Latest / The Capitalism and Freedom in the Twenty-First Century Podcast / Harvard Law's Cass Sunstein on Nudges, Behavioral Economics, Constitutional Law, and Liberalism | Hoover Institution

Transcript

Harvard Law's Cass Sunstein on Nudges, Behavioral Economics, Constitutional Law, and Liberalism | Hoover Institution · The Capitalism and Freedom in the Twenty-First Century Podcast

0:00
[MUSIC]
0:09
>> Jon Hartley: This is the Capitalism and Freedom in the 21st Century podcast,
0:11
an official podcast with Hoover Institution Economic Policy Working Group
0:15
where we talk about economics, markets, and public policy.
0:18
I'm Jon Hartley, your host today.
0:20
My guest is Cass Sunstein,
0:21
who's one of the most cited legal scholars of the 20th and 21st centuries.
0:24
He's a longtime professor at the Harvard Law School, where He's been since 2008 and
0:28
previously was at the University of Chicago law School since 1981.
0:31
Cass also served as the administrator of OIRA, which is the Office of
0:35
Information Regulatory affairs during the Obama administration.
0:40
He's also a prolific author of over 50 books, including Nudge with Richard Thaler
0:45
and Noise A Flaw in Human Judgment with Daniel Kahneman and Oliver Siboney.
0:50
He's also the author of several famous law textbooks on constitutional law and
0:54
administrative law.
0:56
Welcome, Cass.
0:57
>> Cass Sunstein: Thank you.
0:57
Great pleasure to be here.
0:59
>> Jon Hartley: So I want to just start by getting into your early life.
1:02
Where did you grow up and how did you get interested in law and
1:05
behavioral economics?
1:07
>> Cass Sunstein: I grew up in Wabin, Massachusetts and
1:10
I was a literature major in college.
1:13
I got interested in law really my senior year in college, where I thought being
1:18
a graduate student in English literature would be extremely interesting, but
1:22
maybe too passive.
1:24
And law, I thought you could have lots of flexibility and be more in the world.
1:28
That's what my young self thought.
1:32
When I came to the University of Chicago a few years later in teaching,
1:36
I was surrounded by rational choice types with Gary Becker and
1:41
George Stigler leading the charge and Richard Posner and
1:45
Bill Landis at the law school being the formidable giants who were surrounded me.
1:50
I felt like I was 3 foot tall and they were all 9ft tall.
1:54
I was a little person trying to understand.
1:58
They were so high in the clouds I could barely hear their voices,
2:01
though they were booming.
2:03
But I was a literature major and I thought that their conception of
2:07
humanity was not the same as what I had learned from James Joyce and
2:12
William Shakespeare and William Butler Yates and John Keats.
2:16
And so I was skeptical, kind of cluelessly skeptical.
2:21
I learned something about behavioral economics from Jan Elster,
2:25
who was a Norwegian political scientist philosopher who came to the University of
2:30
Chicago.
2:31
He introduced me to notions about pre commitment strategies and
2:35
about adaptive preferences.
2:37
That kind of was an inroad into the behavioral stuff.
2:40
And I started writing material that drew on Thaler and Kahneman Tversky,
2:46
and it was, I think, primitive, was law professor primitive.
2:52
But Thaler came to the University of Chicago at a time
2:56
when I was immersed in behavioral stuff and that kind of Accelerated.
3:01
We became great friends.
3:02
That's kind of accelerated my focus on behavioral economics and
3:07
its relationship to law.
3:09
>> Jon Hartley: Fantastic, so I want to sort of, I guess, for
3:12
the moment just focus on behavioral economics.
3:15
We'll get to law and political philosophy later.
3:19
So you co-authored the seminal book, Nudge, cast on scene in 2008.
3:25
I remember it as a pretty, I think, seminal moment for behavioral economics.
3:32
And it really, I think, increased the popularity of behavioral economics.
3:36
I mean, at the same time you sort of had the global financial crisis was sort of
3:40
around the same time.
3:41
But I just remember, you know, that really the, you know, late 2000s and.
3:46
And 2010s was a very, very a popular time for behavioral economics.
3:51
He also wrote a book more recently, Noise, with Daniel Kahneman and Oliver Sibony.
3:57
Daniel Kahneman being the famous Kahneman Tversky.
4:03
I'm curious what you think about the state of behavioral economics is now
4:08
following what I think has been just as sort of, I see it, I think,
4:12
maybe a more challenging time for behavioral economics.
4:17
We've had some recent controversies, a lot of controversy,
4:21
particularly in the field of psychology.
4:24
A massive fraction of studies there aren't replicable.
4:27
I think there's been some very famous scandals within the behavioral
4:32
economics world, think Dan Arielli, Francesca Gino, Francesca.
4:38
You've been in several of their papers have been retracted,
4:43
following the uncovering from the folks at Data Colada.
4:47
And to me, I mean, it seems like behavioral economics maybe peaked
4:51
around the times of the Richard Thaler Nobel and
4:55
the Robert Shiller Nobel in sort of the mid 2010s.
4:58
And maybe it's been on a bit of a decline.
5:00
I mean, would you agree with that?
5:02
Tell me why either that's wrong or
5:04
where you think the state of behavioral economics is as we see it today in 2025.
5:10
>> Cass Sunstein: Okay, thanks for that.
5:11
I'll tell you a story and
5:12
then we'll maybe talk about something in the vicinity of data.
5:17
So the story is I had dinner not long ago with a behavioral economist.
5:22
And he said it's terrible.
5:25
And I said what's terrible?
5:27
And he said it's so terrible.
5:29
And I said what do you mean it's so terrible?
5:31
And he said the demand for people doing behavioral economics
5:37
is higher than it's ever been by far that people in
5:42
the private sector, Silicon Valley and Saudi Arabia,
5:47
in Germany, in France and Italy, are looking for
5:51
people in the private sector who are trained in behavioral economics.
5:58
Governments are keener than ever that we're seeing interest in hiring behavioral
6:04
economists from people who are running behavioral insights teams or nudge units.
6:10
The demand is off the charts and it's higher than ever.
6:15
And I said, why is it terrible?
6:17
That sounds amazing.
6:18
He said, there aren't enough people.
6:20
So the, the people are holding themselves out as behavioral economists or
6:25
behavioral scientists to get jobs, but they're not trained.
6:30
The field is flourishing with such, you know, kind of exuberance on the part
6:35
of the people who are interested in it, and there just aren't enough people ready.
6:40
And he said, you've got to get people at Harvard ready to go help
6:45
people who are thinking about smoking, secession or road safety or
6:50
how to get people to pay fines when they're not paying fines,
6:55
how to get people to take up programs, how to deal with occupational licensing.
7:01
So on the practitioner side, from government and
7:05
the private sector, behavioral economics is incalculably
7:10
better off than it was even six, seven years ago.
7:15
On the academic side, the same is true.
7:18
If you look at the leading economics journals,
7:21
if there isn't at least one behavioral economics paper in the current issue,
7:27
it's surprising if there isn't one the year before or the year after.
7:32
I'm sorry, the issue before or the issue after a year is too much.
7:36
There are too many papers in behavioral economics in the American Economic Review,
7:41
in the Economic Journal, and the Quarterly Journal of Economics.
7:45
And some of it is kind of normal science where loss aversion is used to
7:50
explain something, and loss aversion is identified in a context
7:55
that we didn't see something might be appearing on default.
7:59
Default rules and their power.
8:02
Benjamin Enki is doing work on complexity, which I find extremely interesting,
8:07
and the role of complexity.
8:10
There's a lot of work on cognitive scarcity.
8:14
This is something that I'm keenly interested in from Malathan and Shafir,
8:18
their work on scarcity and that has implications for
8:21
when behavioral findings will be especially present and when they might
8:26
fall apart because people are focusing on something and so they're not going to show
8:31
loss aversion because if they do, they're going to lose a lot of money in a hurry.
8:36
So that's very vivid.
8:39
John List, of course,
8:40
has done a ton of phenomenal papers in the domain of behavioral economics.
8:45
And if we don't see a good John List paper in any four month period,
8:49
that's a surprising four month period.
8:53
There's keen interest in both economics and
8:57
in other domains in manipulation and how to think about it.
9:03
Sometimes it goes by specific names like dark patterns, or
9:07
sometimes there's a specification of the exploitation of a behavioral bias.
9:13
This also connects with work on algorithms and machine learning.
9:17
So behavioral economics right now is both on the practice side and on
9:22
the theoretical side in a better position, I'd say, by far, than it's ever been.
9:28
There isn't explosive insight, period that we saw with early Kahneman and
9:34
Tversky or Thaler, but we see much more in terms of insights and
9:40
much more specificity and new insights coming that are often firmer,
9:46
they might be consistent with, but firmer than anything we've observed
9:52
before just because the techniques are better on replication.
9:59
Thaler has a new edition of the Winner's Curse coming out,
10:03
I think in September 2025.
10:07
And all of the findings replicate every single one.
10:11
So that's, I think his exposition challenge in this amazing
10:16
new edition of the Winner's Curse is not to be kind of crowing at
10:21
the replication of the core findings, but they all replicate.
10:26
And to say they replicate is one thing to say the magnitude or the conditions
10:31
under which, et cetera, like you can make the endowment effect go down or
10:36
disappear if people are really sad, as I recall, you don't see an endowment effect,
10:43
but the central findings of that defining book all replicate.
10:48
So within the field, by which I mean both practitioners and
10:53
people are doing real academic work.
10:57
I'm thinking of Alex Moss, whom I don't know, but whose work is very, very good.
11:02
The fraud of Gino is a little like you're playing baseball and
11:09
there's some player who took steroids.
11:13
I mean, baseball still pretty good sport and
11:17
shouldn't take steroids and the fraud has hurt horrific.
11:21
I, I don't even understand it.
11:22
I should say I don't understand the state of mind suggests my limits as a behavioral
11:27
scientist that I don't, I don't see how one could do that.
11:31
But that's clear with Gino and, and, but if you look at the,
11:36
the extent of the field, fraud is horrific, but very limited in number.
11:44
The replication challenge, as I'd, as I'd call it, I wouldn't call it a crisis.
11:48
What's happening in Gaza is a crisis.
11:51
Climate change, in my view, is a crisis.
11:54
The replication challenge, that's a better word than crisis,
11:57
is if things don't replicate, that's basically an opportunity for learning.
12:02
And the core findings replicate.
12:05
We're seeing more, you know, more material that tells us more about when default
12:11
rules will be powerful and exactly why you could say that a default rule won't be
12:16
powerful if people hate the thing they're defaulted into and they have freedom.
12:23
Findings don't replicate.
12:25
And that's, that's an advance to know that it's very important to keep at it to see,
12:31
you know, is what are the boundary conditions for the availability heuristic.
12:37
Stock trader probably isn't going to be vulnerable to the availability heuristic
12:41
where they're going to lose a lot of money.
12:44
>> Jon Hartley: So this is kind of, I guess, how I see the maybe evolution,
12:47
like behavioral economics.
12:48
You had like 20, 30 years ago, you had Kahneman, Tversky, Thaler,
12:53
they were sort of challenging traditional models.
12:56
They get into data, find some anomaly or some bias.
12:58
Aha, yes.
12:59
Some new behavioral biases are discovered.
13:02
And then you also had in the sort of two, maybe 2000,
13:06
2010s, a lot of work on behavioral finance and factors and
13:11
things like momentum or herding, Jeremy Stein's hurting theory paper.
13:18
You also had these things like Ernest Fear's theory about fairness around
13:21
that time.
13:23
I think how it's evolved over time is now I think researchers are trying
13:28
to find more of a unifying theory that connects these facts.
13:33
So things like I would say the complexity work, I would maybe put under that.
13:41
I guess maybe more on the empirical side.
13:45
I don't know if it's fair to say that, you know, RCTs and experiments,
13:49
I think it's fair to say that they've certainly been the backbone,
13:52
a backbone of behavioral economics.
13:55
I'm not sure if it's necessarily fair to say that every RCT or
13:58
experiment is behavioral economics.
14:00
I think some might disagree.
14:01
And I think some of these definitions are very loose.
14:04
But, you know, I think one challenge there is, you know, scaling.
14:07
And I don't know, I've seen a lot of work that in my mind, like information
14:12
provisions that maybe don't have significant results in that space.
14:16
But I'm just curious, like, I think one that behavioral economics,
14:19
with things like default rules and information provision experiments,
14:23
can't really hold a candle in terms of magnitude of the effects of these things.
14:27
Is that right or wrong or.
14:29
Obviously, behavioral economics is kind of meant to be the sort of third way,
14:33
like idea into something very new.
14:35
I'm just curious what you think about scaling in behavioral economics ideas and
14:40
implementation.
14:43
>> Cass Sunstein: Okay, so suppose we did an RCT and
14:46
I put on my front lawn a sign that said don't smoke.
14:52
And we had a population who drove past my front lawn and a population that didn't.
14:59
And let's.
14:59
Suppose it's a big size and it's.
15:02
Let's just go with the idea it's randomized and
15:07
the diminution in smoking was zero.
15:12
That this would be a test of a pretty pathetic nudge and
15:16
whether it worked to diminish smoking in New England.
15:21
And I predict that the effect would be zero.
15:24
Or we could design in a cafeteria in Palo Alto or
15:29
in Cambridge, an intervention where there was
15:33
a notice contains meat with respect to meat products,
15:39
which would be to nudge people not to eat meat.
15:43
And I have no prediction about the effect of that.
15:46
But I wouldn't be surprised if the effect of the nudge
15:52
was small or zero if we had a randomized trial.
15:57
So the number of nudges is.
16:00
Is so large, just like, like the number of criminal punishments is so large and
16:06
the number of medical interventions is so large that if you took a class and
16:12
called them criminal punishments throughout, let's say,
16:16
western history and the effect size of the punishment was really small and
16:22
some populations are zero, that wouldn't be amazing.
16:26
And if it turned out that across a category of medicines,
16:31
the effect in diminish, diminishing fatality or
16:35
illness was really little, and that wouldn't be surprising if the.
16:41
I hope medicine's much better now.
16:43
But if you look at medicine over the course of human history,
16:46
you probably get a median effect size of zero.
16:50
So I think the meta analyses of the effect of
16:55
anything is imperfectly informative.
17:01
So I would predict, and
17:02
we have reason to think that default rules are the most powerful nudge.
17:07
And there's a paper that's a meta analysis by Elkie Weber and
17:11
others which shows the average or median effect, I Forget which, it's 26%.
17:17
That's very large.
17:19
But even that I want to be very careful with.
17:21
There's a paper by Todd Rogers and another that finds
17:27
a shift opt in to opt out creates over 90% increase in
17:32
participation in use of an educational technology.
17:38
And there are default rules that sometimes have very limited effect.
17:42
And there are things we'd want to have.
17:44
We want to use data like this to build up a theory of when a nudge has an effect.
17:50
The question does the average nudge have an effect?
17:55
It's kind of a crazy question.
17:57
It's a little like the question,
18:00
does the average increase in civil penalties have an effect?
18:07
We need to know what's the increase from and to.
18:10
And what kind of civil penalty is it and do people even know about it and
18:14
what's the affected population?
18:16
So then we'd build up an account.
18:19
I'd love it if we'd have more precision than we now do about which nudges
18:23
have what kind of effect.
18:26
As a first approximation, architectural nudges have a larger
18:30
effect than educative nudges like labeling or reminders.
18:35
But even the category of labels and reminders, it's,
18:38
it's too broad and too coarse.
18:40
If you have a label that says with a smiley face, this has GMOs.
18:50
I, I predict that will have very little effect in diminishing
18:54
people's consumption of GMOs.
18:56
If you have something that is a graphic wording for cigarettes, it wouldn't
19:01
be at all surprising if that had a significant impact on smoking secession.
19:09
>> Jon Hartley: So I guess, you know, we had this.
19:12
I Think one of the biggest things that came out of all the interest in it
19:16
was created and I think maybe most famously in the United Kingdom,
19:20
I think this might have been during the David Cameron government that,
19:24
you know, the idea was, you know, you would create a team within a government
19:29
that would essentially advise on behavioral matters related to policy.
19:35
I think also around the same time, you and Richard Thaler were also big advocates
19:40
of this idea of libertarian paternalism, this idea of having choice architects.
19:47
The idea of incorporating behavioral findings into policy,
19:50
I think was a very new thing at the time.
19:53
This was around maybe 2010 or so.
19:56
I'm curious, what's the legacy of those groups like
19:59
the original Behavioral Insights Team or Nudge or so called Nudge Unit in the uk?
20:03
I think it's been spun up,
20:04
but there's now groups in many countries that are helping to advise governments.
20:08
I mean, what have been sort of made the largest contributions from nudges
20:12
in that space with advising and working with governments?
20:18
>> Cass Sunstein: It's a great question and
20:19
it would be worthy of a long academic article or book.
20:23
And there are pieces that try to catalog those things.
20:27
I'll give a couple of examples.
20:30
In the United States, there's a program called Direct Certification for
20:35
school Meals where if you're poor and the school knows you're poor and
20:39
you're a kid, you're directly certified, meaning you don't have to apply.
20:45
And at one count, something like 10 million children were benefiting
20:50
from nutritious free lunches and breakfasts to which they were entitled.
20:56
And the Direct certification program was something in which they participated.
21:01
So that's millions of kids getting to participate in school
21:06
meal programs by virtue of the shift from opt in to opt out.
21:12
Another one that is not a product of any particular nudge unit,
21:18
but which is a product of behavioral economics research, is green by default.
21:27
So we're seeing people who are automatically enrolled.
21:31
This is a big program in Switzerland, as I recall.
21:35
It's big in Germany.
21:36
It's big in Southern California,
21:38
where people are automatically enrolled in green energy.
21:42
They can opt out, go into coal if they want, but that has very large
21:46
effects in reducing pollution, including greenhouse gas emissions.
21:50
And in some places it leads to higher electricity bills.
21:54
But people aren't opting out.
21:56
So that's kind of a big legacy item.
22:00
In the United States, people are, whenever they buy a car,
22:05
it has a camera in it that where you can see behind.
22:10
And this was a rule I got to be involved in.
22:13
It's not, it's.
22:14
It has a nudge component that is you're nudged to look at the camera and
22:20
to avoid smashing into a small kid or a dog or
22:23
something physical and strong that'll batter your car.
22:28
So it has a strong behavioral nudge feature in it.
22:32
The mandate itself of the cameras and the cars is a mandate, not a nudge.
22:37
But it was very much, I can attest,
22:39
behaviorally informed by thinking about limited attention on the part of drivers.
22:45
And there are people, including very little people, not because they didn't
22:50
grow, but because they're children who are not dead because of that.
22:55
And that's a, you know, very significant policy intervention.
23:01
Dick Thaler would refer to automatic enrollment and retirement plans.
23:06
And that's kind of the most visible, I confess, because I never want to retire.
23:12
And the idea of retirement gets my system,
23:15
one that is the intuitive automatic system in the brain.
23:20
Sad I I the retirement one isn't the one that comes first to mine for
23:26
me, but is true that automatic enrollment in retirement plans is a big deal,
23:33
but there are 1001 uses of automaticity.
23:38
The OECD now has a big project on sludge sludge audits,
23:42
reducing administrative burdens of various sorts.
23:46
Now, you can think of sludge as a problem even if you have no interest in or
23:51
enthusiasm for behavioral economics.
23:54
But the fact is that sludge route of behavioral economics, the concept and the.
24:00
Concern about the adverse effects of administrative burdens on, like,
24:05
entrepreneurs or innovators, or
24:07
on people who want to build things like build housing, or
24:11
on people who are seeking access to employment or other opportunities.
24:16
Sludge is devastating, often because of behavioral characteristics of our species.
24:27
>> Jon Hartley: I guess.
24:28
Speaking of nudge units and regulatory policy,
24:31
I sort of want to segue into your time at oira.
24:34
So during the Obama administration, you led oira,
24:39
which is the Obama administration's, I guess,
24:43
chief regulatory rulemaking sort of oversight function.
24:49
And during this time, I think you certainly were described as a pioneer and
24:54
someone who'd held the role.
24:58
Some folks, like, I think Richard Epstein, said that because you're one of the more
25:03
conservative folks in the Obama administration.
25:06
But one of the ideas that you champion, I think,
25:09
was this idea of a regulatory budget.
25:11
And some people were calling this regulatory money ball, I think,
25:15
at one point.
25:16
And I think it's something that really, actually was truly adopted in the first
25:21
Trump administration and that they really followed this idea that every new
25:26
rule that it created, for every new rule they create, they would delete a rule.
25:31
>> Cass Sunstein: Okay, so the idea that I was very enthusiastic about and
25:36
certainly didn't originate was in his cost benefit analysis.
25:41
So the idea of careful analysis of the costs and
25:44
benefits of regulatory options, including alternatives to the proposal
25:50
that is being offered to the American people, seems to me essential.
25:56
And you can, and many economists do, like cost benefit analysis,
26:00
who have no interest in behavioral economics.
26:03
The idea of cost benefit analysis outruns behavioral economics,
26:08
was pioneered by people who were not particularly keen necessarily on
26:12
behavioral economics or didn't know what it was.
26:16
But I like behavioral economics because it is a corrective to behavioral
26:21
biases which people in government, either because they're human or
26:26
because they're responsive to people outside government, are subject to.
26:32
So you might think, I heard of a case in which a risk came to fruition.
26:36
There ought to be a law, let's regulate that.
26:38
But it may be that the case was an outlier, very rare, and
26:42
that the costs of regulating outrun the benefits.
26:47
So cost benefit analysis as a foundation for every environmental rule,
26:53
every highway safety rule, every food safety rule, I really like that idea.
27:00
There may be cases where it's challenging to quantify,
27:03
but the economic Constitution of the United States, I say, should and
27:08
does have cost benefit analysis at its heart.
27:11
Let's distinguish that from a regulatory budget.
27:15
The idea of having a ceiling on the cost of regulation in a year,
27:20
I don't like that idea because if.
27:23
And so the regulatory budget idea is one to which I'M opposed.
27:26
I get it, but I don't like it.
27:28
The reason is if you have a regulation in a year, let's say, that costs $5 billion,
27:33
and then you have three others that cost $8 billion, now we're up to $13 billion
27:39
in cost, and then you have four others and the cost is starting to get really high.
27:44
That might be horrific depending on what you get in return for them.
27:49
If you get in return for them $100 million, it's a catastrophe.
27:54
You shouldn.
27:56
But if you get in return for them $100 billion in benefits, then go for it.
28:03
So cost benefit analysis, hooray.
28:07
Regulatory budget.
28:10
A kind of sympathetic and
28:13
respectful boo reaction.
28:18
Cost benefit analysis right now isn't riding as high as I would like
28:24
to in the US Government, and the reason isn't running as high as
28:29
I would like is that the Biden administration didn't, let's just say,
28:34
put cost benefit analysis at the foundation of its regulatory policymaking.
28:41
It had environmental and
28:43
other goals that I hope were broadly compatible with cost benefit balancing.
28:49
But it wasn't, because cost benefit analysis was in the driver's seat.
28:54
And in the Trump administration, the deregulatory efforts,
28:59
some of which I applaud, some of which I don't like so much,
29:03
the deregulatory efforts seem to be founded on a policy commitment
29:08
that is not grounded in empirical analysis of costs and benefits.
29:14
So the regulatory budget idea I'm not for.
29:19
I don't like the idea of one regulation in, 10 out.
29:24
I think that's random and a gimmick.
29:27
If you have a year where you have three regulations out and
29:31
500 in, that might be a good year, depending on their content.
29:37
If you have a year with three regulations in and 500 out,
29:40
that might also be a good year, depending on their content.
29:45
What we want is careful cost benefit assessment of existing regulations.
29:50
And we did that in the Obama administration, at least for starters.
29:54
I would applaud the current administration if it did careful assessment of
29:59
existing regulations in terms of cost benefit analysis in deciding whether
30:04
to get rid of them.
30:06
And you'd get rid of a lot of them if you did cost benefit analysis of whether they
30:10
made sense.
30:11
So there's a happy alternative universe which I hope will arrive at,
30:17
in which the regulatory state really is a cost benefit state.
30:23
>> Jon Hartley: I think one of the other things you were an advocate of in
30:27
the first or in the Obama administration, correct me if I'm wrong,
30:32
was occupational licensing deregulation,
30:35
meaning that workers can take their license from one state to another.
30:41
Don't have to take additional new exams.
30:43
You know, say a nail technician in Maryland can also work in Virginia and
30:48
doesn't have to get an entirely new license.
30:52
This is a huge challenge for people that move a lot.
30:55
So, you know, spouses of folks in the military that are moving quite frequently
30:59
and if, you know, they're a licensed worker,
31:02
they often have to get relicensed, which is a huge tax and huge burden.
31:06
Here's what you think about all the progress on that.
31:09
>> Cass Sunstein: I'm upbeat.
31:10
I think it's good.
31:12
So occupational licensing, you're right,
31:14
is something that I was very concerned about in government.
31:19
And the fact is it's a drain on economic growth,
31:23
on entrepreneurship, and on individual opportunity.
31:27
So there's an old liberal idea.
31:29
By liberal, I don't mean left of center.
31:31
I mean the liberal political tradition of careers open to talents.
31:36
That's the idea.
31:38
And libertarians tend to like it.
31:40
Many progressives like it, too.
31:43
And careers open to talents.
31:45
It's a beautiful idea.
31:47
It was a little bit boring, maybe in the 1980s because it seemed.
31:52
Yeah, of course.
31:52
What else have you got?
31:54
But it's not boring now.
31:56
Where careers open to talents isn't something that.
32:00
Is a lived reality for too many people.
32:03
So the idea, if you're licensed in Arkansas,
32:06
you get to do the same thing in Mississippi.
32:09
A presumption in favor of that seems to me a really good idea.
32:14
Unless there are unique conditions in Mississippi
32:17
such that the Arkansas license doesn't warrant that license, or
32:22
unless there's something wrong with these original states licensing policies.
32:29
Now this could be an opening through which you could drive a truck.
32:33
But I,
32:34
I want to start by saying there's a strong presumption in favor of reciprocity.
32:39
>> Jon Hartley: It's fascinating.
32:42
It's amazing to see how much strides that have been made in occupational licensing.
32:46
It's an area that I, I work in and sort of tracking these trends across states and,
32:51
and across countries and, and I mean the US in general, at least from some of
32:56
the work that I've done in a forthcoming paper with Morris Kleiner, is that the US
33:01
actually has some of the most stringent occupational licensing regulations.
33:05
At least if you were to measure the fraction of workers that have a license
33:10
compared to many other countries is very, very high compared to say, Europe.
33:15
So it's interesting that you would think sometimes that we have a very
33:21
dynamic economy in the US but there are all these sorts of barriers that exist.
33:28
Not saying that we don't need some licensing for doctors and
33:32
certain professions are important, but to what degree do hair braiders and
33:37
nail techs have to be licensed and so forth?
33:40
Dog walkers and so forth.
33:43
It's certainly an interesting topic that's made a lot of headway in recent years.
33:47
Some people might say it's still small compared to other macro things,
33:50
But I'm with you in that I think it, it's more important than people realize.
33:53
I think maybe about 25, 30% of the workforce has a license or so,
33:58
and it's really substantial and
34:01
it's used toward those people who are generally in lower incomes.
34:05
And so it's really, I think, one of those policy ideas, you know,
34:09
occupational license deregulation that can actually improve opportunity for
34:14
lower income individuals and help improve their productivity.
34:18
So I'm with you there.
34:20
I want to sort of segue into legal philosophy.
34:23
You spent a lot of time,
34:24
time in your career teaching law classes across a variety of fields.
34:30
You, I think, spent a lot of time working on administrative law.
34:33
I just was curious, how would you describe your judicial philosophy?
34:36
I kind of understand you're a judicial minimalist.
34:39
You're sort of an advocate for maybe more executive and administrative
34:43
decision making or more intelligent administrative decision making.
34:47
You're a fan of Experts, broadly speaking.
34:50
And you recently wrote a book, Law and Leviathan Redeeming
34:53
the Administrative State with your Harvard Law colleague Adrian Vermeil.
34:57
Tell us more all about your judicial philosophy.
35:01
>> Cass Sunstein: Okay, so there are a couple things I like.
35:03
If there's the idea at a highest level of abstraction,
35:07
the idea is deliberative democracy.
35:10
So a deliberative democracy is something associated
35:14
with the German philosopher Jurgen Habermas.
35:18
John Rawls has endorsed it.
35:20
James Madison was a deliberative democrat in economics.
35:25
Amartya Sen is an enthusiast for deliberative democracy at Stanford.
35:33
James Fishkin is a fan of deliberative democracy.
35:37
And the basic idea here.
35:39
We'll get to the judges in a moment.
35:41
Is that a well functioning constitutional order combines accountability and
35:46
responsiveness.
35:48
That's the democracy part of deliberative democracy, with a commitment to reason
35:53
giving and deliberation among people who are trying to figure things out.
35:59
That's the deliberative part of deliberative democracy.
36:03
So we don't have a government by referendum.
36:09
The right to instruct was rejected by the Constitution's founders on the ground
36:14
that it would destroy the point of the meeting.
36:17
They're supposed to deliberate with one another and figure out what's best.
36:21
So deliberative democracy will naturally lead, lead to some enthusiasm for
36:25
technical expertise.
36:27
If you're trying to figure out how to handle road safety,
36:31
intuition won't be enough.
36:33
The deliberative part requires a commitment to getting really empirical.
36:39
And I see the administrative state at its best, not as it always exists by
36:44
any means, as continuous with the founding commitment to a deliberative democracy.
36:51
Okay, in terms of the judges, we want a judicial role
36:56
that's alert to the ambition to a deliberative democracy.
37:01
And we want courts respectful of the processes of deliberative democracy and
37:09
we want them attuned to ensuring that it doesn't misfire.
37:15
So the most important thing the Administrative Procedure act does,
37:20
maybe is to authorize courts to strike down agency action.
37:25
That's arbitrary.
37:27
So if an agency says that we're going to deem the social cost
37:32
of carbon to be 500 because that's really big and
37:36
the climate crisis is upon us, that's arbitrary.
37:41
That's not reason giving.
37:43
If you say we're going to deem the social cost of carbon to be zero because
37:48
we think climate change is not real, that's almost certainly arbitrary.
37:53
I think the almost is just I'm being very careful here.
37:57
We have to figure out what the grounding is for
38:00
the administrative state doing one thing or another.
38:05
So arbitrariness as A very bad thing is consistent
38:10
with the aspiration to a deliberative democracy.
38:15
A democracy that's not deliberative might be arbitrary, but
38:20
that's a form, I say, of authoritarianism, which runs afoul of
38:24
the founding era's ambitions to have a deliberative democracy, which.
38:30
Okay, then there's the rule of law.
38:32
So the book Law and Leviathan is about what the rule of law entails.
38:40
And here the idea is that the rule of law has an internal morality where,
38:46
and this is Hayek is an inspiration, at least in my mind, for
38:51
this, that if you have a rule in the world that isn't on the books,
38:56
that's not the rule of law.
38:58
So if. If in reality the enforcement officers
39:02
are acting inconsistently with the rules on the books,
39:05
that violates the internal morality of law.
39:08
If the law is changing so rapidly that people can't plan,
39:11
that's inconsistent with the rule of law.
39:14
If there's retroactivity, so
39:17
people do things that are fine by legal requirements in 2025,
39:22
but then in 2027 they're punished that it runs afoul of the ban on retroactivity.
39:30
If people can't understand what the law is, if it's too vague or confusing,
39:34
that's inconsistent with the.
39:36
The rule of law, the morality of law.
39:39
Now, if you have no redistribution or lots of redistribution,
39:43
that might be good or bad, but it has nothing to do with the rule of law.
39:48
At least it doesn't until we specify what the legal sources.
39:54
But you can do lots of stuff consistently with the rule of law.
39:57
Still, the rule of law is.
39:59
Central and a foundational part of the restrictions, as they should exist and
40:05
frequently do exist on the operation of the administrative state.
40:10
In terms of the Supreme Court's role,
40:13
I like the idea of the judges avoiding very ambitious theories and
40:18
avoiding very ambitious, broad rulings.
40:22
So there's some people on the left and some people on the right get
40:27
excited about maximalist rulings where the court, you know,
40:32
strikes down a bunch of things or adopts a theory of liberty or equality,
40:37
and bam, our country is now being ordered to do all sorts of stuff.
40:42
I'm skeptical of that.
40:44
The idea of judges being theoretically humble and
40:48
ruling in ways that are agnostic about what liberty and equality really means.
40:56
They don't have expertise on that.
40:57
And it's a pluralistic society, after all.
41:00
So being theoretically modest and
41:03
also ruling narrowly on the case, these are presumptions.
41:07
These aren't edicts.
41:10
But ruling narrowly rather than broadly, I.
41:13
I like that.
41:14
So here I stand with Chief Justice Roberts, who says,
41:18
if it's not necessary to decide an issue in order to resolve a case,
41:22
it's necessary not to decide an issue when we're resolving a case.
41:27
So that's a plea for a form of minimalism, which, whether or not it's good in art,
41:32
is usually pretty good for the Supreme Court of the United States.
41:37
>> Jon Hartley: I'm curious, I guess, on the topic of both the courts and
41:40
deliberate democracy.
41:41
I'm curious how you think about our current system and
41:46
how much it has maybe strayed from how the founders maybe
41:50
envisioned the sort of original Madisonian form of government.
41:56
And maybe, you know, you think it's important that for
42:00
it to have changed since then.
42:02
I mean, we now live in a world that's very,
42:04
very different from the time of the Founders.
42:06
Certainly from a national security perspective,
42:08
we have nuclear weapons and so forth.
42:10
But I'm just curious, you know, there's this, I would say, criticism
42:15
of the current sort of state of affairs or the current regime in saying that,
42:20
well, you know, Congress doesn't make laws and tries to avoid.
42:25
To deliberate.
42:26
Now they don't want to decide these big issues that it's
42:30
really their responsibility to decide.
42:34
And then what ends up happening is they give a lot of power to
42:37
the administrative state that decides a lot of things, or the presidency,
42:42
the executive that decides a lot of things, and
42:44
then they do things, and then sort of it goes between the executive and the courts,
42:49
and the courts kind of are being forced to weigh in on these things that.
42:53
That perhaps Congress should be weighing in on.
42:56
I'm curious what your thinking is about the, the current regime and
43:00
how the current powers, you know, three branches of government work
43:05
together compared to sort of Madison's original vision.
43:08
And, and how do you think it should be, how you think it should function?
43:12
>> Cass Sunstein: It's great.
43:12
It's complicated.
43:14
So let's take two stabs at it.
43:18
There's a grand narrative out there which your question can be taken to press,
43:24
which is that we've seen successive breaches of Article 1,
43:29
Article 2 and Article 3 of the Constitution.
43:34
Article 1 insofar as Congress grants broad discretionary authority to
43:39
the executive and administrative entities, violation or
43:43
breach of Article 2 insofar as we see independent agencies exercising authority
43:48
outside of the presidency, that's on the run in current constitutional law.
43:54
So the idea of independent agencies is no longer secure.
43:57
It might be gone pretty soon, but it's been around since 1935.
44:03
And then there's a third breach of Article 3 where a lot of adjudication is done by
44:08
the executive agencies.
44:10
You're putting a Spotlight on Article 1.
44:12
I don't think the grand narrative is right, but I think.
44:16
But I'll get to something kind of adjacent to it that I think is right.
44:20
If you look at the founding period,
44:23
a lot of discretionary authority was given to the executive and to agencies and
44:29
the constitutional concerns were either absent or very occasional.
44:35
So the breadth of grants of authority to the executive pursuant to open
44:40
ended words like reasonable is pretty eye opening in the founding period.
44:45
It's not like Congress made all the decisions and so, okay, and
44:49
said to the executive, okay, do what we just said.
44:52
Instead Congress said do what's reasonable or appropriate, things like that.
44:57
So there's real continuity between the first, second Congress and
45:01
first and second Congress and what we now observe, the idea that independent
45:06
agencies are constitutional affront or barnacle, that's also hard to defend.
45:11
Independent agencies were around in the founding period and
45:15
it looks like most people then thought that was constitutionally fine.
45:21
The adjudication issue is a little more complicated and
45:24
the magnitude of adjudication done by the administrative state now is very
45:29
plausibly out of accord with constitutional expectations.
45:33
Okay, there's all that.
45:36
It's clearly the case that the breadth and and
45:41
ex magnitude of executive branch authority was not expected.
45:48
And I think it's less because there's discretion than the number of
45:53
domains in which discretion is exercised.
45:56
And the power that discretion is exercised with respect to
46:01
way outruns the founding period.
46:04
So if you think of the EPA or the Department of Transportation or
46:08
the Social Security Administration or
46:11
the Federal Communications Commission, or the Federal Trade Commission.
46:16
My gosh, they're doing all sorts of stuff that at the founding,
46:21
the national government wasn't expected to do.
46:24
The Department of Labor.
46:26
So whether this is an affront of the Constitution is, I think,
46:30
not at all clear.
46:32
Because what part of the Constitution does it violate is a very fair question.
46:37
It's not clear it violates any part of the Constitution, but
46:40
it's very different from what was anticipated.
46:44
Now, this is a large question.
46:48
No modern industrialized society looks like the founding era or
46:54
40 years after the founding era government.
46:59
So we're not seeing that in Canada or France or Italy or Germany.
47:04
Not that any of those is perfect.
47:06
But the idea that we would have a small national government akin to that of,
47:12
let's say, 1830, that seems cray cray.
47:17
I think.
47:18
So that's cool person talk for crazy.
47:21
I'm trying to be a cool person.
47:22
So failing, evidently.
47:25
But the, the, the mission of trying to discipline
47:29
the administrative state by reference to the rule of law and
47:35
to cost benefit analysis, including elimination of
47:39
administrative burdens and barriers and sludge.
47:44
That's a really honorable mission.
47:48
>> Jon Hartley: That's fascinating.
47:50
I guess you just maybe broadly.
47:54
Thank you. Here, just about political philosophy and
47:57
liberalism, you've written quite a bit.
47:59
Bit about political philosophy as well, in particular about liberalism.
48:07
In 2023, you wrote a New York Times essay titled why I Am a Liberal.
48:12
Does the term liberal mean a whole lot anymore?
48:16
I'm just curious.
48:16
There's everything from economic liberals to social liberals.
48:21
Many who I would say are firmly in one camp and firmly not in the other,
48:26
which traditionally I'd say would be a key distinction between maybe Republicans and
48:33
Democrats, at least from the 1980s through maybe 2016.
48:38
I'd say it would probably go back even further,
48:43
maybe even pre the time of Buckley and others in maybe the 50s.
48:49
Now, Locke also talks about a lot about moralism in book four of
48:53
the essay concerning human understanding.
48:57
And certainly I think it's fair to say that the founders believe that morals
49:01
writing about nature's God and the laws of nature and then declaration.
49:06
So I think there's some today, namely the post liberals,
49:13
that might characterize them as libertarians,
49:18
you know, in social liberals.
49:22
But, you know, I think that's somewhat historical and untrue.
49:28
And I mean, they also were in favor of tariffs as well.
49:31
So, you know, I think calling them full economic libertarians I think wouldn't be
49:35
true either.
49:35
Now, they certainly believed in property rights.
49:38
That was something that was foundational.
49:39
And they certainly believed in, you know, fighting the excesses of government.
49:45
You know, they're fighting, you know,
49:47
very small taxes by compared to the taxes today.
49:50
And that in part the American Revolution.
49:55
Obviously, you know, economic liberalism, property rights has been responsible,
50:00
in my opinion, for enormous amounts of economic growth.
50:04
And maybe there's some unintended consequences of that.
50:07
But I'm curious, what does it mean to be a liberal today as you see it?
50:13
And are there limits on liberalism as you see it, if any?
50:17
>> Cass Sunstein: Okay, so a liberal today believes in freedom and
50:22
pluralism and the rule of law.
50:25
That's a triumvirate, a holy trinity.
50:29
Freedom, pluralism and the rule of law.
50:33
Ronald Reagan was a liberal.
50:35
Barry Goldwater was a liberal.
50:38
Barack Obama is a liberal.
50:41
Franklin Delano Roosevelt was a liberal.
50:44
So this is a big tent.
50:46
Hayek was emphatically a liberal, no question about that.
50:52
John Stuart Mill, also Robert Nozick for sure.
50:57
John Rawls, definitely Susan Okun,
51:00
Stanford's political scientist, no question at all.
51:07
So I have a book coming out in just a couple weeks called Liberalism in Defense
51:11
of Freedom and, and I almost called it Big Tent Liberalism.
51:15
Now, the disagreements between Susan Okun, for example, and John Rawls are fierce.
51:23
The disagreements between Robert Nozick on the one hand,
51:27
and Joseph Raz, a philosopher who was a liberal, died not long ago.
51:33
Those are also fierce.
51:34
But pluralism, freedom, and the rule of law are a shared set of convictions.
51:42
I think it's.
51:43
It's like, really important for those who are part of the liberal
51:48
tradition to stand up and say so, even if they disagree, you know,
51:54
very fiercely with people who are also part of the liberal tradition.
52:00
So I feel, right now, I'm not a libertarian, but
52:03
I feel libertarians are my brothers and sisters.
52:06
I worked in the Reagan administration as the Department of Justice.
52:10
I had some serious, serious disagreements with
52:15
President Reagan, but, you know, my.
52:19
My intellectual sibling and commitment to liberalism.
52:24
So this might not have seemed as important 30 years ago as it is now,
52:29
but given what's happening in Russia and China,
52:33
in various parts of the world, given the illiberal left,
52:38
which is sometimes not that excited about freedom of speech on campus or
52:43
elsewhere, not just the Marxist left, but
52:47
the illiberal left, given the illiberal right in Hungary, for
52:52
example, where the idea is a democracy, that's illiberal,
52:57
I think it's an oxymoron, but we understand what the sentence means.
53:04
To put a spotlight on freedom, pluralism, and
53:09
the rule of law as composing the liberal tradition
53:14
is a big step forward, and it may be, you know, of.
53:21
Of great importance in the next 50 years.
53:25
>> Jon Hartley: Well, Cass, I really want to thank you for coming on.
53:27
This has been an amazing conversation, and you've had such a prolific career.
53:32
It's a true honor to have you.
53:33
>> Cass Sunstein: On, an honor to be able to talk to you.
53:35
Thank you.
53:37
>> Jon Hartley: This is the Capitalism and Freedom the 21st Century podcast,
53:39
an official podcast of the Hoover Institution Economic Policy Working Group,
53:43
where we talk about economics, markets, and public policy.
53:45
I'm John Hurley, your host.
53:47
Thanks so much for joining us.