Latest / The Capitalism and Freedom in the Twenty-First Century Podcast / Economic Growth, Macro-Models, and a Move to the Hoover Institution

Transcript

Economic Growth, Macro-Models, and a Move to the Hoover Institution · The Capitalism and Freedom in the Twenty-First Century Podcast

0:09
This is the Capitalism and Freedom in the Twenty-First Century podcast where we talk
0:12
about economics, markets and public policy. I'm John Hartley, our host. Today my guest is John
0:17
Cochrane, who is the Rose-Marie and Jack Anderson Senior Fellow at the Hoover Institution and is the
0:23
co-chair of the Hoover Economic Policy Working Group. John is a prior guest on the show, but
0:28
this time we're going to do something a little bit different. We have something of an announcement
0:32
to make, that being that the Capitalism and Freedom in the Twenty-First Century podcast
0:36
is becoming the official economics podcast of the Hoover Institution Economic Policy Working Group.
0:42
It's very fitting that the Capitalism and Freedom in the Twenty-First Century podcast become a part
0:46
of Hoover. This podcast gets its name from the 1962 book, Capitalism and Freedom by
0:52
Milton Friedman. Hoover is where Milton Friedman called home from after he left the University
0:59
of Chicago after winning the Nobel Prize in the late 1970s to his death in 2006. Hoover has also
1:06
been the home of many luminaries, including Thomas Sowell, George Shultz, John Taylor,
1:11
and of course John Cochrane. John's going to explain a little bit more about Hoover's Economic
1:17
Policy Working Group and what this means for the podcast, and interview me a bit as well and
1:22
introduce me to the Hoover audience here. Thanks. Yes. To our devoted listeners,
1:30
half of the purpose of this show is to introduce the Economic Policy Working Group angle to Jon's
1:37
already faithful listeners, but the other half is to introduce Jon and the podcast to the Hoover
1:42
community where this podcast will now be hosted. Jon, let me ask you... You get to introduce
1:49
yourself a little bit after I introduce you. Jon is currently a graduate student at the
1:54
Stanford Economics Department, I'd like to say star graduate student. I'm going to embarrass
1:59
you a little Jon. Jon is one of the smartest economists I know as well as remarkable for being
2:07
interested in public policy and public outreach at an age when most economic graduate students
2:14
are simply sitting in the bowels of the library working on equations. Jon runs this podcast,
2:20
he's incredibly well-connected. Everybody I know seems to already know Jon Hartley,
2:25
and you're only in grad school. At some point on this show, I'm going to ask you how you do it.
2:33
[inaudible 00:02:34]. Jon has a background at University
2:35
of Chicago, a little time at Goldman Sachs. Tell the Hoover audience just a little bit of
2:40
your background, Jon, and then a little bit about what you've done with what the philosophy of the
2:46
Capitalism and Freedom Podcast has been, and we'll talk about what it's going to be in the future.
2:51
That's very kind, John. A very, very kind introduction. Perhaps I should be spending more
2:57
time working on research in those same libraries and basements where various graduate students are.
3:06
For those who aren't familiar with this podcast, these are one-on-one interviews with experienced
3:11
economists, policymakers, people in financial markets. We usually ask questions about their
3:18
careers, their main ideas and contributions. We like to focus on a lot of long-term trends
3:26
rather than the news of the day. But the idea is to really just focus on a lot of the key ideas,
3:33
many of which were actually mentioned in Milton Friedman's nineteen-sixty-two book
3:37
Capitalism and Freedom. He talked about a lot of things in that book that I think were quite
3:42
prescient and quite relevant to today. Whether these are topics like occupational licensing,
3:48
universal school choice, these are topics that have been sweeping the nation in
3:54
recent years or sweeping the United States. We've spoken with former Arizona governor
3:59
Doug Ducey, who's the first governor to pass these policies in the US. We spoke with Morris Kleiner,
4:05
who's the world's leading expert on occupational licensing. We've also had really amazing
4:09
discussions with academics over some of their key contributions, and some who've represented
4:18
key transformations in economics. One being Steve Levitt, for example, who co-authored
4:23
the book Freakonomics. He actually announced his retirement on this podcast a few weeks ago,
4:28
after which the podcast kind of exploded in popularity. It's been quite amazing to go to
4:34
economics conferences and talk to people who I've never met before saying that, "Well, I listened to
4:39
your podcast." It's a pretty amazing experience. To give just a little bit of background about
4:45
myself, as John mentioned, I'm an economics PhD student at Stanford. I've had a good number of
4:50
policy stints in a number of places. Most of my research is focused on, one stream is focused
4:57
on business cycle, macro and finance, so many of the same topics that John works on. Then
5:03
another stream is working on economic growth and regulation, so things like zoning, occupational
5:10
licensing, things that are wholly macro or might be causing growth. Things like management being a
5:18
factor of production or antitrust policy to what degree can that hurt or help growth? Those are
5:25
some of the topics that I work on. But in terms of some of the topics that I think are really
5:31
interesting and have been a real theme of this podcast, including some of our discussions that
5:37
we've had, John, are topics like DSGE models. One theme for those who aren't as familiar with
5:44
what a dynamic stochastic general equilibrium model is, they're kind of what is often called
5:50
the workhorse macroeconomic business cycle model that places like central
5:56
banks use like the Federal Reserve and such models. They're maintained by central banks
6:01
around the world and economists like myself and John have to write them down on our papers. But
6:07
I think sometimes we're a bit skeptical of how useful they really are. We've spoken with Larry
6:14
Summers on this podcast and we've talked about some of his original debates with Ed Prescott,
6:21
who is sort of one of the key figures in starting DSGE real business cycle models. But I think John
6:26
and I are both a little bit skeptical whether any central bankers out there actually moved
6:32
their dots based on what's coming out of the house DSGE model compared to say BAR's. I'm
6:38
going to probably pass this back to you, John. Let's come back to that one. I want to do a little
6:43
more intro first. The Hoover institution, those of you don't know it, combines scholarship and policy
6:51
and outreach. Most of us senior fellows have been academics and now work for Hoover and continue
6:59
academic work, but also try to do things in the policy space as well. The economics policy working
7:05
group is sort of the bureaucratic institutional structure that brings together everything we do
7:11
at Hoover in the economic policy space. It includes a whole list of interesting people
7:17
who you will get to know over time. Because one thing I think we're going to do is have
7:22
Jon interview interesting people associated with what we do. We have a seminar series which brings
7:29
in more interesting people that Jon will get to talk about. We have our conferences. There's an
7:35
annual monetary policy conference, which is getting more and more successful I think,
7:39
and various other things going on. A range of topics, not just macro Taylor
7:44
Rule money that we've been talking about here, but healthcare, health insurance, energy, climate,
7:53
state and local. All the zoning regulation stuff that Jon mentioned are prime concerns.
8:00
And we are not just banging on the wonders of the free market, the challenge of course today is that
8:08
those of us of free market bent like Jon and I at the ghost of Milton Friedman who lives in this
8:14
office. The challenges of course on the left, the advance of the progressive state and on the right,
8:21
economic nationalism and those other trends that we need to think about and deal with.
8:28
Those are kind of the broad things of the Economic Policy Working Group.I think it's
8:33
going to be wonderful to bring Jon's great interview style together with the people
8:40
and the topics that we cover in the Economic Policy Working Group, so you'll never get bored.
8:47
Well, that's very kind of you to say, John. Really excited, one, to be working with all
8:57
the wonderful Swedish scholars that Hoover has and really excited to... One, I will say we've
9:03
actually interviewed quite a few people at Hoover already, but there's many still to interview,
9:08
and so I'm very excited for those discussions. Wonderful. Okay, well as long as we've got
9:14
some time left, Jon, why don't we talk about economics a little bit?
9:17
Let's do that. What have you been working on this summer?
9:23
It's funny you ask, one topic that I've been working on quite a bit, it's also a topic that
9:28
you've been working on quite a bit for the past few years, and that being the fiscal theory of
9:33
the price level. Really trying to understand to what degree we can test the fiscal theory of the
9:40
price level. It's at least the key equilibrium condition in the fiscal theory is one that you
9:49
can empirically test because like MV equals PQ or the monetary base velocity prices and quantities,
9:57
you can see all those. Similarly with FTPL, you can observe debt, money, prices,
10:03
and with a lot of history you can look at the future path of surpluses and you can discount
10:09
them back. One thing I think I've been a bit surprised about is actually I think how well
10:16
the fiscal theory actually works. I don't think it works perfectly, but I think it works actually a
10:22
bit better than at least that I expected it to. Let's back up just a second and explain to our
10:28
listeners who... I am sure everybody knows exactly what the fiscal theory of the price level is,
10:33
of course. So the fiscal theory of the price level states that the fundamental determinant
10:40
of inflation in our economy is not so much money, not so much interest rates,
10:47
those matter, but the fundamental question is, is there more government debt than people think our
10:54
government has the will or ability to repay? When that happens, people try to get rid of government
11:01
debt quick before it becomes worthless, and in doing so they drive up inflation. That's the basic
11:07
concept of the fiscal theory of the price level. Now I'm glad Jon, you're taking up this challenge.
11:12
Because the basic idea is just like prices present value of dividends, it's the theory
11:17
of asset pricing applied to government finances. When people think that there aren't going to be
11:23
any dividends, they try to dump the stock, the price goes down. When people think the government
11:27
isn't going to be able to raise taxes over what it spends, they try to dump the government debt
11:33
and that price level goes up. Same mechanism. But we've been asked trying to test prices present
11:39
value dividends for 40, 50 years and it's still contentious. You still say the word efficient
11:46
market, basically that statement, and people go bananas, but of course it's not efficient.
11:52
Now of there's all sorts of false steps. The whole thing is kind of vacuous. If you let
12:00
the prices expect present value dividends, well who's expected, who's discounted? Once you put
12:07
those things in, the usual criticism is that it's vacuous. How are you dealing with this central
12:13
problem which permeates all of asset pricing that take into its full generality, it's very
12:23
hard to test. Because you don't know what the expected dividends are and you don't know what
12:28
the discount rate to take the present value is. Well, I think that's one central challenge. One
12:37
thing that we can do is, we can make some assumptions. With the benefit of hindsight,
12:45
we have hundreds of years of fiscal data that's been compiled by great economists like Carmen
12:51
Reinhart, Ken Rogoff, among many others, others that have compiled data on interest rates.
13:00
What you can do is, you can go back to say 1800 and you can, with the benefit of knowing
13:09
what the future fiscal history will be or what the future path of surpluses will be,
13:13
you can discount those future surpluses back using a long-term discount rate of what it was
13:20
in 1800. You can repeat that and iterate forward and recalculate that, and you can compare that
13:28
present discounted value to the real value of government debt or your debt to GDP ratio.
13:35
This is something that others have done, yourself have done in different ways. There's a paper by
13:44
Hanno Lustig and co-authors that's forthcoming in Econometrica, that there's a sort of a similar
13:52
government debt valuation exercise. I would say it's a little more complicated and it's
13:56
not exactly testing fiscal theory of the price level exactly either, but it's the same idea.
14:02
Trying to take seriously this idea that the real value of government debt, you can compare it to
14:07
the present discounted value of future surpluses. Let me put in a plug here for what you're doing
14:15
in the larger scheme of things, especially if any young researchers are listening. We've got
14:20
a brand new theory, it says that basically prices present value explains inflation in a
14:26
way people haven't really thought about before. We have 50 years of accumulated methodology on
14:33
how you apply that insight to stocks and bonds, and is simple low-hanging fruit to
14:40
apply that methodology to government debt and inflation questions. Now, low-hanging fruit is
14:45
not always as easy to eat as it seems, because there's hundreds of little puzzles to work out
14:53
ways to implement it. But the the big picture of how we would go about to do this is fairly
14:59
straightforward. I think this is a very exciting research area and I'm glad to see you doing it.
15:03
So let me shift the conversation a little. I want to ask you about our other big issue. Which this
15:09
is one of the things that's been on my mind, and I just finished a paper called Expectations in the
15:14
Neutrality of Interest Rates, which talks about this, but part of why it's on my mind. But your
15:19
listeners may have seen there's an Atlantic article, nobody knows how the Fed actually
15:24
does things. There's a Scott Sumner blog post, nobody knows how interest rates actually work.
15:29
The point of my article was really to look at the very basics and to realize that there is
15:35
a standard doctrine of monetary policy that says higher interest rates, lower spending
15:42
makes the economy worse, and through the magic of the Phillips curve that lowers inflation.
15:49
But there is no faintly respectable modern economic model that embodies that doctrine.
15:57
It simply doesn't exist. Of course, even the pieces of the doctrine, the verbal doctrine,
16:04
people are starting to notice that the Fed did raise interest rates. Inflation did come down,
16:09
not clear whether the Fed just jumped in front of the party or not, jumped in front
16:13
of the parade or not, I happen to think it did. But nonetheless, the prime mechanism,
16:20
the Fed didn't have to induce a recession. Now people can spin epicycles on this, "Oh,
16:26
the Fed changed expectations." Well, the Fed's been giving speeches to try to change expectations
16:31
for 30 years, it's the first time it's actually worked. But there is no respectable model. That
16:38
of higher interest rates lower future inflation. Now that contrasts with what goes on in the bowels
16:46
of a policy world and in much of academia where people write incredibly complicated multi-equation
16:52
models and people in the policy world and the central bank and the international organizations,
16:58
they write these new Keynesian DSGE models. They're beautiful models, but they don't work the
17:05
way the words say we think monetary policy works. They do not produce... Higher interest rates,
17:11
do not produce steadily lower future inflation. That's just not how the equations and the
17:17
predictions of the model work. I'm curious, and if you want, we can go into exactly why
17:23
this doesn't work, but in your contact with the policy world, how has this cognitive dissonance,
17:32
how has this schizophrenia that's been going on for 30 years get patched over?
17:38
Well, it's a great question. My sense is this, and this may certainly offend certain sensibilities of
17:46
people that have spent their entire careers on DSGE models. But my experience working
17:52
in various policy organizations and having talked to a lot of people who have worked
17:59
very closely with central bankers or have been former central bankers themselves,
18:02
is that there's basically a bifurcation that goes on. I would kind of put it this way, central banks
18:11
are essentially the biggest subsidizer of DSGE research. And DSGE research broadly speaking,
18:20
is actually lost a lot of, I would say, interest in broader academia, broadly speaking. If you
18:28
look at the number of macro companies, the number of people just working on DSGE models,
18:32
or just say macro models in general in economics departments, has declined really significantly,
18:38
especially since the 1990s or so. Largely they've been replaced by
18:43
applied economists, people that are working on largely applied micro topics but with really
18:50
good identification. What happened I think in tandem in the 1980s beginning with Kittleman and
18:59
Prescott and the rise of real business cycle DSGE models, is that macro-economists were more focused
19:06
on theory and less focused on very good reduced form empirics. We have a lot of central bankers,
19:14
increasingly fewer who have PhDs. There's a great Financial Times article that was written by Ron
19:20
Wigglesworth a few weeks ago that said and made this observation that now only 50% of FOMC members
19:27
have PhDs compared to say 75% say a decade or two ago. I think what's in part going on,
19:35
and the reality is, that central bankers don't really pay attention to DSGE models.
19:41
I think that they use VARs to forecast big macro-economic variables like inflation and growth
19:49
and those sorts of things. But at the end of the day, most central bankers will not change their,
19:55
I say their dot or their monetary policy reaction function based on some wisdom that's coming out
20:03
of the DSGE framework, or coming out of the DSGE model, say like FRB/US that's estimated by the
20:13
economists at the Fed staff in DC. Now that said, there's some insight from DSGE models
20:20
that are used, things like aggregate demand slopes down, maybe this idea that higher interest rates
20:28
can slow down inflation. That's not something that's necessarily exclusive to a DSGE model,
20:35
I think at some level, I think more policymakers have like an ISLM kind
20:40
of model at the back of their minds rather than a highly technical 100 page DSGE macro model.
20:47
But I do think that there really are a lot of empirical examples in the past, think the Volcker
20:54
disinflation and what happened there, or the observation that Milton Friedman and Anna Schwartz
21:01
made in the monetary history of the United States. That money supply that its contraction contributed
21:08
to how deep the Great Depression is. In my mind, at least in terms of the best way forward,
21:15
is really to improve our empirics. I'm a big fan of a good identification. I'm a
21:21
bit of an empirical or contra macro-economist myself, and I think that the way forward for
21:26
macro in my opinion, which is probably a bit different than yours, John, is that I think
21:31
we need better empirics, better identification. That's not easy given that so many things are
21:35
endogenous in macro and perhaps impossible, but to me I think that's a better way forward and perhaps
21:42
a better way to get policymakers thinking about research. Because to be honest, I don't think
21:49
that people on the FOMC are reading new a hundred plus page macro papers that they get released on
21:56
the NBR working papers series on a weekly basis. That's just my view on things. Again, I fully
22:06
agree with you that I think theory is broken in that sense, but I guess the question is, do you
22:12
need a better model to beat a model in the words of Bernanke and many others or should we really
22:17
just follow the data and see what we can glean from that? And have some [inaudible 00:22:25]-
22:24
You said several multiples here. Well, let me react to you about, again, I've got a list of
22:29
the eight things you said that I wanted to react to here. My view is actually, the models might
22:35
be right. The bottom line where I come to is a combination of New-Keynesian DSGE with the
22:43
fiscal theory of the price level, repairs the most outstanding obvious problem of the DSGE models.
22:52
That most outstanding problem is an assumption that inherently there are multiple equilibrium
22:59
and the Fed acts by making everybody jump to an equilibrium that the Fed likes by threatening to
23:05
blow up the economy if we don't do it. Literally that's in there. You got to be kidding. But that's
23:10
easy to fix with fiscal theory of the price level. Then you get what I think is quite sensible,
23:16
which is a view of the economy, which is the way we think it works. There's production and there's
23:23
workers and there's markets and there's various kind of shocks and interest rates do and do their
23:32
thing. I'm not against the models and in fact, what I said was, I said something very careful,
23:38
we don't have a model that encompasses standard doctrine. Now it's not obvious that the doctrine
23:43
is right and the models is wrong rather than the models are right and the doctrines wrong,
23:49
because that doctrine really just comes from stories that are repeated over and over again.
23:53
The doctrine again, that there is interest rates that cools the economy and then the
23:56
magic of the Phillips curve lowers inflation. The latter part is especially troublesome.
24:03
Now you also made fun of a hundred-page models, and that's not really where the trouble is.
24:08
There's a one-page version of the model, that doesn't work either. The famous three-equation
24:13
model. Which includes there's something like price stickiness, that's a crucial part of
24:21
why this monetary policy... Although we don't really know what that is. This is not about
24:27
overly complicated models. This is about models that even boil down to their essence don't work
24:31
the way you think. I don't want to get too deep into models, but there's a clear intuition for
24:36
why. Suppose the Fed raises the interest rate, the nominal interest rate. Now, normally we think of a
24:42
bottom line that real things remain the same. If you measure something in feet or meters,
24:48
that doesn't change it. So the Fed can raise the interest rate, but then the puzzle is, "Well,
24:54
why doesn't then next year's inflation rise so that the real interest rate remains the same?
25:01
Why does next year's inflation go down?" That's the fundamental puzzle, because in the models,
25:08
there is no model where next year's inflation goes down when the Fed raises it this year.
25:13
The verbal doctrine is a static, you said ISLM. This is a static view,
25:20
a view that eliminates time. The central thing, what I meant by a respectable model,
25:26
is one that treats time correctly. The intuition people have says, "Well, higher interest rates,
25:31
that's going to lower output and lowering output somehow is going to lower prices." Then they jump
25:37
from that to, not prices go down today, but prices slowly go down tomorrow. You have sort
25:40
of a verbal, an adjustment mechanism to a static model produces dynamics, but that's not how you do
25:47
respectable economics. Respectable economics today is intertemporal, it's today versus
25:52
the future. When they raise the interest rate, that's going to typically raise tomorrow, that's
25:59
the kind of fundamental problem. It's not about being overly complicated, it's about just sort
26:04
of fundamental economics that might be right but does not comport with the standard story people.
26:11
Now you also said... I'm sorry, I got a laundry list here so I got to go through it. People
26:16
aren't... In graduate school your compatriots are not working on macro and inflation,
26:22
they're all doing difference in difference empirical things, trying to find tiny,
26:29
tiny little exogenous causal changes in the sea of endogenous facts that we have. But it's really
26:39
weird. When we went through a financial crisis, there was an outpouring of what
26:44
caused the financial crisis, and then we had the sovereign debt crisis and there was an outpouring
26:49
of what caused the sovereign debt crisis. We've just had an inflation that we haven't seen since
26:55
the Reagan era, and you would think there would be an outpouring of what caused this inflation.
26:59
The answer is no. People would just, "Oh, it was a supply shock or greed or shrinkflation
27:04
or whatever. Can we get back to work on doing a difference in difference of whether green ties
27:09
or yellow ties raise your GDP by 0.001%. Now you said empirical, and of course the
27:18
problem with macro is that those tools don't work. They don't work because you don't have
27:24
natural experiments. This central problem of causation versus correlation is one
27:33
that is not as easily solvable. The VAR, the vector autoregression tried to do it,
27:39
but that I think is completely played out. It was a brilliant idea when Sims introduced it in 1980,
27:45
but it requires the notion of an exogenous monetary policy shock, and there is no such
27:51
thing as an exogenous monetary policy shock. The best we can find are two or three tiny movements,
27:58
but measuring... And that's not even really the question. The question we want to know of
28:03
monetary policy is not, "What if the Fed sneezes raises 25 basis points?" The question we want to
28:09
ask in monetary policy, what happened in 1980 was not a sneeze, it was a change in the rule.
28:15
The change in the rule changed the expectations. So there's this one data point. Volcker raised
28:20
interest rates, big recession, inflation came down. But you look over history, there's other
28:24
totally contradictory data points. The end of the German hyperinflation, when the fiscal problem
28:29
was solved, interest rates went down, there was no recession, there was an economic boom,
28:34
they printed up way more money and they borrowed more money. What? Wait, that's totally opposite.
28:40
That's the central problem. I think it's an exciting time for research, because our models are
28:48
completely played out and we need to go back to absolute basics. Does raising interest rates lower
28:54
inflation and raise the value of the currency? If so, how? That's like the most basic question you
29:00
can ask. I think I know the outline on how to do it, but our current models do not do that.
29:07
And our current empirical procedures are... VARs, they're largely for forecasting,
29:12
they're used just for finding correlations. For causal analysis that's completely played
29:17
out and is answering the question we're not really interested in answering. And you said forecasts.
29:22
How well have central bank forecasts been doing lately? They've been a complete disaster. In part
29:28
has they ignore supply shocks, they ignore the whole real business cycle. They say,
29:32
"Oh, we had supply shocks." And then I say, "Okay, well where's the supply shock in your model?" "Oh,
29:35
we don't have any supply shocks in the model." "Well, why not, if you think it's a supply shock?"
29:39
Forget about [inaudible 00:29:40]. The last thing I wanted to complain about is you
29:42
said PhDs in central banks. It's not obvious to me that given the state of economic knowledge more
29:49
PhDs in central banks is going to be a good thing rather than a bad thing. Central banks remember,
29:54
never had PhDs. It was not a College of Cardinals or a theological disputation club. It had bankers
30:02
that had regular people and the Fed is set up to be a representative institution. It's
30:07
not obvious to me that PhDs are going to do any better, especially if they know too much of their
30:14
models and not enough of the historical doctrine matters. Doing policy is very different from doing
30:20
research. Doing research, your job is to have a hundred crazy ideas before breakfast on the
30:25
hope that one of them works out. Doing policy is about not screwing up. The loss function is
30:32
entirely different. Loss function for academia is, I can throw away my bad ideas. Loss function
30:37
for policy is, don't screw up even once. Or when are you going to get tenure.
30:42
Sorry? I feel like in academia
30:46
perhaps it creates a lot of risk-loving behavior in the sense that, not getting tenure requires
30:52
you to take a lot of risks on certain ideas. I don't know, getting tenure seems to require
30:59
taking very little risks and doing exactly what AER is likely to publish without too
31:04
much complaining. That's true.
31:06
But having influence after you get tenure and doing something important in life, which may
31:13
or may not mean getting famous in economic. Getting famous in economics is about getting
31:17
a thousand other people to do what you do and cite you, it's not necessarily about being right,
31:22
but long-lasting influence being right. Yes, is about taking an intellectual risk. Which is not,
31:28
if you do policy. Jon, you do policy someday, no more intellectual risks. You got to be very,
31:35
very... I think wise policy is very conservative. Even I, I just gave you a lecture about higher
31:42
interest rates, I don't know how they lower inflation. But you put me in charge of the Fed,
31:46
inflation goes up, I'll tell you what I'm going to do, I'm going to raise interest
31:48
rates. I don't know why it works. I don't know if it works, but until we know for certain something
31:55
else works, that's what you got to do. Well, I think it's quite ironic that Jay
32:00
Powell's idea or central idea or approach to policy, which he refers to is called the Risk
32:12
Management Approach to Monetary Policy. I think that's making it innately clear. It's kind of,
32:20
keep going until you break something type approach. It is definitely interesting.
32:27
I think monetary policy needs more of a risk management approach. They sort of say,
32:31
"Here's the forecast and here's what we're going to do." Rather than what our military
32:36
colleagues here at Hoover do, which is they teach us, you don't say, "Well,
32:38
the forecast is they're coming on the left flank, so that's where we'll meet them." You go, "Well,
32:43
what if it's the left flank? What if it's the right flank? What if they come down the center?
32:45
What if they come around?" You got to think much more about risks than the Fed typically does.
32:52
It's interesting. I don't know how you'd frame it in terms of risk,
32:56
but I think a related problem is really just group think. You go back to, for example,
33:05
October of 2021 after inflation had started to shoot up in May or April, May of 2021. At that
33:14
point by October 2021, it was very clear that it wasn't just your used car prices that were driving
33:22
the big jumps in CPI that we were seeing. We were starting to see things like shelter home prices or
33:28
owner-occupied rent prices increase, and even still the Fed waited another six months to start
33:37
raising interest rates. I think there was this idea that inflation was going to come back down
33:43
on its own and wouldn't need help from the Fed. This was just this idea that was being promoted
33:49
by the media. I really think that forecasters were moving away from what their VARs would suggest,
33:56
just that this supply shock was a one in a hundred-year type event. But even still,
34:05
I think independent thinking is a somewhat rare thing. Another thought is, I think just on sticky
34:11
prices and sticky wages, I've heard the stories of how it came back to modeling, how sticky prices
34:17
and sticky wages first got inserted into macro models in the first place. My understanding is
34:22
basically you had RBC models which said that there was no role for policy to have any real economic
34:30
effects. A lot of the early New-Keynesian modelers basically said, "We want policy to matter and
34:37
this is something that we can throw into the RBC machinery and make policy matter again." It wasn't
34:43
some sort of great empirical insight necessarily that it was sticky wages that were causing
34:53
massive fluctuations in unemployment history. There was some work, written by Bewley and others
34:58
afterward, but I think it's kind of funny how it was something that was inserted into models
35:04
to make them work the way that such modelers wanted them to work without necessarily a ton
35:12
of empirical basis. Just to push back a little bit on some of those natural experiments. I do think
35:20
especially for economic historians out there, I think there are great natural experiments to
35:24
look at when it comes to monetary policy. We have great instances of the money base being
35:31
cut by massive amounts. For example, I think in France there's a great François Velde paper,
35:39
there's the 1863 Currency Reduction Act in the Confederate United States where the money supply
35:49
was contracted and it could only be contracted in one part of the confederacy, the eastern part
35:55
of the Confederacy because the union forces had taken the Mississippi so the new grayback notes
36:01
couldn't make their way to the western part of the Confederacy, which was like Texas and so forth.
36:05
So inflation from your [inaudible 00:36:09] back prices actually went down significantly in the
36:14
Eastern part of the Confederacy versus the Western part. I think things like that are what you see
36:17
in that François Velde paper in France. I think these are obviously historical episodes that the
36:26
economy may look very, very different now, but I think they're interesting in [inaudible 00:36:33]
36:33
exogenous identification. Some would even say some of those monetary policy shock series,
36:38
whether it's from folks like Nick [inaudible 00:36:41] give you some exogenous variation,
36:44
breaking an exchange rate peg or stopping UI benefits or some sort of cutoffs across
36:51
states. I think these are interesting examples of good identification that can help us to answer,
36:57
at least begin to answer some macro questions or identify some [inaudible 00:37:02]. Sure might
37:02
be partially equilibrium, not perfect, but in my opinion gets us some answers maybe with some
37:11
precision, even though they might not be global. You've given me another laundry list here, so hold
37:16
your horses. First of all, I want to violently loudly agree with you on what you just said,
37:23
that I find lately historical episodes are the most convincing natural experiments we
37:32
have for distinguishing macroeconomic theories, as opposed to fancy statistical analysis with a
37:39
lot of assumptions and identified VAR by some identification restriction. I'll just want to
37:45
point up some that I've been making lots of pay in my own writing. I regard the inflation episode
37:52
that we just had, inflation came from seemingly nowhere. The Fed didn't do anything with interest
37:57
rates, particularly unusual. 8% inflation, where'd that come from? Then inflation went
38:02
away without repeating 1980. That's a historical episode. Well, the federal government printed
38:09
up $5 trillion of cash and gave it to people and the fiscal theory of the price level does
38:14
exactly that and the other theories clearly don't. There's a historical episode that I think helps
38:20
you distinguish theories. The other one, the zero bound era. When interest rates hit zero
38:27
and the Fed did massive QE, our existing theories made very clear predictions about what happened.
38:35
At the zero bound you have deflation spirals, in old Keynesian models you have multiple equilibria
38:41
in some spots in New-Keynesian models, and when you do QE you have hyperinflation in monetarist
38:47
models. The comparison of QE, which did nothing, and the Covid expenditures which caused inflation,
38:53
is just crucial. For Monetarism says those two have the same effect. Fiscal theory
38:57
says they don't. This is an example of historical episodes that are just screaming theory A makes a
39:05
clear prediction, theory B makes a clear prediction, the episode tells us what
39:09
happened, and yet nobody's paying any attention. The problem and I love Velde's work on a Chronicle
39:16
of a Deflation Foretold for your listeners that the Velde and Sargent Macroeconomics of the French
39:23
Revolution, another masterpiece, their work on US fiscal and monetary theory more masterpieces.
39:30
I think that's incredibly useful work. The problem right now is sort of our professional
39:36
problem, it doesn't look very technical. You can't impress hiring committees and econometrics
39:42
editors with all the equations you can run, you can just convince people of stuff that's right.
39:46
It's like Tom Sowell's work, it doesn't have lots of regressions, it's utterly persuasive
39:53
by assembling facts. Now maybe hopefully economics will at some point produce reward
40:01
things that are actually producing knowledge, but I totally agree that's kind of where my
40:06
own thinking is going to try to send anything out. You said something about central banks not having
40:14
independent thinking, and that's true. There is kind of this bubble where they talk to each other.
40:21
It's funny because the Fed is set up so much as a responsive structure. The regional banks,
40:27
the FOMC members, they're not supposed to be there to represent schools of thought, they're supposed
40:32
to be there to represent communities. And again, it's just stuck. I think the greatest example is,
40:39
and the disadvantage of having too many PhDs around, is just one, you mentioned one episode,
40:45
but the episode earlier, the flexible average inflation targeting is a beautifully constructed
40:51
New-Keynesian DSGE Maginot line against the supposed problems of deflation at the zero bound,
41:00
which we cure by making promises about what we will do 10 years from now if inflation should
41:06
ever come. The promise of if inflation ever comes 10 years from now we'll be slow to react will be
41:12
the thing that stimulates us today. That was just New-Keynesian DSGE doctrine. It took over within
41:19
the central banks and the original Maginot line, sorry, it was the wrong problem at the wrong time.
41:28
It's a great example of both the kind of group think, but how the group think was affected by
41:35
I would say too many PhDs and not enough people saying, "You got to be kidding about this stuff."
41:40
And last, you said sticky prices. This was a very deep thing that you said. Now sticky prices,
41:48
everybody recognizes the models of sticky prices are kind of silly. This is a problem
41:53
we have in economics. We're like the drunk who looks for our keys where the light is not over
41:58
by the car where we drop the keys. Because we kind of sense, like the historical experience,
42:06
that inflation causes output booms and deflation is sort of associated with recessions and we don't
42:13
know why. So making prices sticky artificially produces models that do that kind of thing and
42:22
lets you explore it. You also mentioned it helps to sell things to central banks. Here
42:30
we just noticed our last two recessions have had nothing to do with monetary policy. There was a
42:35
financial crisis and there was a lockdown that caused the recession. You can't ask for better
42:41
real business cycle view of the world. But you are right that telling central
42:46
banks they're not very powerful is not a really good way to get yourself invited back to central
42:50
banks. It is funny that the whole macro has so totally focused on monetary, but we'll kind of
42:59
agree monetary policy might 1%, 2% here and there, but sort of the central things that
43:05
cause big economic perturbations are not any more monetary policy. Friedman turned out to be wrong,
43:10
maybe not about history, but about current events. Recessions are not caused by monetary
43:15
policy mistakes. This isn't just doctrine, this is what you can see in every empirical estimate,
43:20
even your VARs, monetary policy shocks just aren't that important for recessions.
43:26
In fiscal theory, the Fed on the edge is helping here and there, but fiscal policy is really
43:32
important. Milton Friedman won far too much. In the fifties people thought inflation had nothing
43:39
to do with central banks as wage price spirals and stuff like that. Then Milton Friedman goes,
43:44
"It's all central banks," and now recessions are no longer all central banks and inflation
43:50
we're starting to discover is not all central banks. It's a problem that the customer with
43:58
the deepest pockets wants a theory where central banks are really powerful and it's not obvious
44:03
they are. Okay, stop having so many beautiful ideas and I'll stop having such long responses.
44:09
Well here's another, I guess the last just couple monetary policy questions. I'd love to talk to you
44:14
a bit about economic growth and regulation and other growth related topics as well.
44:21
Let me just interject. Monetary policy, I do it because I think it's interesting,
44:25
but economic growth is the important issue of our time. The falling in economic growth is the
44:32
tragedy, is the Europe's end of economic growth is the tragedy. And it really has
44:35
nothing to do with the Fed and monetary policy, it's about regulation, innovation, dynamism,
44:41
all those things. So yes, that's the important question. We're off, we're the drunk looking
44:46
for the car keys over where the light is and not where we dropped them, economic growth.
44:52
Absolutely. Before I get to growth, one last question on just monetary policy. One
44:57
is, you mentioned FAIT or Flexible Average Inflation Targeting. We've had just inflation
45:05
targeting in the past 30 plus years, and the history behind it I think is really interesting.
45:12
For those that aren't really familiar with how central banks were targeting things,
45:17
there was a period of time at the height of Milton Friedman's thinking
45:20
in the 1980s when central banks were trying to target monetary aggregates,
45:24
they were taking MV equals PQ very seriously. Then they realized they kind of couldn't really
45:30
target it or control M very well, in part because money demand is a bit unpredictable.
45:38
From there that's kind of how this New-Keynesian machinery kind of emerged, or the whole concept of
45:45
targeting an interest rate as well as targeting an inflation rate emerged. It was really the
45:53
central bank of New Zealand, the Reserve bank of New Zealand that began inflation targeting
45:58
and chose this 2% level of targeting, and then other central banks around the world followed,
46:07
the Fed started following but didn't officially declare their 2% target until actually the early
46:13
2010s. But the whole world started following this 2% idea by and large by the end of the 1990s.
46:24
What's interesting is, in those debates, if you look back, there were a number of people
46:28
like for example, Paul Volcker who is very much against 2% was in favor of a 0% inflation target.
46:36
Paul Volcker being president of the Fed or the chairman of the Fed for quite some period of time
46:40
in killing the inflation of the 1970s, 1980s or at least being credited. But with that maybe there's
46:44
more of a fiscal policy story there that you could tell. But other central bankers like John Crow
46:50
who is the governor of the Bank of Canada was in favor of 0% inflation targets. I know John, you've
46:57
spoken quite a bit or written quite a bit about not being as worried about inflation as we should.
47:03
Let's say some recession hits the US economy in the coming years and the Fed drops interest rates
47:11
back to the zero lower bound and say we start getting worried about deflation again. What
47:16
would you say to those people in macro circles? Oh gosh. Once again, there's the whole history
47:22
here. Yes, central banks, there's a question of the target in the instrument. What do central
47:29
banks want to control? Are we driving to Las Vegas or are we going to go to Santa Barbara? Then what
47:36
do they actually is the steering wheel, the money supply or the nominal interest rate? Those are
47:42
central questions that we've changed around on over the years. As you mentioned, it used to be
47:47
the money supply that turned out not to work. Central banks went back to what they've always
47:50
been doing, which is using the nominal interest rate to try and steer inflation where it's going.
47:59
The other question is the set of goals. Should central banks worry about one thing, inflation
48:05
or a whole bunch of other things. Our official mandate in the US is inflation and employment,
48:12
but there's a whole bunch of other things people want central banks to do as well, exchange rates
48:16
and inclusive employment and climate and all sorts of things now. Just quick on the history of this,
48:28
in the 1980s the US Fed and other central banks went back to interest rates and felt
48:33
that higher interest rates was what you do to control inflation. In the early 1990s,
48:38
a bunch of countries, starting with New Zealand, innovated with the idea that the
48:44
central bank would not try to do five things at once, inflation, unemployment, all sorts of other
48:49
things. They would just worry about inflation. I think this is very wise, the best way to raise
48:56
employment within the constraints of what monetary policy can do, is to just have very
49:02
stable inflation. It's when inflation goes up and down and then the Fed has to fight it all
49:06
the time that you get fluctuations in employment. By the way, it is kind of weird that the Fed is
49:13
supposed to be in charge of inflation, but it can't do the one thing that creates inflation
49:17
or gets rid of inflation, it can't give people money and it can't take away people's money.
49:22
The Fed is supposed to worry about employment, but god forbid it should talk about tax rates,
49:27
employment policies and so forth, which is actually what determines employment. Anyway within
49:32
the constraints of what it can do. What happened was, New Zealand and the other countries, it was
49:37
a political deal with the central banks involved, was you worry about inflation, we will evaluate
49:43
you by how you do on inflation and nothing else. I think also they always included fiscal and micro
49:50
economic reforms, which is actually an important part of the deal.we by the way, will pay our debts
49:56
and we won't count on you to print up money to pay our debts. That was a crucial part of the deal,
50:00
and we'll get rid of a bunch of regulations. What happened was not, you mentioned the Volcker
50:05
experience, high interest rates and a big recession to combat inflation. The inflation
50:09
targeting countries got rid of inflation big overnight. The idea was it was supposed to
50:14
insulate central banks from political pressure not to raise interest rates, so they could be tough
50:19
and repeat Volcker if they had to, and they never had to. It was just instantly inflation went away,
50:25
sort of like the end of the German hyperinflation. That was a big success. Through the nine formation
50:30
of the Euro central banks were told, "This is beautiful, just worry about inflation,
50:34
don't worry about anything else. Everything else will take care of itself if you just worry about
50:38
inflation." Now we've slid back on that a lot. Central banks are now doing all sorts of other
50:45
things beyond just targeting inflation. But we are left at least in the doctrine and also
50:51
we're starting to realize central banks maybe don't have as much control with it.
50:55
We're left with the doctrine of exactly what do you mean by target inflation? You're right,
50:58
2% was picked out of a hat. I personally like zero, but our Fed's mandate is price stability.
51:08
The European central bank's mandate is price stability. What do you think they meant when
51:13
they said price stability? Do you think they meant 2% inflation forever or do you think they
51:18
meant price stability? That's legalism, but then there's the economic question,
51:22
which I don't know how deep you want to, I kind of like price stability, in part because
51:27
I'm not that concerned about adding stimulus from the central bank. The central banks just
51:32
not screw up. That's good enough. In the US, it's price stability
51:36
and full employment- And full employment.
51:38
[inaudible 00:51:40]. The price stability part is interpreted as 2%
51:42
inflation. Now the theory is it's the same theory as the theory that it feels good if you wear shoes
51:48
that are too tight all day long, because it feels so good to take them off at night. That if it's
51:52
2% in normal times then you can have higher interest rates and lower them more. All right,
51:56
maybe, maybe not. I don't think so. I guess the divine coincidence idea
52:00
that you're just targeting inflation, you end up- There's a even bigger question of what do we mean
52:05
by this 2% inflation? This really comes up with the past inflation that we've had. If say you have
52:11
a 2% inflation target, does that mean inflation is on average 2% over long periods of time? You might
52:17
think yes, absolutely no. Because they always catch it in a forward-looking direction. It's
52:23
2% where mistakes are forgiven. Your impression of a 2% inflation target might mean, "Well, if it
52:30
goes up 8% in one year, you got to have several years of zero or 1% to bring it back to 2% over
52:38
the average." Uh-uh, central bank's interpretation of price stability is 2% inflation forever, and
52:44
if we ever screw up, we just forget about it and try to get back to 2% in the future. But we bring
52:50
the level of prices back to where it was. That's actually, that's the deepest part of the 2% target
52:58
that I think bears some worry. So yeah, I think zero would be fine. Price stability
53:05
should mean price stability. But that is that last remaining part of the debate still.
53:11
Well, speaking about your books, and it would be remiss not to plug your,
53:17
perhaps dare I say magnum opus, The Fiscal Theory of the Price Level that came out in 2022. You're
53:24
also working on a new book, a related one on my understanding is, of fiscal theory, but in
53:33
the Euro area. Can you explain to just a little bit about that and the idea behind that book?
53:37
I'm very excited about this. I'm working with Klaus Masuch at the European Central Bank and
53:42
Luis Garicano who is, let's see, LSE, University of Chicago and an ex-member of the European
53:48
Parliament great theorist. We are working on it. We're finishing up due July 15th, a manuscript
53:57
which will be available on my website for free until it gets published by Princeton
54:00
in the spring, on reforming the structure of the Euro. Of course the central problem of the Euro,
54:08
it was set up quite clear-headedly, really amazing the things they thought about compared to standard
54:15
monetary doctrine at the time. The problem of a common currency without a fiscal union,
54:23
is that of course each individual country has an incentive to borrow a lot of money and say,
54:28
"Help, help, help. Print money to bail us out." So the ECB and the Euro was set up
54:35
very thoughtfully to try to stop that problem. The ECB would not buy sovereign debts. The ECB
54:42
would not work to hold down spreads, and countries were supposed to follow debt and deficit limits
54:49
to not get in trouble in the first place and we'd solve that problem. Well, not really.
54:56
In part because when putting it together, they didn't really want to write a prenup.
55:00
It would've been impolite to say, "Now if you do feel like defaulting, let me remind you that
55:06
you are now just like a country you default," and to put in a bankruptcy and resolution mechanism,
55:13
which they didn't do. That was a beautifully set up separation of monetary and fiscal policy,
55:20
and it kind of fell apart. Now, it fell apart in the most human of all ways. Things
55:26
that nobody ever imagined in fact happened. The financial crisis happened. Nobody was thinking
55:32
about financial crisis. Financial crisis happened and the ECB started integrating.
55:36
The sovereign debt crisis happened, nobody thought there would be sovereign debt crisis
55:40
anymore. The ECB started buying Greek bonds. QE happened, and the ECB decided,
55:46
"Well, so what? We don't buy sovereign debts. We're going to buy lots of bonds and print up
55:49
lots of euros." Covid happened, the ECB bought more. Step by step, we're now in a situation
55:56
where every time there's a hiccup in the Italian spread, the ECB starts buying Italian debt. We're
56:03
economists, it's about incentives, the incentives for countries to run good fiscal policies,
56:09
for people not to expect the ECB to intervene routinely are gone. That basic structure of the
56:16
Euro, the monetary and fiscal part of the Euro... And it's banking too. By pretending sovereign debt
56:22
was risk-free, it means banks are loaded up with sovereign debt, so banks are like hostages now.
56:27
Because if Greece fails, all the Greek banks fail. Why? Because banks are allowed to load
56:33
up on sovereign debt as if it's risk-free, in a way they can't load up [inaudible 00:56:36]. Well,
56:36
that's nuts, but nobody ever got out fixing that. The book points out this beautiful structure that
56:41
was set up, how it fell apart in the exigencies of four crises. If you put me in a crisis,
56:48
I'm going to bail everybody out just like they're going to... You got to get through
56:51
the crisis. But then nobody fixed the structure in between the crisis. They just said, "Oh great,
56:56
a river of money solved it, so let's go get ready to print up another river money."
57:01
We're at the point now where Europe needs to put that genie back in the
57:06
bottle and get some structures that will get the incentives right again,
57:10
not for everybody not to count on the ECB always to print money to solve every problem, because
57:16
it won't always be. That's the short summary of the book and wait... Well, this probably won't
57:22
come out until July 15th. July 15th the draft will be up on our website and look forward to
57:27
it next year from Princeton University Press, and many more podcasts about this wonderful effort.
57:34
Well, I'm very excited to read that. That's a super exciting development. So lastly, just want
57:42
to talk about economic growth. The question I think is perhaps the most central question
57:50
in the field of economics. Why are countries rich and poor? Why are countries growing or
57:57
why do certain countries have levels of GDP per capita that are much higher than others?
58:02
Let's not solve all the world's problems in one podcast, Jon. I think what we've done
58:07
for our listeners is tee up the many issues that we'll continue talking about. Along with tariffs,
58:14
that's probably going to be in this soon. Industrial policy, that's going to be in
58:19
soon. How much money should the government throw on electric vehicles, especially made in the US
58:26
by union labor? We got a whole bunch of issues coming to talk about over the coming years.
58:32
Absolutely. Well, this is really such an exciting opportunity and Hoover has such a
58:38
great cater of scholars who will be appearing on the podcast soon. And of course, really wonderful
58:46
to have you on, John, as always. I really want to thank you so much for joining us today and
58:52
also for this wonderful collaboration between the podcast and the Hoover Economic Policy Working
58:59
Group. Really excited to be part of the team and really excited to continue to engage on a lot of
59:05
these same ideas that we've been talking about. This is going to be great fun, Jon. I hope
59:10
all of you tune in to the new home of Capitalism in the Twenty-First
59:15
Century at the Hoover Institution, EPWG Wonderful. Today our guest was John Cochrane,
59:23
who is the Rosemarie and Jack Anderson Senior Fellow of the Hoover Institution and the chair
59:28
of the Economic Policy Working Group. This is the John Cochrane podcast, where we talk about
59:33
economics, markets and public policy. I'm Jon Hartley, your host. Thanks so much for joining us.