YT YouTube2Text

YouTube2Text → Transcripts → Live - ECB Annual Research Conference 2026 (Day 2)

Live - ECB Annual Research Conference 2026 (Day 2) — Transcript

by European Central Bank · 23,149 words · 3,384 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

1:39Hey,
4:05Heat. Hey, Heat.
9:55Good morning and welcome to the second
9:58day of the ECB annual research
9:59conference.
10:01We we start this morning with with a
10:03highlight of each uh conference which is
10:06the Jean Manet lecture. So today is the
10:1011th uh Jean Manet lecture. Uh and of
10:13those 11 uh seven have won the Nobel
10:16Prize and I think there's a fair chance
10:18the other four uh will will soon be in
10:21that category. So this is really you
10:23know an amazing experience both for the
10:26people in the room and online to listen
10:28to really some of the most brilliant uh
10:31economists you know uh uh working today.
10:34So I'm delighted that Paul Krugman uh is
10:37uh uh accept accepted the invitation to
10:40to present uh this year. Uh his topic
10:43will be economic size and economic power
10:46which is very much in line with the
10:48overall theme of this year's conference.
10:50But before turning over to Paul, let me
10:52just uh try and uh frame uh Paul's
10:55contributions from an ECB point of view.
10:58So of course um we're we're you know all
11:01the time super interested in the
11:03international trade and economic
11:05geography topics uh which were cited in
11:08in his uh Nobel his first Nobel award.
11:11Um but I would let me also emphasize
11:13from the ECB point of view uh his work
11:16uh if you like on in terms of the
11:18foundations of international monetary
11:20economics have been so important. Uh so
11:23whether that's uh currency crisis which
11:26by the way can uh with some uh mutation
11:28uh also be thought about in terms of
11:31crisis in bond markets whether it's
11:33liquidity traps whether it's the uh pass
11:37through of exchange rate shocks to to
11:39prices all of these are topics we we
11:42turn to all the time and of course in
11:44terms of the uh economics of monetary
11:47union and the economics of the euro uh
11:50Paul has made a very foundational
11:52contribution ions. Uh so if you go back
11:54to the early 90s when people started to
11:56write about the monetary union, you
11:58know, I encourage uh people to go and
12:00and uh reread it. I actually did it
12:02yesterday just to uh in preparation uh
12:05lessons of Massachusetts for EMU. So
12:08that's from the early 90s. Uh and then
12:12you know in 2012 in the NBOR macro
12:14annual he he had a piece called revenge
12:16of the optimum currency area which
12:18basically you know tried to reflect
12:20about the Europe Euro crisis in the
12:23context of the the exanti debate about
12:25the euro. In that uh short paper he made
12:29some recommendations for what is needed
12:32for the euro uh to to to prosper and to
12:35and to uh improve. uh some of those
12:38reforms have been made uh not all of
12:40them but but I I think it's very
12:42insightful piece. Let me say also in
12:46addition to all of those academic
12:48contributions I I suppose Paul um is
12:51known as an educator that includes his
12:54his kind of uh textbook contributions.
12:57uh Paul Samson at some point wrote, you
13:00know, had a quote something like, you
13:01know, let other people write a nation's
13:03laws or policies, you know, so long as I
13:06can write the textbook, you know, that's
13:07the most important job. And, you know,
13:09in many ways that that is, but of
13:12course, uh really going back to the mid
13:141990s, uh Paul has been a you know, an
13:18innovator in in finding new ways to
13:20communicate economics. So I can remember
13:23during the Asian crisis uh reading his
13:25his blog then you know his work for
13:28Slate and the New York Times and now
13:30these days on Substack of which I'm
13:33happy to hand over seven a month uh or
13:36$7 a month uh to to get the full access.
13:39Um but it it's so important that people
13:42not just write the uh foundational basic
13:46research that we heard yesterday but are
13:48also able to use um the economic
13:50knowledge we have. So if you like in the
13:53research we saw yesterday maybe we'll
13:55add 1% a year to the stock of economic
13:58knowledge but let's use the 100% of the
14:00stock of economic knowledge to analyze
14:02the policy problems we have. So an
14:05ancient term in economics is political
14:07economy. So this is you know Adam Smith
14:09was a political economist saying that
14:12you cannot unbundle economics and the
14:15institutions and the politics and I
14:17don't know if Paul would self-describe
14:19as a political economist but but that's
14:20the label I'm going to uh attach him
14:23which is a very proud tradition in
14:25economics. So uh Paul also at the ECB
14:29was uh at the first centra forum uh in
14:312014 he was one of the speakers. So in
14:34other words uh when we looked around at
14:36that time uh say okay who are the best
14:39possible people uh to invite uh to to
14:41the center forum Paul Paul was there. So
14:44it's been uh not always an easy message
14:47but you know we we want to hear uh from
14:50people like Paul uh in the most
14:52straightforward fashion. So having said
14:54all of that h the format will be I will
14:56now invite Paul to to deliver the
14:58lecture and then we will have some Q&A
15:01after that. So, uh, let's welcome Paul.
15:05[applause]
15:11Well,
15:13uh, so thank you, Phillip. Thanks, uh,
15:17to the ECB for having me here. Thank you
15:19all for listening. Um, just a quick
15:22story. I decided to do international
15:25economics
15:26alarmingly 50 years ago um and graduate
15:30school and at that time particularly
15:33people advised me uh don't do
15:37international trade because it's boring
15:40nothing happens and there are no
15:41exciting things occurring uh in the
15:44field of international trade and that
15:46luckily I didn't take that advice and
15:48things are certainly not boring. Um the
15:51topic of
15:53this conference and of a lot of the
15:56recent literature uh is
15:59it it's aside from obviously being
16:01incredibly important. It's also it's
16:03intellectually exciting. Uh it's not
16:06that much that we're using we're seeing
16:08some new tools. I was actually
16:10particularly delighted to see people
16:12making use of new data sources, new ways
16:14of evaluating, but also we're just, you
16:16know, new questions. Um, and um, one of
16:20my old teachers, Essie Dmar, said that,
16:23you know, bad times make for good
16:24economics. And I guess that's kind of
16:27what's happening now. We're getting a
16:29lot of interesting work. Unfortunately,
16:31the the side effect of that is that we
16:33have the bad times to worry about, but
16:35anyway, here we are.
16:37um when I was trying to put together
16:40this talk is going to in a way be two
16:43talks because there are there's uh the
16:45first part is going to be a my
16:47understanding of where we are in terms
16:50of how to think about geoeconomics which
16:52I have been frantically trying to get up
16:54to speed on. Um the second part is going
16:57to be about uh size and power issues
17:00which I I believe are are becoming
17:02really critical and we're the subject of
17:05a lot of what we heard yesterday. Uh and
17:08I will offer I think a somewhat
17:10different take or at least a a way to
17:12look at it and I think there are some
17:14there are some surprises in all of that.
17:17Um, so when I decided I I [sighs]
17:22I don't know what how would you describe
17:24my current role? I think I think I've
17:27become an influencer. Uh, that's what I
17:30do these days. I write these daily uh po
17:32daily posts. Um, and one of the things,
17:35you know, uh, with with that kind of
17:39domain is that you always have to have a
17:42a picture, an eye-catching picture at
17:46the beginning. That's uh, that's comes
17:48from blogging days. You always, you
17:50know, start with something that will
17:51just because no one is forced to read
17:53it. So, you have to have a something
17:54that will catch people's eye. And I
17:56tried to think, how do I come up with a
17:58picture of weaponized interdependence
18:00and and geo geoeconomics? Uh and the
18:03answer is well I don't know if I'm
18:05ashamed or not but I I do read science
18:07fiction and uh this uh I don't know how
18:13many people have read uh the classic
18:15science fiction novel Dune which was the
18:18subject of sever several terrible movies
18:20and then finally Denive got it right u
18:24and the um the premise of the novel is
18:27that there's this planet Arachus which
18:30is the sole source of something called
18:32spice which is essential to the galactic
18:36economy and so everything relies upon it
18:40and so basically Arachus is the straight
18:42of Hormuz with giant sandworms. Uh it is
18:45really very much relevant to current uh
18:49issues. By the way, um the struggle for
18:52control of the spikes um leads initially
18:56to, you know, um weaponized
18:58interdependence and eventually to
19:00outright war, which is kind of where we
19:02want to probably end today's talk. Okay.
19:06Um let me give you two quotations.
19:10Uh the first is from the novel, from the
19:12protagonist of the novel. Um he can
19:15destroy a thing controls the thing. uh
19:18and the second is a rather more verbose
19:21statement uh from Albert Hirschman who
19:24is the patron saint of geoeconomics
19:27uh about how basically gains from trade
19:30are also uh potentially a tool of
19:33coercion. Uh and that's really this it's
19:35really the same statement. Um they're
19:37really saying the same thing. Uh now
19:39what is interesting about both of those
19:42is that both of those claims whether
19:45it's the character in the novel or
19:46Albert Hirshman is basically
19:50implying that you have to be willing to
19:53yourself pay the cost of the
19:55destruction. And that is I think one of
19:58the really big gaps that we will have in
20:00in trying to understand where we're
20:03going with all of this uh geoeconomics.
20:05I guess this I guess um Dune is not
20:08really it's cosmoe economics I guess
20:10rather than geoeconomics since it's
20:11inter interplanetary but anyway um okay
20:16[sighs and gasps] how do we think about
20:17this um so [sighs and gasps] it's a
20:21dirty little secret among people who do
20:23international trade theory that it's a
20:26general equilibrium field that in some
20:29ways it is the general equilibrium field
20:31everything needs to add up and we all
20:34know that we have to do uh consider uh
20:39interlocking markets, comp competition
20:41for factors of production, all of that.
20:44And yet most of the time when you want
20:47to think about policy issues, you end up
20:49with supply and demand. You end up going
20:51partial equilibrium. Uh and so that's
20:54what I ended up doing, trying to come up
20:56with my own version. And I was both
21:00hopeful and afraid that some of the work
21:05uh presented at this conference would
21:09[sighs and gasps] you know make
21:10basically make this look silly by coming
21:12up with a uh a true general general
21:15equilibrium approach to uh to
21:18geoeconomics. Um there is a known
21:22familiar
21:23cluge a way to do partial equilibrium
21:27while saying that it's general
21:29equilibrium which is to presume that
21:32there is a an outside good that is
21:35somehow not the subject of policy and is
21:39perfectly competitive and produced at
21:41constant returns. And uh we all know
21:44that that is it's a dodge. It's a way to
21:47deal with it. Um, and you know, better
21:49to do that than to give up on any kind
21:52of comprehensible framework. And I was I
21:56I thought it was a terrific paper by uh
21:59Yang and Lou yesterday on power and uh
22:02uh but I was both reassured and
22:05disappointed to see that they did the
22:06same thing. Uh they use the same cluge
22:08to deal with this and so so we can do
22:11partial equilibrium. Um it's
22:14there's a good some good uh that we
22:19imagine one country exports to another
22:21and of course then we can imagine that
22:22there are multiple goods uh and the
22:26weaponized interdependence is that a
22:29country can threaten to cut off this
22:32trade. If we do a standard
22:36marshalian whatever we do uh consumer
22:38producer surplus essentially uh we would
22:40say that ordinarily we would expect that
22:43there are gains to both sides. The
22:46exporter of the good gains a producer
22:48surplus which is the the the bottom
22:50triangle. the the importer gains the
22:53consumer surplus and the threat of uh
22:57weaponized interdependence is that well
23:00we can cut off those gains and they will
23:03uh they cut off the gains for both sides
23:07um which do we think is more important
23:10as economists I think although it would
23:14take some it takes um uh you know a lot
23:17of quantification to be sure but almost
23:20surely most of the time uh the the the
23:27pink orange whatever the the pale red
23:29triangle at the top is going to be
23:30bigger. It is much more common for
23:34countries to be really dependent upon
23:36imports of something and to have no good
23:39alternatives than it is for exporters to
23:44have nothing else they can do. Not just
23:47because there may be many exporters but
23:50or other things you can export or other
23:52markets but also because resources can
23:54be re reallocated. So although there may
23:57be a significant number of people
23:59employed producing for a foreign market
24:02uh they could be doing something else.
24:04Um and so we we generally would tend to
24:07think that the uh it's strictly speaking
24:10it's the slope not the elasticity. But
24:11anyway, we we typically think that the
24:13elasticity of demand for imports is
24:16substantially lower than the elasticity
24:19of supply of of exports and that the the
24:21welfare losses from losing access to an
24:24export market will typically be low. Um
24:29two things um one is that people in
24:33actual positions of power if anything
24:36think the opposite. So we have and you
24:41know it's I try to not be too political
24:44here but obviously we have an unusual uh
24:48rather baffling trade dispute between my
24:50country and its northern neighbor. Uh
24:53and the position of my president is that
24:57we don't need Canada for anything that
25:00there are no there are no benefits for
25:02to importing uh oil and aluminum and
25:06electricity and and so on. um from from
25:09our from Canada and that we are
25:12subsidizing Canada and that all of the
25:15benefits from the trade are the bottom
25:16triangle here. Uh okay. Uh that is not
25:21standard uh economics. Uh it is
25:26uh for the most part probably uh I do
25:30worry that we may be a little too glib
25:32about these uh about dismissing the
25:36producer surplus side or the the export
25:38side. Uh partly that's because there are
25:42situations in which the resources that
25:44are being used to produce and export do
25:46not have uh readily available
25:50alternative uses. And if you want an
25:52actual example that is relevant to this
25:55current thing, um, uh, at theabaska tar
25:59sands, uh, a, uh, Australia, I'm mixing
26:03up my my, uh, my my, uh, English
26:06speaking small nations, uh, Canada. Uh,
26:09there at the moment there is no real,
26:13uh, other place. uh you you oil is being
26:17extracted in Alberta uh and is shipped
26:19to the American Midwest through
26:21pipelines and it really you can there's
26:24there is a limited pipeline capacity
26:26that can take it off to the Pacific
26:28Ocean but not much else. So the there is
26:30actually in this sense a a producer
26:33surplus that's being generated by the
26:35ability to export the oil and um and a
26:39cut off of that trade would hurt
26:41Canadian producers. So there would be
26:43some welfare loss there. Uh
26:46time is always a factor and one of the
26:49things in all of these is that given
26:51sufficient time both of these lines get
26:54much flatter. So the Canadians are in
26:57fact working on expanding their pipeline
27:00capacity to the Pacific. Uh so that will
27:03be less um less of an issue. Uh other
27:06thing that is worth saying uh that's
27:08also true on on the on the demand side.
27:11Um, one of the things, you know, all of
27:14this craziness, but the uh one of the
27:16things that we're witnessing happen uh
27:19in the current uh oil crisis is that the
27:24um on day one there was really no
27:27alternative to the straight of Hormuz.
27:30Uh as time goes by,
27:33uh you find ways around, you find ways
27:35to uh to substitute, you find
27:38alternative routes. So the the uh in
27:40some sense the elasticity of demand also
27:43rises over time. So that that the power
27:45that comes from the ability to disrupt
27:48trade is is a wasting asset. You can use
27:52it for a time but not not indefinitely.
27:54Um the uh uh [sighs and gasps]
28:00one point to make also we we are I am
28:04going to be talking about sizes of
28:05countries and power that comes from
28:07size. um small countries, you know, the
28:12classic the small open economy that
28:15produces a homogeneous good that is uh
28:18traded on world markets has no ability
28:20to influence world prices presumably
28:23does not have power in in world trade.
28:26Um but another one of the dirty little
28:28secrets of international economics is
28:31that to a first approximation there are
28:33no small economies. That even economies
28:36that are um uh appear to be you know
28:41that are are in fact very small uh
28:43produce differentiated products that we
28:46uh and we kind of know that from the in
28:49part just by looking at the world but in
28:51part just by knowing what happens when
28:53you try and model world trade. try and
28:55model international patterns of trade
28:57and for now I don't know how many years
28:59ago but we've you know cgee models have
29:02been armingtonized from the beginning
29:04you always have to just assume that
29:05countries produce differentiated
29:07products trying to actually model in a
29:10more fundamental way patterns of trade
29:13you end up doing something like eaten
29:15cortum where you end up with randomized
29:17technological advantages everybody
29:19almost everybody is um is uh to some
29:24extent has has significant market power
29:26in trade in some limited respects. Um,
29:30when I was a student, I everybody my age
29:34you've heard of is was a Rudy Dorbush
29:36student. Uh, and one of the many
29:38instructions that Rudy taught me that
29:41has been very useful in life is when
29:43you're not writing for graduate
29:45students, you never begin consider a
29:48small open economy. You always begin in
29:50Belgium. And uh but Belgium is not in
29:53fact a small economy in that sense. Uh
29:55it is in fact producing differentiated
29:57products and even the example I've been
30:00using of uh Canadian oil. I mean oil is
30:04a globally traded commodity. We do talk
30:06about a world price of oil but in fact
30:09um I did say that the Canadians really
30:12have at the moment no alternative. Um
30:15they their oil must pretty much be
30:17shipped to the US Midwest because that's
30:19where the pipelines run. It's also true
30:22that the US Midwest has basically no
30:24alternative to the Canadian oil. There's
30:26a lot of oil being produced from the
30:27Peran basin. Uh but the pipelines don't
30:30run that way. Uh and also it turns out
30:33that oil is not oil that uh that the
30:35Canadian oil is heavier and the the u
30:38the u the Texas oil is too is is light
30:42sweet crude and they need for certain
30:44things they need the heavy uh icky
30:47sticky stuff that comes from Canada. So
30:49there there are everybody has some
30:53market power pretty much even uh presum
30:56I don't know there must be uh maybe
30:59grain exporters though I suspect that if
31:01I got into the wheat market I would turn
31:03I would learn that wheat is not the not
31:05wheat that there are different varieties
31:06and all of that so everything matters um
31:10okay um everybody has uh market power uh
31:16although it it's a wasting asset if you
31:18tried to use it. Uh so everyone has some
31:21coercive power that comes out of global
31:24trade. Uh nonetheless,
31:27size clearly matters. Um big countries
31:31are
31:32more likely to produce varieties that uh
31:36where there are few if any competitors.
31:38They're more likely to you know just
31:41they just plain have more stuff. So
31:43there's more things that the world
31:45depends upon them for. And to the extent
31:47that we worry about the producer
31:49surplus, they have bigger markets. So
31:53size is clearly uh an an important part
31:56of this discussion. Um and it showed up
32:00in um uh in uh Steve Reading's paper
32:05yesterday, the talk about the role of uh
32:08that they basically had to have a a size
32:11factor uh for for the um for for the
32:15matrices and we're able to indogenously
32:18uh get the weight on it but the factor
32:21was GDP and it turns out that the
32:23elasticity of power with respect to GDP
32:25is 78 which is kind of interesting. uh
32:28but my immediate thought was which GDP
32:33and I'll get to that in a second. Uh the
32:36other is we've had some discussion about
32:38you know we were talking about hegemonic
32:41uh systems of world trade and what
32:43happens when
32:46we have a flip and the United States is
32:48no longer the dominant economy and uh h
32:51how does that affect the system but
32:53that's actually that's not a prospect
32:56that has by many measures already
32:59happened and a lot of what we're trying
33:01to understand now is what what we do in
33:04a world where that has already happened.
33:06So, [sighs]
33:08uh many of you have probably seen me
33:11writing about this because I this is
33:14both important and um and also I think
33:18intellectually very interesting. Uh so
33:20one of the one of the ways we can
33:22compare economies is
33:25purchasing power of their GDP. And uh
33:29now you might say why not just use
33:32dollar values and the answer mostly is
33:35simp is just instability. Uh if you are
33:39going to compare the uh euro area with
33:41the United States uh it looks as if
33:45something terrible happened to Europe uh
33:48after 2008 but mostly it's just that the
33:50euro declined against the dollar. It's
33:52not it's not mostly a real movement in
33:54the in the two economies. So you we
33:56basically use something like purchasing
33:57power parody to to smooth it out. Um and
34:02if that's the criterion we use
34:06uh well there's a clear number one
34:08economy in the world and it's not the
34:10United States. Um and if we do the US
34:14European comparison
34:16uh actually
34:19basically there there are three economic
34:21superpowers in the world by this
34:22measure. There's the United States,
34:24there's China, which is clearly number
34:26one, and then there's the US and the EU,
34:29which are roughly co-equal. Uh not the
34:32world as many people uh not the world as
34:35the US commerce secretary would portray
34:37it, let's say, but it is the world that
34:39that the standard numbers do. And um one
34:42thing that I
34:45sometimes sometimes people ask, you
34:46know, okay, how reliable
34:49are these purchasing power parity
34:51comparisons? And the answer is look,
34:53they're hard. Uh it's a lot of work, but
34:56a lot of work is put in. This is stuff
34:58is coming out of the the World Bank's
35:00international comparisons program. And
35:02so this is not a casual enterprise. This
35:04is something there where people have put
35:06a lot of effort into trying to to make
35:08these calculations. And they mostly
35:10don't I think I'll show you a slide in a
35:12few minutes. They they mostly uh are in
35:15accord with other measures that you
35:17would think would get the same thing. So
35:18this is in one sense by one measure
35:23which in some ways is kind of what we
35:24think is should be the right measure you
35:26know just how big are these economies
35:28how much stuff do they produce so we are
35:30now in a in a in a tripartite world uh
35:34at some level almost you know co-equals
35:38uh except that uh by the numbers China
35:41is clearly the the biggest and u and
35:44then there are questions about political
35:46cohesion ability to act as a unit which
35:49is a arguably the really big difference
35:52between the United States and and Europe
35:54but um but that's the world we're in now
35:56and so any discussion about geoeconomics
36:00any discussion about hegemonic stability
36:02should bear in mind that look there is
36:04no hegeimon not by the numbers um where
36:08China is not yet big enough to be the
36:11hegeimon but the United States truly is
36:13not big enough to be the hgeimon so we
36:15are just not in that world anymore
36:18Okay, there are two puzzles
36:23uh involving these comparisons and as
36:26some people who've been you know who've
36:27been reading my substack know I've been
36:28spending a lot of time on them partly
36:30because I think it's important and uh
36:32but also because it's interesting um so
36:35um start with the one I've been spending
36:38a lot of time on which is the US versus
36:42Europe. So um
36:47we'll go back a slide right by by
36:49purching power parody go back
36:53to uh the year 2000 and the US and the
36:56EU27 were you know countries that are
36:59currently in the EU27 uh were about the
37:02same size as economies and look at them
37:05now and they're about the same size as
37:07economies and yet if we use national
37:10income accounts
37:12uh the United States has vastly outgrown
37:15Europe. So much bigger difference and um
37:20this is mostly uh it turns out
37:23demography is much less of a factor than
37:25I thought it would be. uh although US
37:28has higher fertility and until yesterday
37:31had higher immigration um the um uh it's
37:35not the fact that it includes anyway it
37:37turns out that it's mostly productivity
37:39as measured productivity has risen
37:41faster in Europe uh in the United States
37:44than in Europe um
37:47how can that be these are two and but
37:50you know these are both numbers compiled
37:52with great care by very careful
37:54competent people, national statistical
37:57authorities, the international
37:58comparisons program, how can they be
38:00telling such different stories? Well,
38:03one answer, one thing I have learned
38:05having been in this business for a long
38:07time is that when you see a real
38:08anomaly, something happening that does
38:12not seem to fit your models. Uh, one
38:16strong possibility is bad data and that
38:19has happened to me repeat. There were
38:22hopefully forgotten I think Philip might
38:24have even mentioned it but there was a
38:25there was a huge literature on uh on why
38:29uh the decline in the dollar following
38:32the plaza accord didn't translate into
38:34rising import prices and the answer
38:36turned out to be bad data. We were
38:39actually it turned out to be bad data in
38:41the same way that might be part of the
38:43story here. Uh how do you compute
38:46economic growth?
38:48uh unfortunately I mean we have we often
38:51talk as if economies produced a thing as
38:54our good we talk about one good
38:56economies um and
39:00you know that which is a very helpful
39:01thing but it is of course a metaphor and
39:04sometimes the metaphor it can fall apart
39:07in a couple of ways and one of them is
39:09that obviously many many of the goods
39:13that we use uh that we live by uh in the
39:18year 2026 didn't exist in the year 2000
39:22or have been so transformed as to be
39:25fundamentally incomparable. So how do
39:27you take account of that? The answer is
39:30hedonic measures. You try to measure in
39:33effect what is the value of the extra
39:35stuff that this new good delivers. uh
39:37which is right and it's done and again
39:39the statistical agencies put a lot of
39:42effort into this but there is no
39:44international standard for hydonic
39:47adjustments. We don't actually all do it
39:49the same way and although it's not
39:54hard to document but appears to be the
39:56case uh uh it does look as if the US
40:00statistical agencies are a little more
40:02aggressive in their hydonic adjustments
40:04than the European agencies are. So that
40:07the United States may be assigning
40:09greater value to smartphones in some
40:12sense than than Europe is. Okay. I don't
40:16think that's the main story, although I
40:17think it is part of it. I think the may
40:20I I I think although it's based only
40:22partly on evidence and partly just on
40:24what what makes a a nicer story. Uh I
40:27think a lot of it is simply that in a
40:30trading world one good economies is is a
40:33really problematic metaphor that we do
40:36in fact produce different mixes of
40:37goods. Uh and let me give you I've had
40:41an amazing amount of trouble uh probably
40:43my fault explaining this story to some
40:45people uh including smart people. So let
40:49me give you another version of this. Uh
40:53imagine a world in which there are
40:56different rates of technological
40:58progress in in different goods. Can
41:00imagine if you like a Ricardian one one
41:03factor world is all you really need
41:04here. But they're just for for reasons
41:06that are outside the model. There are
41:08just different rates of of technological
41:10progress and they're p persistent. So
41:12you have much faster productivity growth
41:14in um in information technology than you
41:18do in haircuts, whatever. Um the um and
41:22imagine um that
41:26uh you know two economies uh call them
41:28America and Europe. Imagine that Europe
41:32lags technologically.
41:35Doesn't have to be a large lag. could
41:36just be a year or two uh but just lags a
41:39little bit behind. What would the
41:42pattern of comparative advantage be?
41:44Well, it would be that the America would
41:48have a comparative advantage in the
41:50goods that have the most rapid
41:51technological progress because the
41:53technology gap there would be some
41:55technology gap in everything but it
41:56would be much larger for goods where
41:58productivity rises 30% a year. Um in
42:03that case what would the pattern of
42:04trade be? Well, America would specialize
42:07in the goods with the most rapid
42:09technological advance. Europe would be
42:11in the goods with less rapid
42:12technological advance. Um,
42:16what would happen to relative living
42:17standards in that world? Nothing. This
42:20is a steady state world. The relative
42:23prices of the rapidly advancing goods
42:25would follow would would fall. So that
42:28relative consumption levels uh would be
42:31the same. What would it look like in the
42:34national income accounts?
42:36The answer is well, you know, you could
42:39do a chain linked uh economic growth,
42:42productivity growth would be a weighted
42:44average of productivity growth in every
42:47sector in which you produce. But because
42:50the United States is doing the high
42:52productivity growth stuff, it will have
42:54a higher measured rate of productivity
42:56growth. And which is true, it's
43:00happening. Uh, and nobody's making a
43:04mistake here, but if you drew the
43:06implication that this meant that Europe
43:09was falling ever further behind, you
43:10would be wrong. This would actually be a
43:13uh uh a perfectly sustainable stat uh,
43:17you know, state of affairs. You might
43:18say, well, why doesn't Europe, you know,
43:20how can we close that technology gap?
43:22Which would be a good question, but it
43:24is not a sign of ever falling behind.
43:27It's not a sign that Europe is turning
43:29into a museum of its past glories. So, I
43:32like that story, you know, partly
43:34because I like Europe, but also because
43:35I think that it's a uh it happens to fit
43:39rather neatly into what we actually see
43:41in the data. Um there's a okay uh again
43:48it should not discount the possibility
43:50that we really just have a data problem
43:51but I think there is this real
43:53technology issue and and the
43:54technological lag is is pretty obviously
43:57real. It's not huge. It's and it's it's
44:00almost entirely on the producers side by
44:02the way. You know application looks the
44:04same on either side of the Atlantic. Um
44:07so that's a story. um try to get back to
44:11what that might mean for economic power
44:13in in a few minutes. But let me move to
44:15the other one. Um the other uh place
44:20where we have something funny in the
44:22numbers is
44:24uh China. I'll do China US. It's a
44:28little easier. So uh so on purchasing
44:31power parity, China is by far the bigger
44:34economy in dollars, you know, at current
44:38exchange rates, China is still
44:40substantially smaller than the United
44:42States. Uh okay, this one is actually
44:46although it might sound paradoxical,
44:47this one is actually familiar. The stuff
44:50I've been saying about technology and
44:52Europe versus US is is stuff that for
44:54some reason was neglected. This we
44:56understand extremely well. This is
44:58Belassa Samuelson. This is uh as
45:02countries develop, productivity grows
45:05more rapidly on average in traded
45:07sectors than in non-traded sectors.
45:10There's a systematic relationship uh
45:13between level of per capita GDP as
45:15measured either way but as measured by
45:18purchasing power parody and the apparent
45:20price level. Uh and so you know um uh
45:25electronics is similar in price in the
45:28United States and China but haircuts are
45:30a lot cheaper in China. And so that's
45:33and so we're actually those are both
45:34valid concepts. The dollar value of GDP
45:37uh is still substantially higher in the
45:39United States although the gap is
45:41narrowing. Um but the per PPP uh is much
45:45higher in China. Um, I would show you a
45:51I did a scatter plot of Blossom
45:52Samuelson using the most recent data and
45:55I won't show it to you because it's
45:56embarrassing because it works too well.
45:58It's almost it it's so good a fit that I
46:01almost wonder whether somebody is
46:03cooking the books to make it look right.
46:04But anyway, the uh but it's it it's
46:06that's very clear. So there's no real
46:08mystery. [sighs]
46:09Okay. Um
46:13let me give you another chart that I
46:16think is helpful that is may uh just
46:21kind of fit in with the rest. Um
46:24so actually before I say when we're
46:27talking about um power we are probably
46:32despite the slight defense I gave of
46:35Trumponomics there the uh it is
46:38primarily about um dependence upon
46:41imports which means that it's about
46:44production and it's about production of
46:47traded goods uh and we want to know uh
46:52you know who commands
46:54dominating positions in enough
46:57industries that the threat of disruption
47:00of trade is is really serious really
47:03something to be worried about. Um if we
47:06look at manufacturing value added and
47:09this is not purchasing power adjusted
47:10this is just current exchange rates but
47:13for some reason the big fluctuations
47:15probably because it's traded the big
47:16fluctuations associated with exchange
47:19rate movements are not nearly as visible
47:20here. Um,
47:23no question that uh, China is the
47:26workshop of the world, that China has by
47:28far the biggest manufacturing sector.
47:31Uh, it is not
47:34quite like the United States in 1950
47:37where it's sort of more manufacturing
47:39than the rest of the world put together.
47:41Uh, the US plus the EU combined are
47:43still larger. Uh I didn't actually put
47:46it in here, but it turns out that uh EU
47:50and US manufacturing value add are about
47:52the same. So actually the manufacturing
47:54numbers look like the purchasing power
47:56parody numbers. They're basically say
47:58that on that basis the US and the EU are
48:00roughly co-equal powers both each
48:03inferior to China but combined bigger.
48:07And so all right, that's kind of our
48:10power ranking in the world. um or it's
48:13our it's our economic weight. How does
48:16that translate into power? Okay, there
48:19are several
48:22big questions that we need to ask here.
48:25Um the first is uh just these raw
48:31numbers. They're not quite so raw, but
48:33the these these partially massage
48:36numbers um don't necessarily
48:40uh fully measure the extent to which a
48:44cut off of trade will damage uh one side
48:48uh of of the exchange because we they
48:51need to be adjusted for
48:54elasticity of demand, substitutability
48:57and again the the Yang and Lu paper went
49:00through that uh found some major
49:03disjunctions just to the naked eye
49:05obvious points um
49:08uh Taiwan is a quite small economy and
49:11yet if you really want nightmares uh
49:14about possible disruptions to world
49:16trade the idea that something would
49:18happen that would disrupt Taiwan's
49:22supply of a large part of the world's
49:24semiconductors is a really big deal
49:26there are also of course there are
49:28geographical
49:30uh positions
49:31So, um I don't think we would be talking
49:34nearly as much about uh about Iran if it
49:38were not for the fact that 20% of the
49:40world's oil was passing through the
49:42straight of Hormuz. Uh so there will be
49:45specifics
49:47uh associated with industries and we
49:50[sighs] uh I I I really applaud the you
49:55know the Louis Yang paper is making a
49:58real effort to get at this. I am a
50:00little bit nervous about how well we
50:02really estimate those demand
50:04elasticities which are crucial. Not a
50:06criticism, it's just saying this is
50:07really hard. I mean if you again if you
50:10spent a lot of time on international
50:12trade modeling over the years the the
50:15the long uh um mind-numbing history of
50:19of computable general equilibrium
50:21models. You know that uh elasticities
50:24and trade are one of the hardest things
50:25to pin down. But uh and do we fully
50:29appreciate that? And this by the way has
50:31a really strong
50:34uh role if we're trying to think about
50:37US versus EU comparisons. I believe that
50:40if we're asking about living standards,
50:42what we basically have is
50:46uh I think the picture is reasonably
50:47clear. Northwest Europe at least has
50:50something like you know 10% lower
50:53productivity than the US maybe uh and 20
50:58to 30% lower uh GDP per capita which is
51:03overwhelmingly because uh because
51:06Europeans take vacations uh and
51:08Americans do not uh and uh and that's
51:12not that's nothing to get really worried
51:15about and the difference in productivity
51:17growth growth rates. Again, I think
51:19that's largely just because of
51:21differences in industrial mix that don't
51:24necessarily matter for living standards.
51:27Uh, but I've done an exercise. I'm not
51:29putting it up here, but some of you may
51:30have seen it. Uh, I have been comparing
51:33two countries. Uh, California and not
51:36California. Uh, just divide the United
51:39States into California and the rest. Um,
51:43not California looks a lot like Europe
51:45terms of productivity growth. Basically,
51:48when we talk about rapid US
51:50technological progress, we are it's
51:52yeah, there's also Seattle and a few
51:54other places, but to a remarkable
51:56extent, we really are talking about
51:58actually just part of California. We're
52:00really talking about Silicon Valley. Um,
52:03you know, one of my favorite things
52:04actually when we um
52:06uh I'm a huge admirer of the Draggy
52:09Report. I'm not I'm not a little worried
52:12about some of the productivity stuff,
52:13but I think the recommendations are
52:14almost all things that should happen.
52:17And but uh and he does talk about, you
52:19know, capital and one thing Europe
52:22definitely does lack is a venture
52:24capital sector comparable to the United
52:26States. But where is the venture capital
52:29sector in the United States? It's not
52:31just Silicon Valley. It's Sand Hill
52:33Road. It's really just a few probably a
52:37square kilometer or so of of the United
52:39States, which is where all of that stuff
52:40is. So the the um Okay, why is that
52:44relevant?
52:46Texas Texas is actually a little bit
52:47above the US national average in
52:49productivity growth. Uh but it's, you
52:52know, it's a lot lower than California.
52:54Um,
52:57does Texas worry that California might
53:01weaponize its dominance of advanced
53:03technology by cutting off cutting Texas
53:05off from access to the latest model AI?
53:09Although that might be doing them a
53:10favor given where we are right now. But
53:11anyway, but does Texas worry that
53:13California that that California might
53:15cut them off from access to critical
53:17technology? The answer is of course not
53:19because, you know, we we're we're a
53:21country. we have uh specifically uh a
53:24clause in the constitution that
53:26prohibits doing anything like that. Uh
53:29should Europe worry that a future or
53:34even the current US administration might
53:36try to cut off European access to
53:38critical technology. Well, it's not
53:40something you can discount. And so there
53:42is a power element there. And I'm not
53:45sure at all that trying to calculate
53:47elasticities
53:49much as we we want to do that is going
53:50to get you the answer to all of that. So
53:54there are issues. Now one thing that we
53:57also don't know is we talk a lot about
54:03how much um the um vulnerability the
54:08European lag in the most advanced
54:10technology creates a vulnerability to uh
54:13to uh some kind of weaponized
54:16interdependence. Uh is that I mean it
54:20certainly does but is that unique? How
54:23many other things are there? Are there
54:25European products that the United States
54:28depends upon that um that uh we would be
54:33in real trouble if that access were cut
54:36off? And by the way, it's just insane,
54:38you know, given
54:41uh to be even talking about this stuff.
54:43But that's that's the world we're living
54:45in. Um I'll give you a silly examples
54:49are part of what just makes all this
54:50tolerable. So, I actually have several
54:53shelves in my pantry uh full of boxes of
54:59Italian pasta.
55:01Uh and the reason was there was a point
55:03last year when President Trump
55:06threatened to impose 100% tariffs on
55:09Italian pasta. um and which I would have
55:14been willing to pay, but I knew that
55:15what would happen was not that I would
55:17have to pay double, but that the um it
55:20would disappear from the shelves. And if
55:22there's one thing which is really not a
55:26substitute, it is American pasta. So um
55:30so you know th those are that that's the
55:32reality. And but are there more
55:35significant things? um you know
55:37everybody talks about ASML which may not
55:40be as the moat around that technology
55:42may not be as as deep as people have
55:45been assuming but are there other
55:47examples and the answer is I don't know
55:49but that's you know call for further
55:51research is what we always do here right
55:52but I think these things are really
55:54important um what about so US versus
55:59Europe what about uh China's ability to
56:02weaponize its position in the world
56:04economy what about weaponized uh
56:07interdependence coming from China. Um so
56:10I unfortunately did not have access to
56:13the Yang and and Lu paper when I was
56:17putting this together and there um so I
56:20actually I have a chart from people I
56:22don't know and don't necessarily trust
56:24but oh sorry I should have put this one
56:26in here. Sorry it's just a quick quick
56:29diversion here. um just looking at world
56:31trade
56:33uh this is this is not this is netting
56:38out intraEU exports so if we ask you
56:41know who are who is the superpower in
56:43world trade it's actually here it's
56:45actually the EU not the Chinese and we
56:48know how much you know the flood of
56:49Chinese goods we know about the surplus
56:51but the reality is that that uh the
56:53United States is a quite distant third
56:54and that the EU is actually bigger uh
56:57okay
56:59This is the Mercada Center on China
57:02Studies. I don't know them. I don't know
57:05how reliable they are. They and they
57:07don't
57:09uh the question I kind of wanted to ask
57:11which is about US versus EU uh
57:14interdependence. They don't ask they
57:16only look at dependence on China and
57:19they have a rather arbitrary measure of
57:22dependency on imports from China. uh
57:25which of course shows both uh uh the the
57:30uh the the two solid lines are are uh
57:33respectively uh US and EU dependence on
57:36Chinese uh exports. Uh and uh they
57:41netted out uh textiles. I guess they've
57:44decided that access to Chinese uh uh
57:46pajamas is not critical. Although one
57:50thing I happen to know just from the
57:51economic geography stuff is that most of
57:54the world's buttons are produced in
57:57Chao, China. And I'm not sure, you know,
58:00if if the world loses its supply of
58:02buttons, that might actually be a bigger
58:03deal than people realize. But anyway,
58:04the um uh um but the chart um um uh
58:11figure B in the Yangloo paper yesterday
58:15actually although they don't do it
58:16exactly they don't do a European
58:18aggregate actually does in fact appear
58:21to show the same result which is that
58:23yeah we're the western economies are
58:26quite dependent increasingly dependent
58:28upon China and if anything uh the United
58:32States is more dependent than than the
58:34EU and I don't know quite you know how
58:36seriously to to take this but I I I was
58:40reassured to see that people who whose
58:42methodology I really understood were
58:44coming that peers to roughly the same
58:45conclusion and would be interesting to
58:47delve into the details of all of that.
58:50Uh but I I will um uh let me let me um
58:55just uh not not pursue that further
58:58except to say that it's not as obvious
59:00as you might think. You know who is
59:01vulnerable here? Uh, and part of the
59:04answer may be that we're all vulnerable,
59:06although apparently the Chinese less
59:08than the rest of us, but the that the
59:09three superpowers are all quite
59:11vulnerable to uh to weaponized
59:13interdependence.
59:15Okay, let me back up to my starting
59:19point.
59:20Um
59:23here we are in a world where
59:27um the threat of
59:31economic damage because somebody cuts
59:33off your trade is clearly very real. Um
59:37one of the questions there there are two
59:39questions that you might want to ask
59:40about this. One of them is why why would
59:45you do such a thing? And the uh the
59:50every paper I've seen so far makes a
59:52really excellent choice which is to not
59:54try to answer that question. Uh which is
59:58you know it and again you know all right
1:00:00I mean I I'm obsessed with Canada and
1:00:03Canada US right now and if you try to
1:00:05ask me what is this about uh that's you
1:00:09know then that that's that's the end of
1:00:11your of your next month or so because we
1:00:13have no idea what that's about. But it's
1:00:15uh uh but it does happen. The reality
1:00:18which Albert Hersen uh could have told
1:00:21us uh 80 years ago is that it does
1:00:24happen that countries governments do
1:00:26have objectives. There's a there's an
1:00:29old economics international trade
1:00:31literature about uh non-economic
1:00:34objectives and how best to serve them
1:00:36which doesn't really quite get at this
1:00:38exactly but you know there at least
1:00:40people made the right choice which is to
1:00:42say look at some po at some point that's
1:00:44not a question we need to answer right
1:00:45away. We need to ask what what can you
1:00:47do assuming that your government has
1:00:49those objectives. So um so that is a but
1:00:53that is a a a big question and obviously
1:00:56if we want to say why did the world
1:00:58change so much uh two years ago it was
1:01:02not because there was a real change in
1:01:04objective conditions that that that
1:01:07meant that countries should have
1:01:08different objectives. It's just that
1:01:10countries did have different objectives
1:01:12and there they are. Um so that's the
1:01:15that's the uh that's the strategic
1:01:17choice but at some level eventually we
1:01:19do need to to you know try to understand
1:01:22what's driving this. Um and that's where
1:01:27uh if I can make a plea I I do think we
1:01:30need to bring the international
1:01:32relations people into these discussions.
1:01:35Uh yeah they don't do quantitative
1:01:37models. They have a an annoying
1:01:40addiction to 2x two matrices, but they
1:01:43are smart. They've thought about a lot
1:01:45and have actually picked up some aspects
1:01:47of international interdependence that I
1:01:49think the rest of us that the economists
1:01:51have have neglected. Um the other
1:01:55question which is really uh critical in
1:02:00on a lot of these uh issues now arising
1:02:04is um yeah we can measure hopefully uh
1:02:07at least in principle we can measure how
1:02:09much pain is created by disruption of
1:02:12trade. We can measure what what what
1:02:14happens to the economies when a country
1:02:18uh weaponizes interdependence.
1:02:20uh
1:02:22but how much do countries care what is
1:02:25the what is the relative willingness to
1:02:27bear pain
1:02:29uh that again it's it's um you know
1:02:31again uh uh on US Canada what is clear
1:02:35right now is that yeah almost for sure
1:02:39uh I I don't really know whether we can
1:02:41measure the dollar value of the costs of
1:02:44of trade war um it might be might be
1:02:49less asymmet metric than people think.
1:02:50It might actually almost be comparable
1:02:52in terms of dollar losses, but of course
1:02:54it's a share of of GDP. It will be
1:02:56enormously higher for Canada. On the
1:02:59other hand, the Canadians kind of know
1:03:00why they're doing what they're doing.
1:03:02And the great bulk certainly of the US
1:03:04public has no idea why we're doing what
1:03:06we're doing. and that kind of and that's
1:03:09obviously also if you're trying to
1:03:11figure out how if ever this uh Persian
1:03:13Gulf crisis uh ends uh there's no
1:03:17question that uh Iran is suffering an
1:03:20enormous amount of pain how much are
1:03:22they willing to bear that's the uh
1:03:25that's that's a and that in a way means
1:03:27that we kind of we actually need
1:03:29political scientists probably uh country
1:03:32expert political scientists to try and
1:03:34assess that. So world power is not
1:03:36simply a matter of the amount of damage
1:03:37you can do but of the amount of damage
1:03:39that you're willing to accept. And we
1:03:41don't have a real answer to that. Um
1:03:45last point um
1:03:50so we have we do know now that
1:03:52international interdependence is a
1:03:54potential tool of coercion. It can be
1:03:57used by countries to pursue objectives
1:03:59whatever they are. They have
1:04:01geopolitical objectives. they can use
1:04:03the threat of cutting off the actuality
1:04:05of cutting off trade to uh create pain
1:04:09for other countries in pursuit of those
1:04:11objectives. Um
1:04:15they also obviously first of all have an
1:04:17incentive to try to xanti alter the the
1:04:22calculus to uh to invest in sectors that
1:04:26make them less dependent uh so that
1:04:29they're they're less vulnerable to this
1:04:31kind of coercion uh but conversely to
1:04:35try to shape their trade in ways that
1:04:36make other countries more dependent and
1:04:38that's where Hman came in in 1945 by the
1:04:41way book actually was written in 42 and
1:04:43that just amazing to think of him being
1:04:45able to do that thinking in the middle
1:04:46of that scene but um at least his claim
1:04:50was that the Nazis deliberately created
1:04:52as much economic dependence in
1:04:54southeastern Europe as they could. Uh so
1:04:56that is
1:04:58one uh something that goes beyond this
1:05:02kind of picture. The other is uh well
1:05:05what if the coercion isn't enough
1:05:09uh or if it's too much? So the you know
1:05:12if weaponized interdependence isn't
1:05:14getting you where you want to go uh well
1:05:17the next step is like you know weapons
1:05:21uh and uh if I go back to that is
1:05:26actually the uh that's showing the the
1:05:28the war that happens when uh when the
1:05:32players are not prepared to accept the
1:05:34uh the results of weaponized
1:05:36interdependence. Uh and again in some
1:05:39sense uh one way to interpret
1:05:42I hope there isn't anybody who actually
1:05:44knows what's going on here but the uh
1:05:46what's what's happening right now is
1:05:48that uh you know there is some oil
1:05:51flowing through the strait uh the
1:05:54counter blockade is inflicting a lot of
1:05:56pain. So what do you do if you are the
1:05:59IRGC? Well, one answer is you encourage
1:06:02your allies to expand the war and you
1:06:04and you start uh uh doing drone attacks
1:06:08on the pipelines that bypass the the
1:06:10strait. And so ultimately this is all
1:06:14really interesting and frightening. But
1:06:16I think in a in a way I worry a lot
1:06:18about world trade distorted impeded
1:06:22duplication of effort loss of gains from
1:06:25trade because people are trying to avoid
1:06:27weaponized interdependence. But the
1:06:29biggest thing is of course that uh once
1:06:32you start turning uh trade relations
1:06:35into a source of geopolitical power, you
1:06:38might find yourself starting to do other
1:06:40things in pursuit of geopolitical power.
1:06:43On that happy note, thank you all.
1:06:46[applause]
1:06:48>> All right.
1:06:54So uh thank you Paul for that uh wide
1:06:57ranging and upbeat uh presentation. So
1:07:00I'm sure there's going to be uh plenty
1:07:03of uh questions and comments from the
1:07:05floor and in Paul's spoken for 45
1:07:08minutes. So uh rather than asking
1:07:11respond straight away to the first
1:07:12question I'm going to collect a few
1:07:14comments and questions. Uh Steve.
1:07:21>> Um thanks very much. Fantastic lecture.
1:07:23Um, one of the points you made is that
1:07:25exercising due economic power involves
1:07:27self harm because you're foregoing
1:07:29mutual gains from trade. I wondered
1:07:31about the sort of income distributional
1:07:32effects of trade underlying that. And so
1:07:34one reason why what might be driving
1:07:36some of these changes is a sort of
1:07:38perception that GDP per capita has grown
1:07:40very robustly since the 1970s, but the
1:07:42real wage of the median worker has not
1:07:44uh because of rising inequality. That
1:07:46may be mainly driven by technology
1:07:48rather than trade. But yet that could
1:07:50provide one reason why not just in the
1:07:52US where we see populist pressure but we
1:07:54also see it in Britain with Nigel
1:07:55Farage. We also see it in France with
1:07:57Marine Le Pen. Obviously there are many
1:07:59things happening there but I wonder to
1:08:01what extent this rise of geo economics
1:08:03is also related to the income
1:08:04distributional effects of trade and
1:08:06technology.
1:08:08>> Okay. So so let me see Arno in the front
1:08:10front row.
1:08:14Uh I was actually curious about the the
1:08:16discussion about uh different measures
1:08:19of economic size uh uh purchasing power
1:08:23parity and uh real GDP. So one way to
1:08:27think about it is you may be interested
1:08:28in output or you may be interested in
1:08:30consumption and and the difference
1:08:32between the two is is trade. And so I I
1:08:35was curious about uh whether for the the
1:08:39mechanism you were telling we should see
1:08:41something in the national statistics in
1:08:43terms of terms of trade for Europe that
1:08:45should be improving and and that's the
1:08:46other part of the that story.
1:08:48>> Okay. So let me pick a a s Kevin work
1:08:52just behind. Yeah.
1:08:57Um, I guess a question is should we be
1:09:00worried about dependencies? And, um, it
1:09:03strikes me that if we don't get the
1:09:04buttons anymore, we probably will figure
1:09:06out how to make the buttons, you know,
1:09:07and and maybe people will figure out how
1:09:09to make the Dutch machines if the Dutch
1:09:11machines are are cut off. So I'm
1:09:13wondering whether the the criterion for
1:09:15really worrying about it is is the stuff
1:09:16that's being cut off stuff that you
1:09:18would need in the event of military
1:09:19conflict because by the time you figured
1:09:21out a substitute it may be too late you
1:09:23know and that's that's maybe one
1:09:25distinction that is useful.
1:09:27>> Yeah. Okay. Um
1:09:30first obviously um the income
1:09:34distribution
1:09:35issues are
1:09:38I mean um they're huge in general in
1:09:42talking about international trade. Um
1:09:45and uh it's not at all clear to me
1:09:48however that they're playing a role in
1:09:50all of this. I can't really I mean it
1:09:53it's funny when when um when it looked
1:09:56as if the trade war was largely directed
1:09:59at China, you could make a case that
1:10:02this was about protecting wages. Uh that
1:10:07um and but now it's not. And of course
1:10:10that's partly because the Chinese
1:10:11effectively weaponized rare earths. Um
1:10:14but I don't see that there's any
1:10:17plausible income distribution story
1:10:20about current trade conflicts between
1:10:22you know the the currently hot trade
1:10:25conflicts are between the United States
1:10:26and countries at similar levels of GDP
1:10:29per capita really hard to tell a
1:10:30hexalene story about that so I don't I
1:10:34mean I appreciate and I think that's
1:10:35that's enormously important and I do
1:10:37think that international trade
1:10:39economists were sort of uh alerted to
1:10:42and uh and hip too uh distributional
1:10:47issues long before uh most of the rest
1:10:50of the economics profession but I don't
1:10:51think that's the story here. Um
1:10:57that was a very good question about the
1:10:58terms of trade and I have been
1:11:03uh trying to figure out if there's any
1:11:06way to measure it and it is really
1:11:08really hard. uh the we uh the
1:11:10conventional terms of trade measures
1:11:12just won't capture what we're trying to
1:11:14get here. And you you don't see it. You
1:11:16don't you don't see it. But the you
1:11:18mostly what we have for terms of trade
1:11:19are either uh the um uh uh the
1:11:26statistical agencies do some
1:11:28calculations but the same hedonic stuff
1:11:31applies. In fact, the first time I
1:11:32realized how how crippling, uh,
1:11:36uncertainty about hydonic adjustments
1:11:37could be was precisely on import prices
1:11:40in the in the 1980s. So, uh, so you
1:11:44know, they we we really are not it it
1:11:46should be there, but I'm not sure that
1:11:47we know how to measure it. So, yeah, I
1:11:49mean, definitely that's a clear that's a
1:11:51clear implied counterpart of the story,
1:11:53and I didn't present any numbers on it
1:11:55because I don't have any. um the um the
1:12:00military you know if if we worry uh one
1:12:04interesting question is what is the
1:12:06relevant measure of size for uh you I
1:12:11don't even want to think about this but
1:12:12you know if we if if we are you know if
1:12:16Hersman was writing his book in 1942 and
1:12:18questions of that for him to even be
1:12:20talking about trade as a source of power
1:12:22in 1942 required a a heroic effort to
1:12:25not think about uh uh not not think
1:12:29about um uh uh about the battle of
1:12:32Stalingrad, you know. Um the um um the
1:12:38and that's an interest that that's what
1:12:42if you'd asked me that question a year
1:12:44or two ago I or maybe three years ago I
1:12:48might have actually said that probably
1:12:50it is purchasing power power to GDP
1:12:52which is kind of like general productive
1:12:55capacity. Uh but given where war is now
1:13:00um you know uh in the war that's raging
1:13:03uh uh not too far away from here right
1:13:05now um large part of the drones being
1:13:09used by both sides have Chinese
1:13:11components. You might want to think
1:13:13about really about manufacturing
1:13:15capacity in general and and maybe
1:13:17specifically certain kinds of of of
1:13:20capacity. But no, it it it it is you
1:13:23know back at the beginning of World War
1:13:24II uh if you actually read people were
1:13:27talking a lot about steel production as
1:13:30you never mind and GDP barely existed as
1:13:33a concept I think but the but you could
1:13:36compare steel production and not there
1:13:37were some people on the axis side who
1:13:39said look at how big America's steel
1:13:41production is we can't win this thing
1:13:43now I don't know what the measure is but
1:13:45it's not going to be the obvious
1:13:47measures from the past
1:13:48>> thank you uh Kristoff off.
1:13:51>> Yeah.
1:13:51>> Well, maybe just wait for the mic as the
1:13:53uh online audience. Yeah.
1:13:56>> Um I wonder what you think Europe should
1:13:59do in this geoeconomic world. Um so you
1:14:02[clears throat] said using geconomic
1:14:05tools can increase the risk of military
1:14:07conflict. Others view it differently and
1:14:09say it's maybe a substitute to uh actual
1:14:12war. So I wonder in this world when you
1:14:15know the US and China use genoc
1:14:18should you know what's the right balance
1:14:20of of preventive efforts or actually
1:14:23should it actively use these tools and
1:14:25what's the balance between military
1:14:27investments or investing dollars and
1:14:29kind of sanctions and losses connected
1:14:30to that
1:14:31>> okay um again we'll collect three so so
1:14:34just on the same row at the Ilha
1:14:37>> okay thanks fascinating [clears throat]
1:14:39lecture I'm I was wondering about the
1:14:42different countries countries or
1:14:43countries with different political
1:14:44systems, how does that affect their
1:14:47willingness to use economic coercion and
1:14:49also their sort of willingness to endure
1:14:52pain? Okay,
1:14:53>> of course I'm thinking about China but
1:14:55also some other countries as well.
1:14:57>> Thanks.
1:14:58>> Okay, let me collect a third question
1:14:59before going back to Paul.
1:15:04So, uh, Luke,
1:15:14>> so thank you for this terrific and uh
1:15:17terrifying talk. Um,
1:15:21so if we would expand, I mean, you spoke
1:15:25mostly about trading of goods. We would
1:15:27expand that to financial trade, FDI.
1:15:34Do you mostly see parallels with what
1:15:35you said today or are there important
1:15:38differences?
1:15:40>> Okay.
1:15:42Okay. Um, let me actually respond to
1:15:46Luke first because, you know, I was
1:15:48alerted. I I I was uh I had my road to
1:15:54Damascus moment on all of this stuff
1:15:56from uh from international relations
1:15:58from uh from um Farrell and Newman on
1:16:02weaponized interdependence. Although
1:16:04it's funny there book is brilliant
1:16:06exposed all kinds of things but in 2019
1:16:09when they wrote about it they were
1:16:10mostly worried about the United States
1:16:12abusing its its power. And it's turned
1:16:15out that well yeah that can happen but
1:16:16there but there's a lot it it's by no
1:16:19means unique and they were very much
1:16:20focused on the financial role. It turns
1:16:22out that the the um uh in ways that I
1:16:26had not realized I should know these
1:16:28things but the centrality of the dollar
1:16:31uh also means that there's a centrality
1:16:33of the US banking system because almost
1:16:35everything ends up having to run through
1:16:37US banks at some point. Um, and that
1:16:40means that that it it is very much I
1:16:43mean in fact if you were trying to tell
1:16:45a story of how we got to where we are
1:16:47right now, a lot of it begins with the
1:16:52US attempting to weaponize the role of
1:16:54the dollar against Iran and uh then one
1:16:58thing leads to another. Um so that's
1:17:02very much part of it and again by the
1:17:04way uh these things are a wasting asset.
1:17:07the uh alternative payment systems. I
1:17:09don't think we're anywhere close to to
1:17:11seeing the role of the dollar as the as
1:17:14the the the world's key currency
1:17:16displaced, but the availability of
1:17:18variety of payment systems, crypto, but
1:17:21also Chinese systems and so on as as as
1:17:23workarounds uh develops over time. So it
1:17:27it there's a is fundamentally the same
1:17:29thing. I mean I uh I think the IR people
1:17:32are thinking too much about networks and
1:17:34not enough about just plain train
1:17:35disruption but they they certainly have
1:17:37a point there. That's so that's
1:17:39definitely part of it. Um the uh by the
1:17:43way the odd thing is that also creates
1:17:45some vulnerabilities if you want to you
1:17:47know imagine that that we go completely
1:17:50mad and have uh US European economic uh
1:17:54conflict. If you want to think about the
1:17:56amount of direct investment by US
1:17:58corporations in Europe,
1:18:00which is in a way, you know, we have
1:18:02given the EU hostages here. So, uh um
1:18:06okay. Um
1:18:08what should Europe be doing? Um uh this
1:18:14is I mean
1:18:17the the obvious
1:18:20thing I don't think that uh I would
1:18:23certainly not be advising the EU to try
1:18:25to create its enhance its ability to
1:18:28weaponize its position. I don't think
1:18:31that this is that kind of of continental
1:18:33society but reduce its vulnerability.
1:18:36Yeah. I mean I uh it's in an odd way it
1:18:40I hate to given where we are now to talk
1:18:44about the United States as a role model
1:18:46but you know we have or have at or were
1:18:48doing some things right. the uh the US
1:18:51uh uh was pursuing a fairly significant
1:18:55industrial policy until two years ago of
1:18:59uh the chips act and all of that trying
1:19:01to to create not full self-sufficiency
1:19:05but but reduce its strategic dependence
1:19:08on electronics uh semiconductors from
1:19:11from China and other potentially hostile
1:19:14powers and there really should be a
1:19:16European equivalent of the chips act.
1:19:18Now, of course, we have killed that act
1:19:19ourselves. So, it's one of those funny
1:19:21things. And the other thing, and this is
1:19:23this is what I think is really
1:19:26interesting and ironic. Um um trade
1:19:30policy
1:19:32uh I don't know how much people know,
1:19:34but probably most people in this room
1:19:35do. The United States has a system of um
1:19:39basically quick response trade policy
1:19:41tools that uh instead of alphabet soup,
1:19:44it's number soup. It's uh section 2011's
1:19:47and section 232s and section 301's and
1:19:49and so on. Uh if you ask what act are
1:19:52those sections of the answer is they're
1:19:54all different acts which makes it even
1:19:55worse. But anyway, they but we have
1:19:56these tools which enable quick responses
1:20:00to market disruption by floods of
1:20:03imports. That's a that's a 2011. Unfair
1:20:07foreign practices, that's a 301. Uh
1:20:10national security, very open-ended.
1:20:12That's a section 232. um which is a good
1:20:16system. It's designed to uh that you
1:20:19don't spend years debating about how to
1:20:22respond to a potentially weaponized use
1:20:25of interdependence. You can act quickly.
1:20:28The unfortunately the way the system was
1:20:30set up was that pretty much vested
1:20:34allegedly temporary but uh still more or
1:20:38less unilateral unilateral power in the
1:20:41executive branch. And all of which was
1:20:44based upon the assumption that the
1:20:46president would have a wider view of
1:20:48Congress and take a responsible position
1:20:51and be concerned with maintaining good
1:20:53relations with our allies. And and so it
1:20:56turns out that every thing has a flaw.
1:20:58And it was a system which was extremely
1:21:01effective uh in the past was almost
1:21:05customd designed to empower the current
1:21:08administration to do a lot of really
1:21:09destructive stuff. Uh and it's
1:21:12interesting again people
1:21:14um uh again anyone who's interested in
1:21:17this stuff I hope has or was about to
1:21:19read Chad Down and Sumea Kane's on how
1:21:22to win a trade war and it is fascinating
1:21:25to me that Chad and Sumea who are very
1:21:28much you know uh free trade pro
1:21:31globalization types are practically
1:21:34screaming that Europe needs it needs its
1:21:38equivalent of 2011 232 2s and 301's that
1:21:41you Europe does not have a process for
1:21:44responding to these threats and you know
1:21:47uh you don't need to put uh uh current
1:21:52US administration officials in charge
1:21:54but you need to have some kind of system
1:21:56that that the the sluggishness of the
1:21:58response to China shock 2.0 do and all
1:22:01of that is is really you need more. So I
1:22:05you know um I guess maybe this comes
1:22:08down to Europe should be more like a uh
1:22:11uh what the United States is supposed to
1:22:13be should there should be effective
1:22:16precautionary industrial policy. There
1:22:18should be effective uh precautionary
1:22:20trade policies. I don't know you know
1:22:22how that will work. Oh, and that then
1:22:24the political systems that that's
1:22:26actually that's a really kind of
1:22:27interesting question, right? Because you
1:22:29normally would think that well
1:22:30authoritarian regimes don't care and
1:22:33they can
1:22:34uh get away with anything. I don't think
1:22:38that that's what history tells us. I
1:22:40think what history actually says is that
1:22:43is that democratic systems are often
1:22:45given a given a good reason are if
1:22:47anything more robust more willing to to
1:22:50accept the sacrifices involved in in uh
1:22:53in holding their place in in geopolitics
1:22:56that if we ask uh uh there's a lot of
1:23:00back and forth but if we ask you know
1:23:02what happened during World War II the uh
1:23:04mobilization was more thorough earlier
1:23:09in the democratic nations than it was in
1:23:11the axis powers. If we look at uh uh uh
1:23:17the obvious extreme hesitance of of the
1:23:20Putin regime to go for full
1:23:22mobilization. Uh it's actually
1:23:24suggesting that democratic consensus in
1:23:27some ways that that authoritarian
1:23:29regimes have no disscent but are are
1:23:33more fragile or perceive themselves to
1:23:36be more fragile than you think. So no, I
1:23:38think on on balance uh it turns out that
1:23:42that
1:23:44you know democracies are frustrating and
1:23:45they debate and they sometimes are slow
1:23:47but they actually are are much more
1:23:48effective in conflict than people
1:23:50imagine.
1:23:51Thank you Paul. I suppose I I have a
1:23:54question which maybe connects the trade
1:23:56and and the macro. So um is it clear who
1:24:01are more vulnerable people running
1:24:03chronic current account surpluses or
1:24:05chronic current account deficits or does
1:24:08that margin uh make a difference? Ah
1:24:10that's a that's a good question and I
1:24:13would say actually if we think that it's
1:24:15mostly imports then it would be the the
1:24:18deficit countries that are more
1:24:20vulnerable
1:24:21um and um but it's going to depend upon
1:24:25the goods involved but yeah the uh uh um
1:24:30if I mean and not at all for the reasons
1:24:33that uh that you know that the
1:24:38mercantalists uh in in in my country
1:24:42think it's not that it's not that that
1:24:44uh that deficit countries uh uh actually
1:24:50I'm not sure I may have turned the
1:24:52argument around but look uh the the
1:24:54macro side the stimulus effects of ch of
1:24:59of eliminating current account
1:25:00imbalances are basically not relevant
1:25:03now really you know it it they in 2010
1:25:07with the world depressed and and all
1:25:09kinds of constraints preventing us from
1:25:11from doing enough to boost aggregate
1:25:13demand that mattered but it's it's not
1:25:16at all relevant now. Um and you know
1:25:18plenty of room uh to uh to cut interest
1:25:21rates or or uh if necessary. So those
1:25:24are not relevant. Um the and if we get
1:25:28into a conflict the so I think I have
1:25:30actually just turned this exactly
1:25:32around. The United States, I think that
1:25:35current leadership thinks that we're a
1:25:37winner in a trade war because we can
1:25:38eliminate our trade deficits and that's
1:25:40good. But the reality is because we are
1:25:42so dependent upon a lot of for for on
1:25:45imports in a number of sectors, we're
1:25:47actually the losers in a trade war. I
1:25:49think that was a completely nonlinear
1:25:52response, but I hope that that someone
1:25:54can disentangle it. Feed it give it to
1:25:56Claude and ask it to figure out what it
1:25:58was I just said.
1:25:59>> Okay. Uh, very good. So, so we're we're
1:26:02at the end of the session, but I'm just
1:26:04going to pause in case someone indeed
1:26:06has a devastating uh comment or
1:26:08question. Uh, so Luke,
1:26:10>> it's not my question. I was asked.
1:26:13>> Yeah. Um,
1:26:18so you showed us what we will soon and I
1:26:20guess in December see in Dune 3. This is
1:26:23where you started.
1:26:24>> No, it's it's actually a still from Dune
1:26:272.
1:26:27>> Yeah. Yeah. But I understand that it's
1:26:30coming.
1:26:31>> Um,
1:26:33how likely this is coming in the world
1:26:36today?
1:26:37>> What probability do you put on it?
1:26:40>> Wow. I mean,
1:26:44[sighs]
1:26:47>> not what you want,
1:26:48>> right? No, it's an interesting question.
1:26:52Look, there there's the catastrophe that
1:26:54has happened which has turned out to be
1:26:58uh in the micro sense worse than almost
1:27:00anyone imagined which is the you know US
1:27:03mil military collapse in the Middle East
1:27:05but in terms of the economic
1:27:07consequences has been actually much more
1:27:09muted than people feared. You know we
1:27:11we're actually having an oil shock that
1:27:13by the numbers is worse at least
1:27:15relative to world oil supply than 73 or
1:27:1779 but no global recession. So, uh, so
1:27:22that one I'm
1:27:24maybe excessively complacent, although
1:27:27check check in again in a couple of
1:27:29weeks. Um, the one that scares me is is
1:27:33uh is Taiwan.
1:27:35Um, and now you would hope, you know,
1:27:38clearly it would be completely
1:27:41stupid and counter to Chinese interests
1:27:44to
1:27:46to to start a war over Taiwan. you know
1:27:48they there there's no economic gain from
1:27:51that it would only hurt them definitely
1:27:53on the other hand um if it's one thing
1:27:57we've learned I think uh if we didn't
1:27:58know that before in the last few years
1:28:00is that governments sometimes do stupid
1:28:02things and we do know in particular that
1:28:05uh irudentism
1:28:07you know uh um Italy wanting Trieste
1:28:11back uh was a completely crazy reason to
1:28:15get involved in World War I and yet
1:28:17there it does uh so um the answer is
1:28:20that was that I am really there is a
1:28:22nightmare scenario there is a nightmare
1:28:24scenario where the uh and it isn't even
1:28:26exactly weaponized interdependence
1:28:28although it's it it ends up uh but that
1:28:30because uh these irrational all of the
1:28:34stuff that uh the Norman Angel told us
1:28:38was outmoded in 1909 because it's still
1:28:41out there uh will lead to major
1:28:43disruptions and we really have you know
1:28:45we given time We can reconstruct uh a
1:28:49semiconductor supply chain without
1:28:51Taiwan, but my god, the um the the the
1:28:56the few years it would take to do that
1:28:58would be very very ugly.
1:29:02>> Very good. So, uh let me thank Paul once
1:29:05again. Uh now going to have a break
1:29:07until 10:45. So, uh let us thank Paul.
1:29:11Thank you. [applause]
1:29:28Good morning.
1:29:43relatively new.
1:30:30Heat. Heat.
1:31:34Hey,
1:31:47baby.
1:33:07Hey,
1:33:13hey, hey.
1:34:12Hey.
1:35:16Hey.
1:35:26Hey. Hey.
1:35:46Hey,
1:35:52hey, hey.
1:39:11Hey.
1:39:30Hey,
1:39:50hey, hey.
1:42:09Hey.
1:44:03Heat. Heat.
1:46:39Heat
1:46:56up
1:47:07here.
1:52:12Heat. Heat.
1:54:19Hey,
1:54:33hey, hey.
1:55:05Hey, hey, hey.
1:55:45Okay. So everybody welcome back um to
1:55:48this last session of the conference and
1:55:50after this inspiring lecture by Paul
1:55:52Krugman. I have some bad and some good
1:55:54news for all of you. So I start with the
1:55:56bad news. There is uh one paper that
1:55:59will not be presented today because
1:56:01Fodora is unfortunately not well. So
1:56:03that's the paper on evading sanctions.
1:56:05So Kristoff Tish is also now released
1:56:08from his duty as discussant as a result.
1:56:10Um but the good news is we have another
1:56:12really great paper in this session um on
1:56:15uh tariffs and in particular what are
1:56:17the welfare effects from tariffs and in
1:56:20particular to what extent uh countries
1:56:22take into account what others are
1:56:24affected or not um by tariffs and how
1:56:27what the implications of that are and
1:56:29for that we have a great presenter cost
1:56:32from MIT um who's going to present the
1:56:35paper today it's a paper together with
1:56:37three co-authors um Rodrigo Dao John
1:56:41Stom Becko and Dave Donaldson. And we
1:56:44also have a discussant Meredith Crowley
1:56:47from Cambridge University. And so with
1:56:49much not much further ado, I would give
1:56:51the floor to um Arno to present his
1:56:54paper.
1:56:55>> Thank you.
1:57:01All right. Um thanks so much for the the
1:57:03invitation uh to uh to present this work
1:57:05in this uh in this conference. It's been
1:57:08great to um to attend. Um so this is a
1:57:12paper very much we've been talking about
1:57:13geoc geoeconomics power coercion. This
1:57:17is a paper that in a way is going to
1:57:19look like economic history. I'm going to
1:57:20be talking about the the rulebased order
1:57:23and international cooperation
1:57:26and and most of the data I'm going to be
1:57:27using is from 2001 right after the Euro
1:57:30great round which is not so far uh away
1:57:33in time but but but feels today uh very
1:57:36different. All right. So the the the
1:57:40starting point of of of what we're doing
1:57:42or or idea is that international
1:57:45cooperation at some fundamental level is
1:57:49about countries internalizing or not the
1:57:54impact of their own action say their
1:57:56tariffs uh on other countries. To the
1:57:58extent that they do internalize there is
1:58:01international cooperation. If they don't
1:58:03there isn't. And we know that in theory
1:58:06there are many reasons why countries may
1:58:09cooperate. There may be formal
1:58:11institutions, formal rules. So think
1:58:13about the the World Trade Organization
1:58:16and and before that the GAT as the
1:58:17poster child for that. There may also be
1:58:21informal rules where countries cooperate
1:58:25not because they're really externally
1:58:27forced to do it, not because they want
1:58:29to, but because they fear that if they
1:58:31don't cooperate, if they don't
1:58:32internalize the impact of their own
1:58:34actions on other countries, other
1:58:35countries will punish them, will
1:58:37retaliate. And it's that fear of
1:58:38retaliation that make them um cooperate.
1:58:42So you'll see that the paper today is
1:58:45very much an empirical paper. The object
1:58:48here is to try to make progress
1:58:52empirically on the extent to which
1:58:54countries perhaps with the rules of the
1:58:57world trading system and the WTO were
1:59:00able to to cooperate. And so what we're
1:59:03going to start with is some some metric
1:59:06uh for for international cooperation
1:59:08that that I'm going to come back to
1:59:10throughout this talk which is if I were
1:59:14to take a dollar from from your country
1:59:17from the European Union and I were to
1:59:19give it to some other country say Canada
1:59:23what would be the discount associated
1:59:25with that transfer is a dollar in the
1:59:27pocket of the EU just like a dollar in
1:59:30the pocket of Canada, do you not value
1:59:33it at all or do you have a 50% discount?
1:59:37So if you don't value it at all, I'm
1:59:39going to say that you don't internalize
1:59:40the impact of your own action on Canada.
1:59:42You don't care. There's no international
1:59:43cooperation. To the extent that this
1:59:45discount is very small, that mean that
1:59:48there's actually quite a bit of uh
1:59:50international cooperation going on. So
1:59:52think about the the end goal here of the
1:59:54paper empirically is going to be to look
1:59:57at the world. So think of a matrix.
2:00:00There's a bunch of trading partners
2:00:01around the world and they're going to
2:00:03put different weights on each other and
2:00:05I'd like to know what those discounts
2:00:06those weights are. The goal here is to
2:00:08try to identify this and and what we're
2:00:11going to be interested in are two main
2:00:13things. One is on average how big is the
2:00:17discount? So is there a lot of
2:00:19international cooperation in the world?
2:00:21Again circa 2001 right after the Euro
2:00:24gray world. That's one thing we'd like
2:00:26to know. And then the title of the paper
2:00:28is a world trading system for whom we're
2:00:31going to be also interested in the
2:00:32heterogeneity in those weights. So are
2:00:35the interest of some countries do they
2:00:38appear to carry more weights other
2:00:40trading partners to care more about them
2:00:44uh than others? And is it the case
2:00:46conversely that some countries appear to
2:00:49put more weight on on everybody else?
2:00:51Okay. So that's that that heterogenity
2:00:54uh that we're we're going to try to shed
2:00:56light on. And the way we're going to go
2:00:58about it is uh very much in the spirit
2:01:01of of what Steve presented yesterday.
2:01:03It's going to be a a reveal preference
2:01:05approach if you want where the idea is
2:01:08to go from action from choices that
2:01:11countries make to reveal their
2:01:14preferences how much they care about
2:01:16each other. And so I I've told you that
2:01:18this is very much an empirical endeavor
2:01:21here. But the the first thing we're
2:01:23going to do is is build some the theory,
2:01:26use it as a bridge to go from the
2:01:29actions that we observe. So in our case,
2:01:31it's just going to be tariffs. So we're
2:01:33going to observe the tariffs that
2:01:34countries impose on each other on on
2:01:36very disagregated products. And from
2:01:39that, so these are the the choices we're
2:01:41going to observe. We're gonna try to
2:01:44identify the the weights that uh that
2:01:47they put on on each other's welfare or
2:01:51real income. So the formula is going to
2:01:53look like this. Uh so let me walk you
2:01:56through through that uh briefly. So
2:01:59tariff GJ so think about G as a a
2:02:02product. So say it's banana J is an
2:02:04importer say the United States. So this
2:02:06is something I observe in the data.
2:02:08That's the tariff that the US uh is
2:02:11imposing on uh on bananas.
2:02:14I could compute uh what a selfish United
2:02:19States would do if it was only
2:02:21interested in US interest. Okay, that's
2:02:24the classical optimal tariff formula.
2:02:27You can identify that object. We're
2:02:29going to estimate it. That's what I'm
2:02:32calling here the opportunistic tariff.
2:02:34And the idea now is that well if on top
2:02:37of that I do care about my trading
2:02:40partner. So I is a trading partner of J.
2:02:42It's thei it's the European Union. It's
2:02:44it's Canada. Then what I may want to do
2:02:47as a country is to impose a tariff
2:02:50that's going to make sure that my firms
2:02:52my consumers whenever they import an
2:02:54export they're internalizing the p the
2:02:56impact of their own actions on my
2:02:58trading partners. So the what you have
2:03:02here in blue it's the sensitivity of the
2:03:04relearnings of the trading partners
2:03:06welfare if you want as country J is
2:03:10affecting its import of say bananas. So
2:03:13one one way to think about it is imagine
2:03:16that I knew the the welfare weights how
2:03:20much Jay cares about its different
2:03:22trading partner. I'm an economist. I'm
2:03:24hired as a consultant. They ask me what
2:03:26should be our tariff given the betas
2:03:28that we have. And I would say well look
2:03:32I've I know that when you import more
2:03:34bananas it's going to have a positive
2:03:37effect on the price of bananas. If you
2:03:39like that exporting country say Costa
2:03:41Rica. What I think you should do because
2:03:44your beta is positive and there's a
2:03:46negative sign here is have a tariff on
2:03:48banana that's lower than what a a
2:03:50self-interested countries would have
2:03:52done. Okay. So that would be sort of one
2:03:54way to to use the formula which is okay
2:03:57if I have this concern as a government I
2:04:01should act on it and I should adjust my
2:04:03tariff accordingly.
2:04:06There's nobody out there who's telling
2:04:07me I think these are orbitas. Instead,
2:04:10what as a researcher, I'm going to
2:04:12observe choices. I'm going to observe
2:04:14the tariffs and I'm going to use that
2:04:16relationship inverted to go from the
2:04:19tariff I observe to what the betas must
2:04:21have been. So if indeed I see that uh uh
2:04:26the United States tends to impose
2:04:28tariffs that are very low on the goods
2:04:30that Costa Rica exports, then I'm going
2:04:33to infer from that that the United
2:04:36States tends to put more weight on on
2:04:38Costa Rica than otherwise. So that's how
2:04:40the the identification
2:04:42um is going to work. So there's a bunch
2:04:45of colors in in in in my formula. So the
2:04:48black stuff is the stuff I really like.
2:04:51This is data. This is just choice. This
2:04:53is tariff.
2:04:56The blue stuff is things that I wish was
2:04:58data. I wish somebody had estimated for
2:05:02me the causal effect of various import
2:05:04restrictions on prices and therefore
2:05:06real income around the world. Even if I
2:05:08knew that, I would just use that as part
2:05:11of my regression. And my opportunistic
2:05:13tariff is the exact same thing. It's
2:05:15just a function of how prices respond to
2:05:18various import restrictions. So
2:05:20potentially one could estimate that it's
2:05:23it's a tall order empirically because
2:05:25well that's a lot of goods and a lot of
2:05:27countries around the world and we don't
2:05:29have that full knowledge. So what we're
2:05:31going to do instead to make progress is
2:05:34we're going to use a model. We're going
2:05:35to estimate that model. We're going to
2:05:37validate that model. Meaning that we're
2:05:39going to look at the predictions of that
2:05:41model for changes in real income when
2:05:44the the variation in the type that we
2:05:46have in the data is occurring and check
2:05:48whether for that particular subset of
2:05:51variation the predictions of the model
2:05:53line up with what we've observed in the
2:05:55data. We're not going to be able to
2:05:57reject our model. That's going to be our
2:05:58sanity check. And so we're going to
2:06:00proceed now taking the blue stuff which
2:06:03is a very structural object coming from
2:06:05the world now treating it as as good as
2:06:06data. So black stuff was data blue stuff
2:06:10is stuff I have. And now if you look at
2:06:13this expression it looks like
2:06:17a regression a linear regression that
2:06:20one could run and that's how we're going
2:06:21to implement it. So I observe choices I
2:06:23observe your tariffs that may or may not
2:06:25be what the selfish countries would have
2:06:27wanted to do. I know the sensitivity of
2:06:31real income around the world to various
2:06:32income restriction. That's how I'm going
2:06:34to identify the weight that different
2:06:37countries put on each other. Okay,
2:06:39that's the that's the end goal. Here's
2:06:42what we find. So, we're going to run
2:06:43that regression. We're going to collect
2:06:45our data. We're going to estimate them
2:06:47all. What we're going to find with those
2:06:50with those uh weights is that there's
2:06:52quite a bit of international cooperation
2:06:55going on circa 2001. So it's not like
2:06:58countries are selfish. They're they're
2:07:00putting a a weight the discount that I
2:07:02was describing at the beginning. It's
2:07:04about u it's about 20%. So a dollar that
2:07:07you that the EU would give to Canada
2:07:09that's still 80 cents. That's what comes
2:07:12out of the of the exercise.
2:07:16So you internalize.
2:07:18One thing you could ask is do you go all
2:07:21the way to efficiency? So do we all
2:07:23agree about what the weight should be?
2:07:24Do we have global efficiency? The answer
2:07:26is no. So there's significant
2:07:28international cooperation, but that that
2:07:30that stops short of uh of of full
2:07:33efficiency. One thing that that we find
2:07:36interesting about uh about our exercise
2:07:39is that something that that that's going
2:07:41to pop out of those traces of
2:07:43international cooperation, if you want,
2:07:45those those revealed measure of welfare
2:07:48is some some reciprocal behavior. And
2:07:51one way to think about it is that there
2:07:52will be in the data a positive
2:07:54correlation between the weight that you
2:07:56put on your trading partners and the
2:07:58weight that trading partners put on you.
2:08:00So some countries are going to be very
2:08:01cooperative if you want. They care a lot
2:08:04about the world and the world care a lot
2:08:05about them. And for some countries it's
2:08:07going to be the opposite. That's
2:08:08something that that we're going to see
2:08:10as well. And one thing that that maybe
2:08:12is more um intriguing and and and we may
2:08:16not have the the full answer on that,
2:08:17but but but I I want to point out is
2:08:20there's nothing in what we're going to
2:08:22do that is going to be screaming that
2:08:24there's something very different about
2:08:26WTO versus nonWTO. So how much time you
2:08:29spent in the WTO as a country, you may
2:08:31think that that make you much more
2:08:33cooperative or not, we we're not going
2:08:36to see that. So there seems to be
2:08:37something in 2001 about that reciprocal
2:08:40behavior that expands above and beyond
2:08:44uh the the World Trade Organization.
2:08:47All right. So let me walk you briefly
2:08:50through the theory and then uh uh give
2:08:53you the results. So the way we approach
2:08:55the theory, it's a general uh
2:08:58competitive environment where countries
2:09:00can impose tariff and that's the way
2:09:03they're going to incentivize consumers
2:09:05and firms to do what they want. Okay? So
2:09:06you don't get to decide exactly what
2:09:08consumers and firms do. But you said
2:09:10tariffs, they affect the price, they
2:09:12affect therefore how much people consume
2:09:14and how much firms produce. That's uh
2:09:18the setup we're going to consider. Now
2:09:20what the government can do is choose
2:09:22that vector of tariffs and so implement
2:09:25a particular equilibrium uh with
2:09:27tariffs. The way people would typically
2:09:29proceed is have a a full description of
2:09:33the environment the constraints that the
2:09:35different governments are facing. We're
2:09:38going to propose something more reduced
2:09:40form without without apology which is
2:09:42now imagine that at the end of the day
2:09:46once you you put in place those
2:09:48institutions once you have this fear of
2:09:50retaliation the only thing it does is
2:09:53the reduced form of it is inducing you
2:09:56to put some weights on the welfare of
2:09:58your trading partners and that's that
2:10:00that's going to be our starting point of
2:10:02characterizing
2:10:05tariffs if you had this as if altruistic
2:10:08motives and and as if here is emphasized
2:10:10because I'm not pretending that
2:10:12countries are alruistic. I'm pretending
2:10:15that they may behave as if they were
2:10:18because of various institutional
2:10:21consideration. So what we're going to do
2:10:22is make that assumption and then
2:10:25characterize the tariff associated with
2:10:27that uh with that assumption. So just to
2:10:30illustrate some some polar cases, the
2:10:34first one here on your left is if I'm
2:10:36selfish, I do what's best for me. You
2:10:38have country one on the x- axis, country
2:10:40two on the y- axis. You do what's best
2:10:42for you. So if you're country one,
2:10:44that's why you see a horizontal
2:10:46tangency, vertical tangency for country
2:10:48two. Uh that's an inefficiency. There's
2:10:51a little lens here. That's the the
2:10:54problem with the tariff imposed by
2:10:57selfish countries. You could have global
2:10:59efficiency. That's also a special case
2:11:00of what we're doing. Then you would have
2:11:02tangency. That's the situation where all
2:11:05countries agrees on the weight that
2:11:07everybody should get. That's what a
2:11:08world planner would want to do. What we
2:11:11find more interesting is the case in the
2:11:13middle that may look uh something like
2:11:15that where you put some weight uh on
2:11:19other country that doesn't have to be
2:11:21the same as everybody else but doesn't
2:11:22have to be zero either. So why is it
2:11:24that you may want to put some weight? So
2:11:27let me give you a feel for the how you
2:11:29you can microphone that reduced form
2:11:31that uh that we're using here. Suppose
2:11:33that you're playing this repeated game
2:11:34and it may be a very complicated
2:11:36repeated game but it's going to have one
2:11:38particular feature which is other
2:11:40countries they're only going to
2:11:41retaliate. They're only going to punish
2:11:43you if you actually affect them. If you
2:11:46don't affect them they stay put. Okay,
2:11:48which we think is a is a reasonable
2:11:50assumption to make. Then I'm going to
2:11:52claim that well if that's the case then
2:11:55whatever the details of the game may be
2:11:57it must be the case that what each
2:12:00country is doing along the equilibrium
2:12:02path is maximizing its own utility
2:12:05subject to giving the other countries
2:12:08the utility they're supposed to get as
2:12:10well along the path. Why must they do
2:12:12that? Well, remember nobody is
2:12:14retaliating if you're giving them their
2:12:16their full utility. So you'd better get
2:12:18what's best for you. That's that's the
2:12:20idea. Well, in that world, that means
2:12:23that those weights that I'm giving you,
2:12:25they're structural interpretation. These
2:12:27are the lag range multiplier associated
2:12:30with that constraint. These are very end
2:12:32and endogenous objects. These are
2:12:34complicated objects. If I had to
2:12:36structurally tell you how they're going
2:12:38to be affected by particular
2:12:39institutional reform, that's very
2:12:41complicated. We're not going to do that.
2:12:42I'm going to tell you I'm just going to
2:12:44be able to measure at a given point in
2:12:46time what those things are. Are they
2:12:48large? Are they small? that's what we're
2:12:50interested in. Okay, that's the formula
2:12:52that comes out once you make uh that
2:12:55assumption. Again, the the logic here is
2:12:59is that if you give tariff advantages,
2:13:03if you impose lower tariff on the goods
2:13:05that benefit certain countries, it must
2:13:07be that you care about them and and and
2:13:10we're going to tell you how much. Okay.
2:13:13So, how uh are we going to estimate
2:13:16that? So we're going to treat the that's
2:13:18the regression that we're going to use.
2:13:20So one way to think about that condition
2:13:22is that it's really really a first order
2:13:24condition. So a necessary condition for
2:13:26the tariff I observe to be optimal. In
2:13:30practice the maybe other reason why you
2:13:32impose tar distributional consideration
2:13:35within your own country. Maybe you're
2:13:37trying to fix something and maybe you
2:13:39know governments make mistake. Tariffs
2:13:41are measured with error. that would all
2:13:44feed into uh the the error term that we
2:13:47have on the right. And so our
2:13:48identifying assumption here is going to
2:13:49be that all those other motives for the
2:13:52types they're orthogonal to that
2:13:54international uh redistribution motives
2:13:57and that's how we can uh identify the
2:13:59the betas. We have a model again
2:14:02standard sort of a quantitative model
2:14:05that we calibrate and then we're just
2:14:08going to test uh whether we can
2:14:10replicate or not. the change didn't
2:14:12really come with we with observed. Okay,
2:14:15I so for the last uh minutes I have I
2:14:19want to show you some uh some results.
2:14:21So I I I promised you international as a
2:14:24bit of a nerdy way to uh uh to present
2:14:27things but but here we are. Uh so we
2:14:30have 28 countries. The EU is one. Uh
2:14:34then we have the 27
2:14:36uh biggest uh uh uh trading countries in
2:14:40the world non-EU and and the rest of the
2:14:42world. So that's 28 country 27 trading
2:14:45partners. So that's 28 * 27 that's 756
2:14:50coefficient that we've just estimated
2:14:52and that's a little crazy but I'm giving
2:14:53you the 756 coefficient. That's the the
2:14:56matrix. That's the heat maps. These are
2:14:58the betas that we've estimated.
2:15:00Everything is relative to self. Okay.
2:15:03Okay. So this is a discount compared to
2:15:04how much you value a dollar for you. So
2:15:07forget about the diagonal. It's it's one
2:15:09as a matter of normalization. That's how
2:15:11we are uh we're computing things. Okay.
2:15:16So that's how we're thinking about
2:15:17international cooperation. So we went
2:15:19from the general idea to something we've
2:15:21operationalized. And then what I like
2:15:23about the this way of of looking at the
2:15:26world now is that the properties of that
2:15:28matrix are going to speak to things that
2:15:31we as economists may find interesting.
2:15:33So our country self-interested yes or no
2:15:37in terms of that matrix. This is all
2:15:39about the off diagonal elements. Are
2:15:40they zero? The selfish case is the case
2:15:43where nobody cares about anybody else in
2:15:45the world. That would look like you put
2:15:47very little weight of the diagonal.
2:15:50Circa 2001. What we find is that on
2:15:53average uh the the welfare weight is
2:15:56about uh 80 81. So the discount for a
2:16:02dollar that you give to a trading
2:16:03partner on average is only uh is only 19
2:16:07uh uh percent. Okay. So this is we can
2:16:10reject this opportunistic tariff
2:16:12setting. That's something that that
2:16:13comes out uh very directly out of our of
2:16:17our estimate. The other thing you could
2:16:19ask is how close are we to something
2:16:23that would be globally efficient. So one
2:16:25way to think about global efficiency is
2:16:27what you could observe if some global
2:16:30planner with some particular weights
2:16:32about you know who should receive high
2:16:35income or not around the world had
2:16:37designed the uh the world tariffs. Well
2:16:41with that planner if if you are at a a
2:16:44global par optimum everybody should have
2:16:47the same weight. So in terms of that
2:16:49matrix it means that the rank should be
2:16:51one. Uh it means that all those columns
2:16:54should look the same. We should all
2:16:56agree on who should receive uh the same
2:16:59weight. It's not the case in practice.
2:17:02We can reject that partly because well
2:17:04look if we all agree then we shouldn't
2:17:07all impose a discounts on the rest of
2:17:09the world. So the fact that it's not
2:17:10full altruism is already telling you
2:17:12that we don't have uh global efficiency.
2:17:16So do we have the same weights? We
2:17:19don't. The the the the the remains
2:17:22uh some substantial amount of
2:17:24heterogenity and and so we can we we can
2:17:26reject global efficiency as well. The
2:17:29part that that we find the the the more
2:17:32interesting about that investigation is
2:17:34the one that has to do with the restal
2:17:37behavior. Okay. So the idea that uh you
2:17:40exchange a good for a good, a bad for a
2:17:43bad tit for a tat, you scratch my back,
2:17:45I scratch your so so that's that's an
2:17:46important idea in social sciences, in
2:17:49international relations and that's one
2:17:51that we can investigate here by looking
2:17:54at whether when you tend to have a high
2:17:56beta then you receive a high beta. Okay,
2:17:59so being good to someone is having a
2:18:01high beta and do others reciprocate. One
2:18:04way to look at it visually is to uh
2:18:07through this scatter plot. So I look at
2:18:09the average beta that you have as an
2:18:11importer. So this is how much weight you
2:18:13put on each other and this is uh how
2:18:18much weight you tend to receive and you
2:18:20see that that positive correlation.
2:18:22Okay, so it looks like those reciprocal
2:18:24behavior. Something I should have said
2:18:27is when you're thinking about that
2:18:28matrix, what what does this reciprocity
2:18:31speak to? Well, it speaks to whether or
2:18:34not it's symmetric. Okay, so symmetry in
2:18:37the matrix is the trace of of reciprocal
2:18:40behavior. And the one data point you
2:18:44probably can't read given the size of
2:18:46the font, but the one uh there's a
2:18:49little cross here floating around in in
2:18:51the middle that that is this this
2:18:53striking I have a low beta and everybody
2:18:56else has a low beta on me. So that's
2:18:58very much reciprocity.
2:19:00That's actually India. Okay. So that
2:19:02that's that's the country uh that that
2:19:04that stands out uh in the in the matrix
2:19:09and that's also our outlier here in the
2:19:11regression in terms of a very low
2:19:13average beta uh that it imposes and that
2:19:17it receives. You could say, okay, that's
2:19:19that's that perhaps that's not an
2:19:21interesting, but what if you had just
2:19:23looked at average tariffs? You know, my
2:19:25my metric for whether or not I care
2:19:28about the world and the world cares
2:19:29about me, why not look at the average
2:19:32tariff that I impose, the free trader
2:19:35probably cares more than uh than the
2:19:37others. One thing we find interesting is
2:19:39that if you do the exact same exercise
2:19:42on the average tariff,
2:19:44then you actually don't see it. So you
2:19:46see it in the beta. So through the lens
2:19:48of our exercise you see traces of
2:19:52reciprocal behavior that would be
2:19:54invisible if you just look at the
2:19:55average tariff. So you know what is
2:19:57going on because the beta is obviously
2:20:00identified out of the tariff choices. So
2:20:03the idea is that yeah but we're
2:20:04identifying things not purely out of the
2:20:06level of the average tariffs but also
2:20:09their composition. So do I see that I
2:20:12tend to impose tariffs? It's not just
2:20:15that they're low on average is that they
2:20:17tend to be particularly low for the
2:20:19goods that my trading partners value and
2:20:22and and that's what we we see and that's
2:20:24how we can uh get our our result. I
2:20:27promised you we we we looked at at the
2:20:29the role of the GAT and whether it's
2:20:32important in in sustaining
2:20:34um uh that reciprocal behavior. So one
2:20:38thing we did is run the regression. So
2:20:40this is country by country. I'm going to
2:20:43run at the weight you impose on your
2:20:45trading partner on the weights that your
2:20:47trading partners impose on you. So J is
2:20:48fixed. I'm just varying I the the
2:20:50trading partner and so I'm doing that
2:20:53country by country and I'm reporting the
2:20:55coefficient. So in this case we see that
2:20:58there is this this this positive
2:21:00association correlation between uh beta
2:21:03and beta ji but there doesn't seem to be
2:21:06any systematic relationship with tenure
2:21:08in the in the WTO which which we find
2:21:10interesting. So that's sort of pervasive
2:21:12and above and beyond uh uh the WTO.
2:21:16Okay, this was all for 2001. Uh you can
2:21:20repeat the same exercise. We have data
2:21:22from 97 to 2019. And so here I'm giving
2:21:25you a feel for okay, this is what
2:21:27happened to international cooperation
2:21:29through the lens of uh of our analysis.
2:21:32So you see that in the 2000 the average
2:21:36level of cooperation according to that
2:21:37metric that's still going up and then
2:21:39that starts plateauing in the in the
2:21:422010 you can look at the dispersion in
2:21:45the betas which again you know there
2:21:47shouldn't be too much if we if we're
2:21:49getting closer uh to uh to global
2:21:51efficiency and you also see a decrease
2:21:54so more international cooperation
2:21:56according to that metric and then uh and
2:21:58then plateau so the world has changed we
2:22:01we we don't go all the to current
2:22:03events. But you know, let let me finish
2:22:05with that picture. That's the first US
2:22:07China trade war. Okay. So if if you
2:22:09implement or our approach, if you look
2:22:13at the the tariffs that that were
2:22:15imposed by the first Trump
2:22:16administration on China and the
2:22:18retaliatory tariff imposed by China,
2:22:21yes, that that's a collapse, you
2:22:23revealed collapse of how much you care
2:22:25about that trading partners and how much
2:22:28that trading partner uh cares about you.
2:22:31Okay, so I'm out of time. So that was
2:22:33the idea uh behind this paper. So try to
2:22:35use tariff as a way to to measure uh
2:22:40uh international cooperation. So we find
2:22:43a significant amount of international
2:22:45cooperation. Reciprocity seems to be a a
2:22:48key feature and it's not easy to explain
2:22:50only with the with the WTO rules. Right.
2:22:54So thank you very much. [applause]
2:23:00So pause the floor to me for the
2:23:01discussion.
2:23:05>> Okay. Thanks uh very much to Arno for
2:23:08writing such an interesting paper and
2:23:10thanks for having me here to discuss it.
2:23:12Um
2:23:14so I'll skip the the overview and I'll
2:23:17I'll jump into it. So the core idea that
2:23:20they're after in the paper is if we look
2:23:23at a snapshot in time and they'll start
2:23:25with 2001
2:23:27and the first question is can we measure
2:23:30international commercial cooperation? Is
2:23:32there a way to measure this object?
2:23:36Then additional questions are if we look
2:23:39back at 2001
2:23:41which countries were benefiting from
2:23:44other countries sort of pro-social
2:23:46cooperative behavior. So who are the
2:23:48beneficiaries back in 2001 of the global
2:23:52structure of tariffs as it existed at
2:23:53the time? And then are some countries
2:23:58systematically more generous and more
2:24:01outwardlooking in how they set their
2:24:03trade policy? are some countries
2:24:05inherently more self-interested
2:24:08in some ways less cooperative. And so
2:24:11the way they define it is they put
2:24:13together this concept and they say, can
2:24:14we measure how much what they call is as
2:24:17if altruism that we see in the tariff
2:24:20structure back in 2001. And so the idea
2:24:24conceptually is if I'm the United
2:24:27States, I cut my tariff on Canada, how
2:24:30much do I value the income transfer that
2:24:33Canadians will receive because what they
2:24:36export to me becomes more valuable when
2:24:38I have this tariff cut? How much do I
2:24:41value that gain in income that I'm
2:24:42giving to Canada? Do I think Canadians
2:24:45are worth as much as income generated
2:24:47within the United States?
2:24:50And so very simply, this is a very
2:24:53elegant thing they're doing. They're
2:24:54just going to estimate a measure of
2:24:56bilateral cooperation that's defined as
2:25:00the actual tariff I see in the world in
2:25:022001
2:25:04less the optimal self-interested
2:25:08uh revenue surplus extraction tariff
2:25:10tariff that I could be charging. and
2:25:12they're going to look at that difference
2:25:14and see how it relates to the
2:25:17responsiveness of real income growth in
2:25:19the exporting country to the tariff the
2:25:22importers imposing. And so the core idea
2:25:26is that if I'm the United States and I
2:25:28cut my import tariff on Canada, that
2:25:31enhances Canada's real income. That's a
2:25:34measure of cooperation and American
2:25:36altruism.
2:25:38Okay. So to get started, I'll just put
2:25:40up some summary statistics. This is very
2:25:43related to the last picture um or one of
2:25:45the last pictures Arno showed us where
2:25:47he looked at just raw tariffs. Okay? And
2:25:50so this is just snapshot taken from a
2:25:53paper Chad Bound and I wrote where we're
2:25:55looking at the multilateral tariff
2:25:57structures. This is only WTO tariffs
2:25:59back in 2013 after the Urgue round is
2:26:02implemented. And I just picked three
2:26:04high, middle, and low-income countries.
2:26:06And the first thing that emerges when
2:26:08you look at this is there's two tariffs.
2:26:11One is the MFN tariff applied on WTO
2:26:14members at the border in 2013. This is
2:26:17what most countries are doing. And then
2:26:19the second one is the one they bound the
2:26:21rate at under the WTO. So they promise
2:26:24all WTO members they won't charge them
2:26:26more than this. If you look at this
2:26:28bound commitment, you see that as
2:26:30countries have lower income, they've
2:26:33made fewer commitments to keep their
2:26:35markets open. So first thing is
2:26:37high-income countries have in a tariff
2:26:39sense more open markets. They give more
2:26:42market access to their trading partners.
2:26:45And then the second thing highlighted in
2:26:47yellow has been raised a lot in US
2:26:49policy debate recently. The multilateral
2:26:52system US tariffs are lower than
2:26:53everybody's bound and applied. So we had
2:26:57this in a paper in 2013.
2:27:00Wilbur Ross published this in the Wall
2:27:02Street Journal in 2017. and his version
2:27:04was
2:27:06the existing tariff structure in the
2:27:08world is screwing over the United
2:27:10States. The US is giving away market
2:27:12access. Its trading partners don't treat
2:27:15it very well. And so rather than the US
2:27:18being this hegemonic power that has
2:27:22control over markets globally and that
2:27:24set global rules for trade, the US is
2:27:26this weak country that is basically
2:27:29being exploited by all the other
2:27:31countries in the world. And so the nice
2:27:33thing in the paper today is they can do
2:27:36very precisely a measurement of well
2:27:39actually who is benefiting and who is
2:27:42hurting and is there any evidence to
2:27:44support this claim uh by the Trump
2:27:47administration.
2:27:49So the papers
2:27:52it's embedded in this long tradition in
2:27:56general equilibrium trade theory going
2:27:58back to Harry Johnson but even before
2:28:00Johnson. Johnson's just one of the
2:28:02modern papers but the basic idea is
2:28:04trade policy can be used by nation to
2:28:06increase one's own real national income.
2:28:10Johnson has a paper back uh in the 50s
2:28:13saying well if countries are serving
2:28:15their own self-interest in a national
2:28:17national income sense they should set
2:28:19the optimal tariff roughly one over the
2:28:22export supply elasticity they're facing
2:28:24and even in the face of retaliation by
2:28:26trading partners they will do so. So
2:28:28countries will set tariffs optimally in
2:28:30an opportunistic way. Bagwell and
2:28:33stagger then embedded this in a series
2:28:35of models in the 1990s in which they
2:28:38said if we take the Johnson worldview
2:28:43why do countries not set optimal
2:28:46tariffs? And they said well we can think
2:28:47about this in a game theoretic way.
2:28:50Countries opportunistically do want to
2:28:52set tariffs that extract real national
2:28:55income from their trading partners. But
2:28:58if they both are doing it at the same
2:29:00time, they can together agree, I'll cut
2:29:04my tariff on machine parts, you cut your
2:29:06tariff on uh apparel, I'll export
2:29:09apparel to you, you will transfer real
2:29:11income to me, I will transfer real
2:29:13income to you. The world will have uh
2:29:16more efficient and higher national
2:29:18income. Now that framework in Bagwell
2:29:21Stagger assumes perfect cooperation and
2:29:25so the idea in their model would be as
2:29:28the US as the European Union reduced
2:29:30tariffs against one another over 70
2:29:32years, they were always perfectly
2:29:35equally valuing the real income transfer
2:29:38they gave to their trading partner as
2:29:40much as they were valuing the real
2:29:42income transfer they received.
2:29:45And so what Arno and his co-authors are
2:29:49doing is they're saying, "Okay, well,
2:29:51let's take that concept and embed it in
2:29:54a broader model." And so now the utility
2:29:56of a country is not going to be its own
2:29:58national income, but we're going to also
2:30:01consider the benefits to other
2:30:03countries. And these lambdas are going
2:30:06to be something that are uh we can
2:30:09measure. And so if I'm in the bagwell
2:30:12stagger world, the lambda coefficient in
2:30:15our nose model will be exactly equal to
2:30:17one. If I'm in a different world, we can
2:30:20measure this as something different.
2:30:22Harry Johnson's world would say lambda
2:30:25is equal to zero. So their model
2:30:27encapsulates this whole large uh
2:30:29theoretical literature. And then as Arno
2:30:32showed the second equation is what he is
2:30:35estimating.
2:30:36Now the left-hand side variable is the
2:30:40actual tariff we observe in the world
2:30:42either under WTO or under a preferential
2:30:45trade agreement. So these are both
2:30:46embedded into the data set. So it' be
2:30:48the lower the tariff with the little O
2:30:51on it is a counterfactual object which
2:30:53we don't observe in 2001 because that's
2:30:56not what countries are implementing. So
2:30:58that is Harry Johnson's opportunistic
2:31:01tariff or perfect optimal tariff. It's
2:31:03what Bagwell and Stagger would call
2:31:05their Nash tariff. And so that tariff
2:31:08they pull out of a quantitative trade
2:31:10model. And so they have to estimate that
2:31:13tariff for every importer, every
2:31:15exporter, and for every one of about
2:31:185,000 goods. They're going to get a
2:31:20model simulated value for the
2:31:22opportunistic tariff. And then they're
2:31:24also going to get a model simulated
2:31:26value for the responsiveness of income
2:31:28to trade flows. And then they're going
2:31:31to be able to estimate this. And the
2:31:33idea is that the beta if it is one that
2:31:36will tell us Bagwell and Stagger's world
2:31:38was perfectly theoretically correct. If
2:31:40it was zero it would say we're back in
2:31:42the world of Harry Johnson. But the core
2:31:44idea is that as beta gets closer and
2:31:46closer to one. We're getting more toward
2:31:49a world in which countries are valuing
2:31:52their own income gains from trade policy
2:31:55equally to the way they're valuing the
2:31:57income gains they receive when their
2:31:59trading partners give them more market
2:32:01access.
2:32:03Okay. And so the core results is
2:32:06actually pretty uh optimistic and very
2:32:08like a happy one. So I've got here that
2:32:10the beta was about 75. Um there are some
2:32:13other estimates in the in the paper but
2:32:17toward the end of the period the bait is
2:32:19actually getting a bit higher. So as
2:32:21Arno said it rejects perfect prao
2:32:23efficiency. Not surprising we kind of
2:32:26understood we weren't in that world. The
2:32:29other important thing though and this is
2:32:30something you could sort of see right in
2:32:32the raw tariff data. Countries that
2:32:34extend a lot of market access through
2:32:36low tariffs and through tariffs that are
2:32:38far below their optimal tariff rates
2:32:41also benefit a lot from lower tariffs.
2:32:44So the US and Canada. So when I looked
2:32:46at his heat map, I looked at well what
2:32:48what is the value of Canadian welfare in
2:32:51America. It was one of the highest ones.
2:32:52It's one of the dark blue ones. And
2:32:54similarly the highincome countries are
2:32:56extending a lot of market access to one
2:32:58another. Um and then he also can go
2:33:02through and look at a bunch of different
2:33:04counterfactuals once he's got um the
2:33:07beta weights. And so most countries are
2:33:12going to gain in income. Um and if all
2:33:15if sorry most gain median gain across
2:33:19the countries is about 1.2 two 1.9% of
2:33:23income and then um if all countries give
2:33:28the US just a huge amount of of benefit
2:33:32the US will gain um only a small amount
2:33:35but the core narrative is if we come
2:33:38back to Wilbur Ross's view of the world
2:33:41this is saying well the US is not
2:33:42uniquely giving away market access or
2:33:45giving away benefits actually they're
2:33:47quite beneficial they're benefiting in
2:33:50the system most countries are
2:33:51participating. Most countries are giving
2:33:53away some amount of real income to their
2:33:56trading partners in order to receive
2:33:58some back. Um, but everybody still
2:34:01values their own national income more
2:34:03than they do that of the trading
2:34:05partners. So, everyone's trying to be a
2:34:06little bit strategic as they're
2:34:08negotiating their trade policy.
2:34:10Um, I've said some of this already.
2:34:13There's a a vocabulary mapping. They use
2:34:17the term uh optimal tariff with as if
2:34:20altruism.
2:34:22That tariff is the sort of observed the
2:34:27observed tariff is equal to what we
2:34:30would call Harry Johnson would call the
2:34:32opportunistic tariff plus this component
2:34:34that's going to be measuring altruism.
2:34:37Um
2:34:39and then they can also go through and
2:34:42show in the model that they can embed
2:34:46domestic redistributive weights in the
2:34:48model. The results basically will hold
2:34:50up under that case. So you can say that
2:34:52some consumers are worth more than
2:34:53producers or different sectors are worth
2:34:55more building political economy in
2:34:58you're going to get the same basic idea.
2:35:00In that case, their optimal tariff with
2:35:03Aliff altruism is going to end up being
2:35:06pretty similar to Bagwell and Stagler's
2:35:08politically optimal tariff, which would
2:35:10be the one that countries successfully
2:35:12negotiate to under the GAT WTO system.
2:35:15Now, in terms of comparing uh to the
2:35:17literature, um Bagwell and Stagger do
2:35:20something looking at how countries have
2:35:24reciprocated or reciprocated negotiating
2:35:26tariffs in the Uruguay round. they start
2:35:30with an assumption of full reciprocity.
2:35:33The difference here is that this paper
2:35:36is going to say we don't know if there's
2:35:37reciprocity. We're going to start with
2:35:39the tariffs and then we're going to back
2:35:41out um how countries value one another.
2:35:44Um do country. So stepping back and
2:35:48thinking more broadly, do we believe the
2:35:50results? Well, there's a first question.
2:35:52Do we actually think countries do
2:35:54exploit market power with opportunistic
2:35:56tariffs? There's some papers in the
2:35:59literature that are kind of claiming
2:36:01that there's some evidence for this. Um
2:36:03the probably most cited one would be Bro
2:36:06Weinstein. I have one looking at uh
2:36:08trade policy in the um early 2000s.
2:36:15That said, the more recent studies
2:36:17looking at the 2018 trade war all find
2:36:21perfect pass through of the US import
2:36:24tariffs onto US consumers. And so this
2:36:27is one subtle point here in the model.
2:36:30If we just look at the prices of output
2:36:34markets for traded goods when we look at
2:36:36the 2018 trade war, the results on the
2:36:39surface are quite puzzling. Why? Because
2:36:42it suggests the US doesn't have market
2:36:44power. And so that would suggest that
2:36:46they don't have opportunistic tariffs
2:36:49that are much higher than zero. So I was
2:36:51discussing this with Arno before the
2:36:53session. Part of what's going on here is
2:36:56that in these models we think and
2:36:59including like the fosial bomb model the
2:37:02optimal tariff and the market power
2:37:04comes and is reflected in real wages. So
2:37:07it's coming on the real wage side that
2:37:08we see the market power that the US can
2:37:11exploit. We're not easily able to
2:37:13measure it in the traded commodities
2:37:15markets.
2:37:17Um,
2:37:19but
2:37:20this empirical evidence would seem to be
2:37:23against the idea of an optimal or
2:37:26opportunistic tariff if we're just
2:37:28looking at pricing in the final output
2:37:30market. So, the literature is a little
2:37:32bit all over the place. Um, I'll skip
2:37:35over this, but I'll just say if we think
2:37:38through the strengths and limitations,
2:37:41um, the key question about the paper
2:37:44would be do we trust the model? Because
2:37:46what's really important is going to be
2:37:48the quantitative model opportunistic
2:37:50tariffs that come out. If we believe
2:37:52those are correct, then we can very
2:37:54strongly believe the conclusions of the
2:37:56paper. They're very fair in how they're
2:37:58trying to present this. They're trying
2:37:59to show some sensitivity. Um but we are
2:38:03always a little bit, you know, it's a
2:38:05little difficult when we're trying to
2:38:07construct the real world against a
2:38:09counterfactual world that we never
2:38:11observe. Um and then the second
2:38:14limitation would be that there are
2:38:16multiple stories and multiple things
2:38:18that countries are concerned about not
2:38:20simply national income but we might have
2:38:23other measures. Okay. So very quickly um
2:38:26this has already been alluded to this
2:38:27morning um professor Krugman's talk but
2:38:30there's a question of what's trade
2:38:31policy for is it about maximizing
2:38:34national income or there other
2:38:36objectives in the objective function and
2:38:39so people in this room have done work uh
2:38:41Steve has worked on voting patterns in
2:38:43the UN Kevin's looked at preferential
2:38:46treatment of imperial colonies but the
2:38:48basic idea is there's one view in
2:38:50political science that says part of the
2:38:52objective of low tariffs and commerce
2:38:54nurse is to promote peace and this would
2:38:56be the idea of the Montescu hypothesis.
2:38:58It was very strongly lobbyed for by
2:39:00Cordell Hull when he was US Secretary of
2:39:02State in the 30s is thought to be some
2:39:04of the foundation of the gat. At the
2:39:07other hand we have Albert Hirschman
2:39:09telling us well trade is a tool of
2:39:11strategic state craft. You create
2:39:14dependencies once you have trading
2:39:16partners that are dependent on you. You
2:39:18use this to exert your political power.
2:39:22um in the context of this paper, both of
2:39:26these motives are going to lead us to
2:39:28tariffs below the optimal. And so that
2:39:32suggests that in the real income world,
2:39:35we can have some understanding of this
2:39:36is about reciprocity. But as a
2:39:39observationally equivalent thing, a
2:39:43country that's setting low tariffs in
2:39:45order to exert domination
2:39:47would also be setting tariffs below this
2:39:50optimal level. So it can be a little bit
2:39:51hard if we want to extend our our
2:39:53thoughts out to political science. So I
2:39:55will I'm over so I will wrap up there.
2:39:57Um but thanks very much to Arno for a
2:40:00really interesting an important paper
2:40:02that answers an important policy
2:40:04question.
2:40:07[applause]
2:40:10>> Thank you very much Meredith for the
2:40:12discussion. I'm going to pass the floor
2:40:13to Anu in case you would want to have a
2:40:15reaction to the discussion.
2:40:16>> Uh thanks so much for the comments.
2:40:18These were these were great. Uh maybe
2:40:21I'll take that as an opportunity because
2:40:23I I I skipped that part in in in my talk
2:40:25to talk about the the US narrative that
2:40:28that we do mention in the in the paper
2:40:31and
2:40:33it wasn't how we started this project
2:40:35and sort of a world events caught up
2:40:38with us. So we we decided to think a
2:40:41little bit about that narrative. So
2:40:42something that that's that that you do
2:40:44see both in the the average tariff and
2:40:47and in our betas is
2:40:51the the US if you think about the way it
2:40:54is being treated by its trading
2:40:57partners. So in terms of the beta it
2:40:59receives the United States receives a
2:41:01pretty high beta. So twothird of the
2:41:03countries in our data set receive less
2:41:05weight than the US does. So, so in that
2:41:08sense it doesn't look like uh uh the
2:41:12United States is being treated unfairly.
2:41:14What it is true if if you if you look at
2:41:18uh the US versus other countries is that
2:41:21circa 2001
2:41:24conditional on the weight it gives
2:41:26others it doesn't receive as much. So,
2:41:29so there is something to that benevolent
2:41:32hedgemon
2:41:34perspective on the United States in the
2:41:352001. They have lower tariff than the
2:41:38the rest of the world is imposing on
2:41:40them and and and you see that. And then
2:41:43the the the last thing we we did in the
2:41:45paper which we do a little sheepishly
2:41:47because you know I've told you those
2:41:50those betas that were estimated we think
2:41:52they're interesting but but they are
2:41:54they're not structural object. They're
2:41:56they're a mixture of things. So that's a
2:41:58snapshot and so any sort of
2:42:00counterfactual exercise where you you
2:42:02start treating them more structurally
2:42:05should should give you pose but we we
2:42:07went ahead and uh we we asked okay
2:42:10imagine just for the sake of the
2:42:11argument that uh all countries around
2:42:14the world put exactly the same weight on
2:42:17themselves and on the United States and
2:42:20they didn't care at all about anybody
2:42:22else. So everybody is just behaving
2:42:25maybe in the best possible way from the
2:42:27point of view of the United States.
2:42:29What's the upshot of that type of
2:42:31coercion maybe if you want and yeah we
2:42:35came up with a number that was something
2:42:37like 68%
2:42:40and that that's something that that's
2:42:41very common in those trade mall is that
2:42:43yeah the the the gains the welfare gains
2:42:45from that type of terms of trade
2:42:48manipulation I improve my terms of trade
2:42:51maybe not only using my tariff but all
2:42:53the tariffs around the world they're not
2:42:55going to be very very large the
2:42:57potential losses is inside the mall that
2:43:00they were an order of magnitude bigger.
2:43:04And then maybe I want to echo something
2:43:05that I was said yesterday by by by Kevin
2:43:08the you know the importance of the the
2:43:11geoeconomic consideration versus the the
2:43:14economic one
2:43:16when when you worried about what's
2:43:18happening now with the world trading
2:43:20system and what's changing
2:43:23I don't think this is cost of living
2:43:25consideration Harer triangle distortion
2:43:29this is more the fear that lack of
2:43:32international cooperation on trade
2:43:34policy implies lack of cooperation on a
2:43:37number of dimension like climate change
2:43:40like the war in Ukraine and that that to
2:43:42me is the the real concern. We're not
2:43:44speaking to these numbers but that's a
2:43:45real concern.
2:43:47>> Thank you very much Ando. Now I'm
2:43:49turning to the audience to see if
2:43:50there's any questions
2:43:53Alberto.
2:43:56>> Yeah. So uh thanks very interesting
2:43:59paper. I I have two questions. The first
2:44:01one is um you showed here the case of
2:44:04the US and China like a collapse in the
2:44:05beta. Do you see these collapses often
2:44:09in your data set and when they happen do
2:44:11they happen for instance for the US many
2:44:13betas fall at once or is it pair wise
2:44:15specific because this could help
2:44:17disentangle for instance in the backwell
2:44:19stiger view of the world a lot is about
2:44:21intertemporal cooperation. So if my rate
2:44:24of discount changed for instance well I
2:44:26would expect cooperation to fall across
2:44:28the board. So I don't know if we can
2:44:29learn anything from that. And another
2:44:32point which refers to the discussion
2:44:33which I thought was interesting was do
2:44:36we see low tariffs because of
2:44:37cooperation or do we see low tariffs
2:44:39because I'm trying to make you dependent
2:44:41somehow. Would a symmetry in the betas
2:44:44help you disentangle between these if I
2:44:46treat you in a way that looks too good
2:44:47to be true? Uh might that reveal
2:44:50something? Thanks.
2:44:53>> You want to take that?
2:44:55>> Uh yeah. Yeah. Uh both uh both excellent
2:44:58question collapse in the betas we we
2:45:01haven't done it yet. I I think using the
2:45:04the the more recent tariff changes that
2:45:06you know that's something one uh one
2:45:08could do. We were also thinking about
2:45:10going back in time using more of what's
2:45:12happening during the cold war and you
2:45:14could imagine some correlation in the
2:45:16betas across the different blocks uh
2:45:19etc. I I think uh
2:45:23you know both both your points
2:45:26have the flavor of okay what what can we
2:45:29learn structurally from the betas how
2:45:31does that help us discriminate between
2:45:34between different theories and so seeing
2:45:36some patterns in the betas if I think
2:45:38it's a discount factor shock in a
2:45:40repeated game I should see some
2:45:42association some correlation and so do
2:45:44more on that front to uh to help us shed
2:45:47light on mechanism I I think That's uh
2:45:51that's super interesting. Um and to your
2:45:55second point,
2:45:57yes, I think you know the pros and cons
2:46:00of of what we're doing that's more
2:46:01reduced from that's not speaking
2:46:03directly right now at the mechanism is
2:46:05that yeah it's very robust but it is
2:46:07possible that the reason why uh I put a
2:46:10high beta on you. So so think about
2:46:12Europe and the lack of tariff
2:46:14retaliation visa v the US. What what
2:46:17does that capture? I imagine that
2:46:18through the lens of my exercise, I would
2:46:20say, well, it looks like Europe still
2:46:22cares about the United States. And yeah,
2:46:24that's a reduced form for we worry that
2:46:27if we retaliate, we we're not going to
2:46:29get help on Ukraine and we need it right
2:46:31now. And and that's the that's what
2:46:32that's the structural uh channel.
2:46:37>> Yes, in the back.
2:46:43>> Thank you. Um
2:46:46tariffs are just one part of the
2:46:48protectionism right there is the
2:46:50non-tariff part as well right um and and
2:46:52I think this is very interesting I mean
2:46:54the the beta suggests there is a high
2:46:57free international trade um but there is
2:46:59the unknown part or the unobserved part
2:47:02which is the non-tariffer barriers right
2:47:04and WTO has been working with that for
2:47:07for many years to lower the non-tariff
2:47:10barriers they are unobserved there are
2:47:12regulations try to buy some some
2:47:14bakeries from France in Germany you know
2:47:17you don't see them right so there are
2:47:18all kinds of regulations that that
2:47:20protects so the question is um you know
2:47:22if we have a measure uh can we trust
2:47:26this these numbers or we should have a
2:47:28kind of a confidence band around them if
2:47:31we introduce the non-tariff non-tariff
2:47:34non-tariff barriers thank you
2:47:38>> thanks uh yes I mean the
2:47:41you can think about what we're doing as
2:47:43saying I'm going to observe the choices
2:47:46of your trade barriers and from that I'm
2:47:48going to infer how much you care about
2:47:49the world setting aside the mole and
2:47:52everything else it it does assume that
2:47:54I'm able to observe those trade barriers
2:47:56as you were saying even that part that I
2:47:59was saying this is what we prefer the
2:48:01rotary if we have the data well the
2:48:03trade barriers are the the rotary plus
2:48:05anti-dumping duties plus other things
2:48:08and so the true measure should take all
2:48:11those things into account so So we've
2:48:12done some uh I I'm forgetting whether in
2:48:15this paper we we did it as well. In
2:48:17other work we we've looked for example
2:48:19at anti-dumping duty. So everything that
2:48:21we can measure we can add to get closer
2:48:24to the true trade barrier and then redo
2:48:26the the exercise then uh when we can't
2:48:31uh because the data just isn't there.
2:48:34There's just a claim on our end that
2:48:36those considerations
2:48:38have to be either small in magnitude or
2:48:40uncorrelated
2:48:42with our altruistic motives. In that
2:48:45case, our identification works otherwise
2:48:47we have omitted variable bias and we'll
2:48:49we'll live with it. So, so we hope it's
2:48:51not too large but you know conceptually
2:48:53that's that's a concern.
2:48:55>> Kevin,
2:49:00>> thanks. If I've understood correctly,
2:49:02you're kind of looking at a world which
2:49:05is more or less non-discriminatory. And
2:49:07so if I particularly like you, I lower
2:49:09my tariffs particularly on goods that
2:49:11you export to me and so on. But I mean
2:49:14now, you know, Trump obviously does not
2:49:15care about the most favored nation
2:49:17principle. And the Europeans are
2:49:18increasingly saying we're not sure
2:49:20whether we are all that keen on it
2:49:22either, which I find very alarming. So I
2:49:26mean maybe your qualifications about
2:49:28your betas not being structural are are
2:49:31is is a very important qualification in
2:49:32this context but if we were to end up in
2:49:35a world without MFN there'd be more
2:49:38direct ways of helping people that we
2:49:41particularly like and could your
2:49:44framework subject to that beta
2:49:45qualification help us thinking about
2:49:47where we might end up.
2:49:49>> Okay that that's an excellent question.
2:49:51So MFN
2:49:53um I didn't mention that explicitly
2:49:57during the talk. I I probably should
2:49:59have. So the way we're thinking about G,
2:50:02a good G, I said bananas, but really I
2:50:06meant bananas from Costa Rica versus
2:50:08bananas from another country. So it's
2:50:10this arro view of a good uh meaning that
2:50:15we are allowing for discrimination.
2:50:18We're not imposing MFN at all and if
2:50:23countries abide by it, we're we're
2:50:25treating this as a choice from which
2:50:27we're learning about their their
2:50:29preferences. And what you see in the
2:50:32case of China at the end is a reflection
2:50:35that I can do less subtle thing than
2:50:38play with the the product mix of my
2:50:40time. I just discriminate against the
2:50:42country I want to hurt and that would
2:50:44show up as a as a decline in in beta. So
2:50:46actually we were not imposing MFN and
2:50:49and treated everything as a choice
2:50:51including you abiding uh by it. Thanks.
2:50:55>> Another question there.
2:51:01>> Terrific paper Arno. Um the real
2:51:03preference approach is extremely general
2:51:05and as Meredith was hinting in the paper
2:51:07you sort of consider alternative
2:51:09objectives such as including political
2:51:11economy forces which I guess shows up in
2:51:13the model in terms of how the
2:51:14opportunistic tariff changes and how the
2:51:15elasticity weighted by beta changes. I
2:51:18was curious when you incorporate
2:51:20political economy forces it seems like
2:51:21the results are very robust but I was
2:51:22wondering do you find on average higher
2:51:24betas or lower betas because you could
2:51:26tell different sorts of stories once you
2:51:28incorporate political economy how I care
2:51:30about foreigners. So for example, maybe
2:51:32it's partly that US firms lobbying for
2:51:34lower tariffs abroad uh because they got
2:51:37multinational subsidiaries there and so
2:51:38on. So I was just curious whether you
2:51:40see any pattern in the betas. And then
2:51:42the other side of the generalities is
2:51:44that you can estimate it without taking
2:51:45a stand about where the betas come from.
2:51:48Uh and I found your finding that um it
2:51:51doesn't look like WTO members versus
2:51:52nonWTO members that makes much of a
2:51:54difference. But I was I was thinking
2:51:56it'd be really interesting to kind of
2:51:57reestimate the model in the 1930s. And
2:51:59could you tell a story that the reason
2:52:00the betas are high today is partly
2:52:02because we've lived through this
2:52:03rules-based international order which
2:52:04has kind of promoted cooperation. And so
2:52:07if we looked in the 1930s, we might see
2:52:08much lower betas before we had that sort
2:52:10of rules-based order. So I'd be sort of
2:52:12curious about any thoughts on that.
2:52:15>> All right. Thanks a lot. Uh uh great
2:52:17questions on on the political economy.
2:52:22So the main thing that that we were
2:52:24worried about was something like uh
2:52:28trade protection in agriculture that you
2:52:31see very high tariffs in agriculture and
2:52:34through the lens of a mall where there
2:52:37is no concern for redistribution towards
2:52:39farmers you end up thinking that you
2:52:42don't like countries that export
2:52:43agricultural goods and that that's
2:52:45that's I mean that sounds like a crazy
2:52:47inference to make. So the the the mold
2:52:50we we've built the one with political
2:52:52economy motive. It doesn't have firms
2:52:54and multinational
2:52:56uh uh production. What it has is
2:52:59different types of workers in broad
2:53:02sectors, manufacturing, services,
2:53:04agriculture and allow nonparametrically
2:53:08countries to put different weights on
2:53:10those different constituents of of
2:53:12society
2:53:14and I don't remember the the details of
2:53:17how it changed. So I'm sure they they
2:53:19bounce around but but the average level
2:53:22of cooperation was was still there which
2:53:25we found uh reassuring on the the WTO
2:53:30the importance of the WTO I I fully
2:53:32agree look there's one view which is we
2:53:36see no difference today between WTO and
2:53:38nonWTO countries not because the WTO is
2:53:42useless but because it's super useful
2:53:43and its ideas have just propagated
2:53:46throughout the world and that's why
2:53:47there's no difference today uh and an
2:53:50interesting uh experiment and if you
2:53:53want to get at the the causal effect of
2:53:55the WTO on international cooperation is
2:53:58a pre-post exercise where you look at
2:54:00particular rounds of negotiation
2:54:03implement our methodology get the betas
2:54:05before get the betas after do we see
2:54:07this going up that that would be direct
2:54:09evidence speaking to that I think that
2:54:11would be super interesting we don't have
2:54:12the data for that but but I think that
2:54:14that would be that would be great
2:54:17>> see if there's any Other questions from
2:54:19the audience?
2:54:23>> One question on the cross product
2:54:25dimension. So within the WTO you're
2:54:27negotiating over a menu of such tariffs.
2:54:31So that means that you might make
2:54:33concession on one good in exchange of
2:54:36another. So maybe this is hidden in your
2:54:38optimal tariff. But then it it sounds
2:54:41like this optimal tariff becomes a very
2:54:43complicated object because you need to
2:54:45take into account all those
2:54:46interactions. So are you just ignoring
2:54:49those or are they hidden? And
2:54:53>> yeah, so
2:54:55you can think about this issue. So this
2:54:59is a real thing that uh one approach
2:55:02would be and that's what uh Bob Stagger
2:55:06and K Bagwell have been doing in more
2:55:08recent work with Euroglu. one view of
2:55:12which I think is super valuable of I
2:55:14want to understand the WTO is I'm going
2:55:16to model negotiation within the WTO. So
2:55:19I'm going to get data on bargaining and
2:55:21I know that they're bargaining
2:55:23simultaneously on different products and
2:55:24this is all part of how strategically
2:55:27they should think about uh the tariffs
2:55:30that they're setting. So optimal tariffs
2:55:31would be within that uh that set of
2:55:34institutional constraints. we're
2:55:36modeling that and and we're getting the
2:55:38the uh whatever equilibrium tariff comes
2:55:41out of that. We're not doing that. So,
2:55:43so the reason why we're not facing the
2:55:45problem is what I say at the beginning
2:55:49which is the world is complicated. There
2:55:51are all those reasons why you may
2:55:53cooperate and and the way this is you
2:55:55know going to be link across products
2:55:58etc may be complicated as well. But let
2:56:00me argue that for a number of morals the
2:56:04reduced form of that is going to be that
2:56:06as if altruistic motive and then let me
2:56:10let me estimate that. So you know the
2:56:12concern the real concern I think with
2:56:14your question would be to what extent
2:56:17that reduced form representation nest
2:56:20that type of we think it nest some
2:56:23repeated games etc. But but we should
2:56:25think more about whether it addresses
2:56:28that that type of negotiation.
2:56:34>> Last chance for a question
2:56:37LA.
2:56:42So I I briefly looked at your model so
2:56:47apologies if it's there. I was wondering
2:56:50how you're treating the intertemporal
2:56:52part of this like you do it in 2001.
2:56:542001 is a year where the US exchange
2:56:56rate appreciates. So I was wondering
2:56:58where where the exchange rate moves are
2:57:01in your paper but mostly where the trade
2:57:04deficit current account deficit or
2:57:05surpluses would end up that may affect
2:57:08perceived income or what they're buying
2:57:11and importing.
2:57:13And the other one is Costa Rica's main
2:57:15export is not bananas, it's biomedical
2:57:18devices.
2:57:19>> [laughter]
2:57:21>> Yeah. Okay. On Costa Rica, I'm going to
2:57:23start stop using that example. Diana
2:57:26made the same [laughter]
2:57:28comment to me last time. I know. I know.
2:57:30I I I I worried. I shouldn't have said
2:57:32that. [laughter]
2:57:36Uh on the the macroeconomic conditions,
2:57:40yes, the apology standard trademark,
2:57:43it's static. We don't have any interal
2:57:47consideration. Nothing interesting on
2:57:49that front. There are deficits in the
2:57:51data. Are we treating them as exogenous,
2:57:53not satisfactory overall, but depending
2:57:55on your question, this may or may not be
2:57:57important. In our case, most of what
2:57:59we're leveraging is this composition.
2:58:02You know, what type of products do you
2:58:04target with your tariffs? And so we see
2:58:08variation there. We according to to our
2:58:12model have very different sensitivity of
2:58:14real income to import restriction
2:58:16targeting different products. So the
2:58:18claim implicit is that those aggregate
2:58:21considerations of current account
2:58:22deficits etc the level of exchange rate
2:58:25we can abstract from it to get at those
2:58:27uh at those question but you know trade
2:58:30and macro should get together have more
2:58:32of an interaction but yeah not in this
2:58:34paper
2:58:37I don't see any further hands being
2:58:40raised with that I would like to thank
2:58:41you very much Arnu and also Meredith for
2:58:44[applause] presenting and discussing
2:58:45this great
2:58:49And with that one hour earlier than
2:58:51initially for seen um I can close this
2:58:54conference and I would very much like to
2:58:56thank actually all the organizers from
2:58:58the conference in particular Luca,
2:59:00Alistister, Eustina, Yorggo, Luke and
2:59:03Gok from the organizing committee but
2:59:06also the people from the audio
2:59:07multimedia team all the practical
2:59:09organization making sure you get food
2:59:11you get nice seats. Um just a word of
2:59:13thank you to all of them. There are too
2:59:14many to mention. So if I start, I will
2:59:16forget somebody. But uh I hope you give
2:59:18a round of applause to all of them as
2:59:20well. [applause]
2:59:27And also thank you for the audience
2:59:29online and here for attending the
2:59:31conference. And just one more logistical
2:59:33piece of information for those who are
2:59:34already hungry. There is being food
2:59:36served on this floor. I think if you
2:59:38follow those who look knowledgeable,
2:59:40you'll find the way uh to food or you
2:59:41may smell it as well. So, [laughter]
2:59:43okay. Thank you very much.
2:59:47[applause]

About this transcript

This page contains the full transcript of Live - ECB Annual Research Conference 2026 (Day 2) by European Central Bank, generated from the public captions YouTube serves with the video. The transcript has 23,149 words across 3,384 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.