechanisms.
Streamlining housing approvals—especially efforts that cut delays, veto points, and discretionary holds—targets the actual driver of high rents: constrained supply. When it’s slow, uncertain, and expensive to build, you get fewer units and higher prices. Speeding the process up is boring, but it’s how you get more housing.

Reducing startup fees and civil penalties is similarly pro-market. Those are fixed costs: barriers that hit new entrants hardest and protect incumbents by making competition expensive. Lower them and you get more experimentation, more small firms, and more pressure on prices.

I’ve run through these quickly because this isn’t an analytical piece about optimal policy design. The point is simpler: why, among small-c conservatives and self-described free-market types, do I see an outpouring of commentary over a stray line praising the “warmth of collectivism,” and so little interest in reforms that would actually make the city richer and more livable?

You should criticize the rhetoric. You should criticize the red policies. I’m asking for symmetry—and for the discipline to react to policy with at least as much intensity as we react to slogans.

Because if we can’t acknowledge good policy when it comes from someone we disagree with, persuasion has already failed. Not persuasion of voters in some abstract sense—persuasion of the politicians themselves. If the only feedback a politician ever gets from the pro-market world is contempt, the rational response is to ignore that world entirely. You’re training them to cater to their base and treat you as unreachable.

Coalition politics requires something harder than opposition: discernment. You don’t have to stop being worried about collectivism to say “yes” to permitting reform.

Big‑L Liberals have been moving toward market-oriented ideas in the name of abundance and growth. If we want that shift to continue—and if we want it to show up in policy rather than just rhetoric—small‑l liberals and small‑c conservatives have to prove the bridge runs both ways. Right now, we need that big tent more than ever.

I’ve come to believe that over-delegation is one of the most under-discussed moral phenomena of our times.

My favourite example of over-delegation is the case of the British person who, on overhearing a sexist joke in the pub, pops outside to phone the police. You think I’m making this up! I’ll write more on the UK’s disastrous free-speech culture some other time. But the point I want to make today is a simple one, about the risks of shirking responsibility. About why it’s bad to delegate the moral obligations that you’re supposed to meet.

The pub example is useful here because it forces us to recognise the importance of moral authority. It reminds us that you can believe that someone should indeed intervene in the sexist joke-maker’s behaviour — by telling them they’re being a bit of a dick, for instance — without believing it’s the police that should do this.1 Of course, the ‘someone’ who should intervene here may well not be the person who just happened to overhear the joke. More likely, it’s the person who the joke was directed at. Or perhaps, in this particular instance, there’s no third party whose business it is to intervene, and it’s all on the joke-teller to reflect on whether they should behave differently next time.

But the point I’m making is that there’s a question of authority to be addressed. Even when intervention is morally permissible — even when intervention is morally required — this doesn’t mean that any old person can, or should, intervene! The same applies to the kind of intervention that’s warranted. I mean, just because you’re sure that your friend is about to run a stop light doesn’t justify you chopping off their hands to prevent this happening, does it?

The sexist joke example is also useful, however, because it points to the particular problem of the state getting involved when it isn’t its business. One serious cost of such a thing is that the law is blunt. The law is famously bad at parsing and evaluating complex moral matters. Often, this is sufficient reason to counsel against the involvement of the law. But another, underrated, reason for concern about over-delegation to the law relates to its cost to the individual — and to society, over time.

This cost arises because failing to meet your moral obligations doesn’t simply put you at risk of serious wrongdoing — although clearly it does.2 It’s also bad for you, in itself. We are reasoning creatures, with the capacity to evaluate and determine how we should behave. But we have to keep practising doing so! And I mean ‘practising’ here both in the sense of actually doing the thing, and in the sense of doing the thing with a view to getting better at it.

This is quite an Aristotelian way of talking. And the Aristotelians might well associate my concerns about moral over-delegation with their concerns about ‘crowding out virtue’. Think here about when you’re at the train station and you see a sign telling you to be kind to your fellow passengers. Such signs are usually well-intentioned, I’m sure. But the Aristotelians would remind us how important it is that we do the hard work, ourselves, of working out how to behave well towards other people. It’s important that we attend to such things! We shouldn’t need the state to remind us to be good, and when it takes on this role for us, our moral reasoning can become lazy.

Of course, this doesn’t mean that we shouldn’t ever help each other in these matters. Help is different from delegation, and delegation isn’t always wrong, anyway. It’s over-delegation I have a problem with, remember!

A great example of an appropriate kind of moral delegation relates to the obligation that all adults have to help children to learn how to use their reasoning skills. It’s good and right for adults to shoulder the burden of decision-making — including moral decision-making — when children aren’t capable of doing so. That said, we probably shouldn’t call this particular example ‘delegation’, because delegation implies intention, and it’s not really as if children are asking adults to shoulder the burden here! Indeed, part of what’s going on is that children don’t know when or how to ask for help with this kind of thing.

The children example does show us, however, that even when a lot of people share some specific kind of moral obligation, this doesn’t necessarily mean that all of those obligations are identical. Parents, for instance, have a very particular set of obligations to help children to develop their reasoning powers. And particular parents have particular obligations to their particular children!

The children example also helps to show us that it’s not always the person who can do the job the most efficiently — or even the most effectively — who’s the best person to do it. Again, this comes back to the idea of holding the right kind of authority. I mean, sure, the state actor has the power of guns and laws, but the use of such things is clearly not the right way to get little Jimmy to stop teasing his sister!

Acting morally is hard. It’s hard at the level of doing the right thing: when you’re the person in the pub who’s supposed to be telling your friend that they’re being a bit of a dick, then of course you want to look for a way out. Of course you want to just be able to laugh at the joke.3 So, you tell yourself that you’ll say something next time. Or you ask someone else to have a word. Or yes, you pop outside and take out your phone!

Then again, sometimes it won’t be right for you to intervene. And sometimes it won’t be clear whether you should or not. Because acting morally is hard even at the level of working out who’s supposed to be doing what. Nonetheless, there are some really easy cases of over-delegation to spot. Examples like dialling the police outside of the pub. Like bringing in the army to trample the campus protesters. Like getting your friend to break up with your boyfriend for you. Like threatening little Jimmy with a bazooka.

As soon as you start thinking about the over-delegation problem, you’ll see it everywhere.

This is supposed to be part of a larger set of measures to improve road safety. But really it’s a way of avoiding the real problem: the million test backlog.

In the last few years, the waiting time to take a driving test in the UK has gone up to about six months. The Guardian reported in December:

About 70% of driving test centres have every slot fully booked for the maximum 24 available weeks ahead.

This causes obvious problems—young people are already facing historically high house prices thanks to restricted supply, now they can’t even get a driving licence.

Many UK towns and cities lack adequate public transport too. A city the size of Leeds or Bristol really ought to have a tram, and in most European countries would do so. London has excellent public transport, but such high house prices that many people cannot afford to move there.1

And why does it take so long to get a driving test? Undersupply.

There aren’t enough examiners. The DVSA has only hired a net of 83 more examiners since 2021. Their pay is too low. On top of that, the Driver & Vehicle Standards Agency loses money on every test. A large secondary market has opened up where test slots are resold at eight times their original value.

All of this is a glaring case of a broken market. Allow prices to rise and more examiners will be able to administer more tests. The alternative is to continue the delays.

The DVLA expects the 24 week wait time to persist into 2027. According to End the Backlog, ’s campaign to improve this situation, it will take five years to get through the outstanding one million driving tests. As they say, it is likely to be worthwhile for the government to spend money to reduce it faster as so many jobs require a driving licence.

Worst of all, according to the DVLA’s own data, the pass rate for a driving test is just under half. So most of the people trying to get licences then have to wait for another six months.

In contrast, here in Arlington, I walked into the DMV one day and took my theory and practical test. I didn’t need an appointment. It was all done within an hour. I do think the standard of driving is worse in Virginia than in England. But British roads are really quite safe. Casualties have fallen by 90% since 1960 and halved since 2006. Whitehall doesn’t need to consult about making the roads safer: it needs to solve the problem that means a million young people will have to wait six to twelve months to get a licence. If not longer.

The real reason the government is going to make it official policy to wait for six months between the theory and practical test is that it is much easier to make a law that says the problem isn’t a problem than it is to actually fix the underlying issues. End the Backlog has a plan to solve the problem.

Instead sounds depressingly like the British government is going to act like the Circumlocution Office.

Whatever was required to be done, the Circumlocution Office was beforehand with all the public departments in the art of perceiving—HOW NOT TO DO IT.

A British man has been sent to jail for 18 months because of two tweets that were said to be inciting racial hatred. The prosecution argued that his tweets had the “potential to trigger discord.”

What Luke Yarwood said was awful. No doubt about that. Calling for migrant hotels to be burned down and people to be slaughtered in the street is abhorrent.

But Yarwood should not have got a prison sentence.

Some people compare tweets like this to inciting a mob. J.S. Mill himself makes this exception to free speech.

An opinion that corn-dealers are starvers of the poor, or that private property is robbery, ought to be unmolested when simply circulated through the press, but may justly incur punishment when delivered orally to an excited mob assembled before the house of a corn-dealer, or when handed about among the same mob in the form of a placard.

People who read tweets are not assembled together, not physically present outside migrant hotels, and are not being addressed as a unit. Twitter is akin to the press here, not an oral declamation outside the hotel. There was no mob in Bournemouth when Yarwood tweeted. No crowd was gathered at the hotels. (The events that prompted some of his tweets were in Germany.)

In the debate about this story, people have reshared an image of the original tweet defending the prosecution. Clearly they feel that it was safe for them to publish those words. But if you think the tweet is an incitement akin to standing in front of the mob, you would surely object to this wider sharing. Yarwood’s tweets were seen by 33 people.

One’s intent at sharing the image makes no difference if the wrong person reads the “inciting” words. If the inciting incident can be in another country from the tweeter, then the test for potential harm is quite wide.

There are ongoing tensions in Britain about migrants. How clear is it that the tweet is safe to share now but wasn’t then? Unless the threat is credible and proximate, Mill’s conditions are not met. Potential harm is too a wide criteria.

This sort of injustice is becoming normal in the UK.

When it comes to free speech, Britain is living through a sustained institutional failure. The number of people being arrested for speech offences is astonishing. Earlier this year, The Times reported:

Custody data obtained by The Times shows that officers are making about 12,000 arrests a year under section 127 of the Communications Act 2003 and section 1 of the Malicious Communications Act 1988.

The acts make it illegal to cause distress by sending “grossly offensive” messages or sharing content of an “indecent, obscene or menacing character” on an electronic communications network.

Officers from 37 police forces made 12,183 arrests in 2023, the equivalent of about 33 per day. This marks an almost 58 per cent rise in arrests since before the pandemic. In 2019, forces logged 7,734 detentions.

Policy Exchange estimates that the police spend more than 60,000 hours per year dealing with “non-crime hate-incidents”.

There are plenty more stories of how the UK is slipping into this illiberalism.

In January, six officers turned up to arrest a married couple because they made complaints about their children’s school in WhatsApp. The constabulary later had to pay £20,000 in damages.

More recently, eleven police officers are reported to have entered Elizabeth Kinney’s home while she was in the bath and arrested her. You might expect that such dramatic action would be in response to a serious crime of some sort.

But no.

Kinney had used the word “fa**ot” in a text message. Her crime was sending an “offensive, indecent, obscene or menacing” message.

For this, she was given a twelve-month community order: 72 hours of unpaid work and 10 rehabilitation activity days. She is also ordered to pay £364 in costs. This is an extraordinary way for the law to treat a single mother for the sake of a text message reported as an act of spite by a feuding friend.

And the laws are creeping into more mundane, everyday areas too.

Other speech restrictions are coming into force in the UK. Stories are coming out regularly about the effect of the new Online Safety Act, which means internet users have to verify their age before they can read something like Ed West’s history Substack. Online fiction which contains sex or violence also requires age verification. As do columns by the Free Press. I had to have my picture taken recently to be allowed to read my own direct messages in the Substack App.

If Substack doesn’t comply with these rules, it can be fined 10% of its qualifying worldwide revenue, even though it is not a British company.1 What happened to Luke Yarwood is part of a broader pattern of illiberalism.

Britain didn’t plan any of this.

It got here slowly, decision by decision. Legislation that sounds sensible at the time is later put to illiberal purposes. Each case finds people to justify the sentence. Calling for migrants to be slaughtered in the streets is a line many people don’t think should be crossed.

But the total effect will gradually become overwhelming. Injustice becomes normalise by every exception. And it erodes public trust in the law. While thousands are arrested for speech offences, in the year to June 2024, 1.9 million violent crimes were closed without being solved. That is 89% of offences. So while hundreds of thousands of violent criminals walk free, thirty people are arrested every day for speech offences.

And illiberal laws creep into ordinary life in sinister ways. The man with 18 months jail time and the single mother who sent offensive texts were both reported to the police by someone they were in a personal feud with. In one case a brother-in-law, in another case a friend.

This is what illiberal laws do. They pit friends and relatives against each other. The police won’t retrieve your stolen phone. They won’t solve millions of violent crimes. But they will send several officers to your house because your friend dobbed you in for using hate speech in a private message.

One common defence of Yarwood’s imprisonment is that he called for the killing of MPs, for the country to be taken over by force. It was once a principle of English law that words alone could be no treason. To be guilty, there had to be evidence of plotting and conspiracy—not mere words thrown out in heat. If we think—truly think—that everyone who declaims that MPs are useless and ought to be hanged should be put in jail, then there will be a lot of awkward moments at family Christmases this year.

Make Britain Liberal Again

Britain was once a bastion of free speech. It was a country that was able to tolerate nasty words that didn’t lead to nasty actions. Now that is slipping away we must be careful about what we support. It is easy to say “Do you think it’s OK to say that” or “Do you think it is OK to incite violence?” Easy to say that someone else should be locked up.

But if Britain keeps going like this, it will not like the future it makes. More and more people will use these laws to get revenge on friends and relatives. More and more people who have worse impulse control than others will suffer extreme consequences for petty acts. More and more people will think twice before they say something.

Illiberalism is starting to creep across England like a frost. Let’s hope we do not have to wait too long for the thaw, and that we don’t create too much harm in the meantime.

How often do you think about this?

The more you do think about it, the more you realise how extreme it is.

It’s also extreme in a historical context. Over the past century, the number of people in prison has increased by more than four times in the UK, and more than ten times in America, with a particular increase since the 1990s in the UK, and since the 1970s in the USA.2 And beyond numbers, locking people up has only really existed as a response to crime — anywhere — for about 200 years!

This isn’t to deny that other illiberal responses to crime were practiced previously. It’s also not to deny the further illiberal concerns we should attend to when thinking about current and historical imprisonment practices — not least the way in which, in the USA particularly, such practices have often been driven by racism.

But again, the more you think about imprisonment as a practice, per se, the more you realise how extreme it is.

Shut up in a little room for most of each day. Told exactly where to go, and often exactly what to do, when you’re finally allowed to spend time in the other tight spaces down the corridor. Kept in line by the persistent threat of even more intimate kinds of constraint. Yelling! Bright lights! Hard surfaces! Chains! Batons! Hands!

And these are just the basic expectations. I mean, all of this is before we get on to the terrible things that aren’t supposed to happen in prisons, but obviously do.

This is the life being led by millions of people in our supposedly advanced modern societies — and we should think about it more.3 Particularly those of us who are committed to liberalism.

Minimally, liberals should be concerned about imprisonment because it offends against the following three core liberal values: free agency, wellbeing, and equality of moral status.

Imprisonment violates free agency

First, liberals should be concerned about imprisonment because locking someone up violates their free agency.

I’ve written on Substack before about how free agency — in the sense of being able to act on your reasoned decisions — is both a distinctive quality of being human, and a morally significant one. Free agency is morally significant because it affords us responsibility as well as power.

Free agency is such a central part of being human that it’s easy to argue that, generally, it’s bad when humans are prevented from exercising their free agency. And that if one human deliberately interferes with another human’s free agency, then generally that’s both bad and wrong.

This doesn’t mean that all instances of interference in free agency are equally bad and wrong. Or even that they’re all impermissible — otherwise, we’d have to avoid anaesthesia!

Rather, I want to emphasise that the way in which imprisonment is aimed at comprehensively constraining this thing of deep human value presents a serious moral problem. And that we should acknowledge that the comprehensive constraint of imprisonment lies not only in its nature — in the persistence of its controlling imposition — but in many of its effects.

Think, for instance, about how prisoners can become institutionalised: how they often struggle to return to normal life outside the top-down directions of the prison estate. Think also about how difficult it can be for them to become re-accepted as full members of society. And how some even develop unhealthy attachments to their captors. All of these effects of imprisonment degrade the capacity for free agency.

Imprisonment is damaging to wellbeing

Second, liberals should be concerned about imprisonment because locking someone up is damaging to their wellbeing.

Not all prisoners suffer direct harmful violence, like forcible drugging, shackling, or being made to undertake hard labour. Nonetheless, I’d argue that all people who are imprisoned suffer some physical and psychological harm as a result.

At heart, this argument trades on the importance of free agency. If free agency is a matter of bodily capacity as well as mental capacity, then comprehensive constraints on free agency can count as instances of both physical and psychological harm.

Beyond this, however, are concerns about the many bad things that prisoners routinely experience that you might not want to count as instances of harm. These range from being made fun of, to lacking access to outdoor space. They are found in the denial of a person’s access to the goods and services that are necessary to human physical and mental wellbeing — something that goes beyond the avoidance of harm.

Imprisonment breaches equality of moral status

Third, liberals should be concerned about imprisonment because locking someone up breaches their equality of moral status.

By ‘equality of moral status’, I’m referring to two foundational liberal ideas. First, the idea that all human beings, regardless of circumstance, share a fundamental equal status, as a matter of moral truth. This kind of equality is grounded in features particular to being human, including the capacity for free agency.

Second is the idea that there’s a different but overlapping basic moral status held by all members of a legitimate political society. This second idea depends on the first, but comes with additional demands to behave towards each other in certain ways. It overlaps with the discussions of ‘relational equality’ you can find in the work of Elizabeth Anderson.

Imprisonment offends against these two foundational liberal ideas in part, again, because of the way in which imprisonment constrains free agency. But also because it’s hard to square imprisonment with the culture of non-domination required in political societies that are committed to respecting this equality of basic moral status.

Of course, crime is also bad

At this point, however, I want to stop and acknowledge that it’s not just imprisonment that’s bad. Crime is also bad!

There’s a chapter I like by Jonathan Wolff, in which he focuses on arguing that crime is bad.4 Now, on some level, it’s funny that Wolff felt the need to write this chapter, because it’s really very obvious that crime is bad. But unfortunately, his chapter is also much needed. This is not least because some of the people who make the loudest arguments against current responses to crime also sadly underplay the badness of crime itself. This is a serious mistake.

Wolff tells us that crime is bad not only because the victims of crime suffer physical harms such as broken arms, and psychological harms such as debilitating fear, but also because crime, as a representation of contempt, brings about damaging inequalities in societal standing. He’s clearly right. Crime is bad for its victims, its perpetrators, the wider public, and in itself.

Good societies don’t put up with crime. And societies that fail to attend to violent crime are particularly negligent.

When is imprisonment necessary?

Most people believe that imprisonment is necessary for preventing at least some violent crimes. Let’s consider an easy example. Imagine Mr X, who openly and reliably admits his desire to continue his ongoing campaign of killing as many people as possible.

Now, if we take imprisonment off the table, then our options for preventing Mr X are limited. Indeed, it seems likely that other fully preventative options — such as killing Mr X, or chemically changing his desires — will count as even more problematic than imprisonment, when assessed against liberal principles.

I’ll discuss some other alternatives briefly below. But for now, we seem to face the conflict that not only: 1) imprisonment is generally bad and wrong on a liberal account; but also 2) at least some instances of imprisonment are required in a liberal society. And we’re left where we began.

Or are we?

Well, you can surely accept both of the premises above — that imprisonment is generally bad and wrong, and also that at least some instances of imprisonment are required — without concluding that millions of people need to be locked up!

I mean, what percentage of those millions pose a clear violent threat? I’m sure there are excellent studies on this, but even if the answer were as high as 90 per cent, then that would still leave hundreds of thousands of non-violent people unnecessarily locked up.

My guess and hope is that our descendants will look back in horror, and struggle to believe that we really imprisoned all these non-violent people. And that’s before we consider the human-rights-violating standards of many contemporary prisons.

What about punishment?

But perhaps you’re annoyed that I haven’t talked about punishment yet. Perhaps you think punishment is the elephant in the room — that there’s a simple reason why so many people are in prison, and this reason is punishment!

Punishment is a hard topic for liberals, however.

What I mean by this is that people who are particularly committed to the value of freedom often feel discomfort — and, I think, should feel discomfort! — about the idea of punishment, itself. I’ll write about this in more detail some other time. But, broadly, this discomfort is related to the way in which justifications for punishment typically depend on either retribution or deterrence.

Retribution arguments tell us things like: ‘punishing someone in order to wreak vengeance on them for committing a crime is permissible’ and ‘you can rely on vengeance as a justification for punishment’. These arguments often make liberals anxious, not least because they depend on the idea that it can be a good thing — the right thing, even — to intentionally do something bad to someone who’s under your control.

Meanwhile, deterrence arguments tell us things like: ‘punishing someone in order to deter further crime is permissible’ and ‘you can rely on deterrence as a justification for punishment’.5 These arguments should make liberals anxious, too. I mean, okay, these arguments might work if you think that only consequences matter when evaluating right and wrong. But, as I wrote here recently, consequentialist reasoning leaves liberal principles and values at risk. And punishing wrongdoers in order to try to prevent future crime involves treating them as a means to an end.

Of course, these are simplistic descriptions of the arguments at hand. And many liberals find ways to reconcile the problem of punishment. But now I’m going to propose a different way forward, at least in the case of imprisonment.

A proposal for defensive imprisonment

My proposal takes punishment out of the equation. That is, if we only locked up the people who posed a clear violent threat, then we wouldn’t have to rely on punitive justifications at all. Rather, on such an approach, we could see prison not as punishment, but as defence.

Now, I accept that this might seem like a conceptually confusing move — mainly because many people take punishment to be part of what prison does, by definition. Tommie Shelby, for instance, tells us that prison is an “incarceration facility that functions to impose punishment”.6 Indeed, throughout his recent comprehensive book on prison reform and abolition, Shelby accepts that prison is punitive, and that this is non-controversial.

That’s fine! If you want to keep a punitive element within the concept of prison, then all you have to do is imagine a second kind of institution called ‘shmison’. Shmison is just like prison, except it’s not a place for punishment. Nobody gets sent to shmison for punitive reasons! People are only locked up in shmison for defensive reasons.

If this seems weird, then remember that people are already often locked up for non-punitive reasons. Consider here the differences between being locked up because you are: 1) infectious with a dangerous new contagious disease; 2) an immediate danger to yourself; 3) an immediate danger to other people; and 4) have committed a crime. Even if these experiences looked and felt very similar, the differences in reasoning would persist.

Now, let’s return to shmison, where people are locked up only if they pose a danger to the public.

Of course, it wouldn’t always be easy to work out if someone posed a sufficient danger to be admitted to shmison. And, as is currently the case, some of those dangerous people would be better suited to psychiatric hospitals. But there will be many easy cases that we can rule out immediately. People who are currently in prison for having committed non-violent crimes, and who have made no violent threats throughout their time within the criminal justice system, count as easy cases, for instance.

Those people have no place in shmison!

Two big payoffs

Beyond the moral imperative for putting an end to imprisoning non-violent criminals, there are two big payoffs that would come from imprisoning people only on the grounds of defence.

First, doing so would seriously reduce the prison population. This would allow for a reduction in prison spending, and a reallocation of prison resources. The reduction in prison spending could either enable a reduction in taxes, or increased spend on alternative justice measures. And prison resources — including land, buildings, and equipment — could be clearly used for many other purposes.

Second, using prison for defence rather than punishment would lead to improved conditions for prisoners. I don’t really mean by this that decreasing the prison population would make prisons less crowded, improve the ratio of guards to prisoners, and enable increased spend per prisoner. Sure, those things would likely happen, at least at first. But resource reallocation might change this over time.

Rather, I want to emphasise that removing punitive reasoning from the equation gives reason in itself to improve prison conditions.

I mean, think about how, whenever there’s a news story about a prison that has unusually good conditions, someone will respond, “But hey, prisoners aren’t supposed to have nice things! They’re supposed to be in prison to be punished!” This kind of response aligns with both the retribution and the deterrence accounts of punishment. On the former, prisoners deserve bad conditions. And on the latter, bad conditions provide a disincentive to commit crime.

But if prison — sorry, shmison — is no longer punitive, then those reasons no longer stand, and we can focus on ensuring good conditions for the inshmisoned. Shmison, indeed, should be a place in which not only the rights of the shmisoners are respected and their basic needs are met, but all the fundamental goods of human life can be accessed, from knowledge to work to love.

A few objections

Now, there are some obvious objections you might make to my shmisons proposal. For a start, perhaps you simply don’t share my moral concern with prison as punishment. So you want to continue pushing the position that we need prisons in order to wreak vengeance, or to deter future crime. Okay! I will continue to try to dissuade you.

Or perhaps you’re concerned with feasibility and practical upshots. In particular, maybe you’re worried about the effects of suddenly releasing a large proportion of the prison population. Where would these people go? What would they do? Well, these are questions that already exist in relation to current prisoner-release procedures — questions that often aren’t answered adequately. Sadly, some former prisoners end up homeless, for instance. But that’s hardly a reason to keep them locked up!

Moreover, if the main concern you have with the idea of suddenly releasing a large number of prisoners is less to do with their needs, and more to do with the risk they might pose to members of the public, then remember the condition on which these particular prisoners would be released. They must not pose a clear violent threat!

Perhaps, however, you think that these former prisoners would pose some other threat. Perhaps you think they would continue to swindle people, or steal things, or do whatever it was they were locked up for doing in the first place. And yes, perhaps these released prisoners would indeed pose those threats. But if prison were reserved for people who posed a violent threat, then we would have to find other ways to deal with non-violent threats! And this could include, if you wanted, other means of punishment — which, again, I’m not ruling out, for now.

Rather, all I’m arguing is that there are many people who are currently in prison who would pose exactly the same physical risk to the public, if they were outside of prison — and that therefore these people should be released.

That said, if releasing these non-violent people really were too much for you, then of course we could start instead by refraining from locking up any more of them. And if even that felt like too much, then I suppose we could expand our focus beyond violent threat. I mean, sure, keep the swindlers locked up, if you really can’t find another way to deal with them! Why not put them in the cells alongside the granny who sends offensive tweets, if you really think she also needs the same constraints as Mr X.

Just don’t call yourself a liberal!

Whatever details you insist upon, however, it seems undeniable that shifting to a model on which we only imprisoned people who posed a clear threat, would lead to the imprisonment of thousands — if not hundreds of thousands — of fewer people.

I want to end by emphasising that obviously we should also think much harder about alternatives to prison. That is, even if we accept that some people do need to be kept away from the public for defensive reasons, then we should ask whether this must be done by collectively housing them in a panopticon.

Solutions to this problem that involve emerging technologies should be considered alongside classic alternatives. These solutions range from banishment, to house arrest, to various kinds of invasive surveillance, to combinations of the above, and things we haven’t conceived of yet. Of course, as already mentioned, some of these solutions could prove more problematic than prison on liberal grounds — particularly when prison is shmison!

But again, we should think much harder.

A simple starting point

It’s often assumed that extreme problems require radical responses — responses that not only enable structural changes to the status quo, but also require a vast amount of resource commitment, alongside the serious shifting of societal norms.

What I’ve tried to persuade you here, however, is that there’s something that could be done to address the extreme problem of imprisonment that not only doesn’t require too much shifting of social norms — it would actually free up valuable resources! Shmison makes sense and shmison comes cheap.

Of course, it’s not these kinds of considerations that should primarily lead us to releasing non-threatening prisoners. It’s the moral horror of the fact that they’re currently locked up.

People who care about freedom should lead the charge on changing this.

----

Thanks to GPT for the panopticon leviathan picture.

[Transcript coming soon!]

Thanks to GPT for the picture!

[Transcript coming soon!]

This comment, however, prompted Henry Hill, deputy editor of Conservative Home, perhaps the publication most in touch with England’s Conservative grass-roots, to say: “It’s weird how near-explicit “we need serfs” is progressive-coded now.”

This is horrible framing, and wrong-headed to boot. Immigrants who take jobs in a capitalistic democracy are not serfs. And Polanksi didn’t say anything approximating to “we need serfs”. He said, essentially, “we need care workers, they should come from other countries”.

However crass Polanksi was, and whatever it reveals about his attitudes to immigrants, it isn’t sensible to use a word that means “a worker who is legally obliged to work a piece of land under the dominion of a feudal lord” to mean “low paid worker in care job”.

This is like the left characterising everything the right says as fascist.

Henry Hill surely doesn’t mean to deny the importance of markets, the division of labour, or the inherent dignity of every individual whatever their work. But in his pursuit of an effective immigration rhetoric, he mirrored Polanksi’s “us and them” framing.

It’s fine to oppose immigration. I happen to think immigration is a good thing, both for the immigrants and for England, and I think the evidence supports that view. But as the right debates this issue, it would be better to avoid this sort of framing.

Maintaining a language of free individuals is essential for liberalism today, whether you are in favour of immigration or not. Just because the Green Party leader says something demeaning doesn’t mean we need to follow-suit.

This set him firmly apart from his contemporaries like Harold Pinter. The Royal Court had an unofficial policy to not perform his plays.

Stoppard once, famously, said he was in favour of “Western liberal democracy, favouring an intellectual élite and a progressive middle class and based on a moral order derived from Christian absolutes.”

What can we make of this apparent contradiction? Was Stoppard a political writer or wasn’t he?

I propose that Stoppard was a liberal. He was not a liberal writer in the sense that he polemicized for party or platform. But he was a dedicated liberal individualist—his work is suffused with the liberal idea that the individual is paramount and must be respected, protected, and encouraged.

Writing to the theatre critic Kenneth Tynan in the 1970s, discussing how he was going to stage a debate about moral philosophy in a play (Jumpers) that was going to be funny, Stoppard said: “it is a mistake to assume that plays are the end-products of ideas (which would be limiting): the ideas are the end-products of the plays.” Stoppard, that is, thinks as a playwright; he does not write plays as a philosophical thinker.

Stoppard’s plays are especially playful and his humor is his first essential liberal commitment. What will get you killed or locked up in an authoritarian country can be mocked in a liberal one. Liberalism allows the individual to laugh at the state. Liberalism allows free play.

At the end of Every Good Boy Deserves Favour, Stoppard’s 1977 television play about Soviet political prisoners, the “Colonel-Doctor” visits the mental hospital where two political prisoners are being held as “patients”. One is actually insane, and hears an orchestra wherever he goes; the other is deemed to be insane because he spoke out against the regime and refuses to state that criticizing the authorities is a sign of madness.

Both prisoners have the same name and the “Colonel-Doctor” gets them mixed up and asks them the wrong questions: the lunatic tells him it is insanity to criticize the regime and the sane man tells him he hears no orchestra. Thus they are both released.

It is a brief but splendidly funny scene. And it reinforces the basic message: you who are watching this play can laugh at Soviet totalitarianism: the ordinary people living under the regime cannot.

In Professional Foul, written at about the same time, Stoppard exhibited another crucial sense in which he is a liberal writer. He is not governed by ideology or theory. He believed in the moral obligation to individuals as individuals.

In Professional Foul, two philosophers attend a colloquium in Czechoslovakia. They discuss moral theories which allow them to wriggle out of basic moral commitments. The main character, who is the J.S. Mill Professor of Ethics at Cambridge, tells a former student of his—a Czech man, who works as a cleaner—that he cannot smuggle his dissertation back to England. That would be a breach of the “rules” of his visit to Soviet-controlled Czechoslovakia, which would count as an immoral act.

Later on, the Professor drops in to see his pupil and is detained by Soviet authorities. He experiences for himself the terror of being at the mercy of an authoritarian regime. Something he said earlier becomes newly relevant: “the essentials of a given situation speak for themselves and language is as capable of obscuring the truth as revealing it.”

Under the interrogatory gaze of Soviet officials, the moral philosopher has to face the facts: rules and manners are very different from morals and principles. Call it what you will—not smuggling the dissertation would be wrong; it would be a sort of acceptance of the communist repression. He cannot hide behind words and relativism.

Or as Jan (a Czech exile returning to Prague) says about England in Rock ‘n’ Roll, “A thousand years of knowing who you are gives a people a confidence in its judgement. Words mean what they have always meant.” From his early play Jumpers, which satirized moral relativism in the work of A.J. Ayer, to Rock ‘n’ Roll, a play about the “Prague Spring” in Czechoslovakia, Stoppard remained committed to the idea that right is right, wrong is wrong, and words mean what they mean.

So the professor decides to help. But, he is now being watched, and his bags will be searched at the airport. So, he hides the dissertation in his colleague’s bag, knowing that man is much less likely to be searched. And so it goes. He broke the rules in order to achieve a higher good.

This play was broadcast internationally and Stoppard was making a very clear point about oppressive regimes that imprison writers. His principles are liberal principles—free expression, individual rights, artistic integrity. The professor even delivers a short lecture about the importance of individual rights, which is so offensive to the Soviets that they set off the fire alarm to interrupt him.

But Stoppard is not taking sides in UK politics. These are Stoppardian plays, not political plays.

Stoppard was honest when he said he had no political agenda and honest when he said he was a “timid libertarian.” Stoppard has the liberal’s pre-political commitment to the individual.1

This liberalism is a particular literary commitment, too. Harold Bloom used to tell his students: “The only method is yourself.” He was warning them against the sort of literary criticism that started from ideological commitments, be they feminist, Marxist, post-modern, structuralist, deconstructionist, or anything else.

Bloom was expressing the liberal idea—that we are individuals, and it is as individuals that we form groups, hold beliefs, and make judgements. As Stoppard once said, our beliefs “owe their existence to individual acts between individuals, which themselves are derived from an individual’s intuitive sense of what is right and wrong.”

It is in this sense that Stoppard was a liberal writer and it was his abiding commitment to individuals—be they Soviet dissidents or English writers that made his ideas the end-products of his plays and gave him the capacity to surprise, entertain, and delight his audiences.