00:00 Good morning,
00:00 everybody.
00:01 Welcome to this seminar organized by the Development Economics Research Group
00:06 on uh
00:09 Global value chains and trade in the age of COVID-19.
00:17 As you all know,
00:18 since the early 1990s,
00:21 we have
00:22 seen remarkable improvements in
00:25 transport,
00:26 information and communication technology
00:29 which,
00:30 together with
00:31 greater openness to trade and investment,
00:34 have led to the progressive deepening of integration in the global economy.
00:40 As production has fragmented across national boundaries,
00:46 we have seen
00:48 also deepening input-output relationships between countries,
00:53 and that has meant that the destinies of nations are increasingly intertwined,
01:01 uh,
01:01 which is a grand way of saying that GDP is increasingly correlated.
01:09 But this period of hyperglobalization.
01:13 Received the first big shock
01:16 with the
01:17 Great Recession
01:18 in 2008 and 2009,
01:21 and then we entered this period of what has been referred to as a great flattening
01:27 where trade as a share of GDP stopped growing.
01:32 And
01:34 Where global value chains stopped
01:37 their expansion.
01:40 Even as we were struggling to recover from that great recession,
01:44 a period of trade tensions and increasing
01:48 mutual protection between the United States and China
01:52 generated
01:54 further
01:54 uncertainty,
01:55 which also had a chilling effect on
01:59 trade
01:59 and global value chains.
02:02 Now we see
02:03 a third shock,
02:05 perhaps the most serious shock,
02:07 an unusual shock.
02:09 It's a shock to both demand.
02:12 Because people cannot spend.
02:14 And to supply because it's hard for people to work.
02:18 And it's affecting not one country,
02:20 like for example,
02:22 the
02:23 Japanese tsunami or the American mortgage crisis,
02:26 but it's almost at the same time affecting many countries,
02:31 as Daria Tagleoni has shown
02:33 that some of the
02:34 most important nodes in the global economy,
02:37 as
02:38 she calculated some time ago that 17 countries which had seen the
02:42 largest number of cases accounted for more than 70% of global trade.
02:49 And finally,
02:50 because it's an unusually
02:52 uncertain
02:53 shock both in depth and duration.
02:57 Therefore,
02:57 it's also turning into a financial shock.
03:01 So how these developments
03:05 interact
03:07 in the age of
03:09 flattening global value chains
03:11 will be discussed today by 3 luminaries in the field.
03:17 First,
03:18 we begin
03:19 with
03:21 Richard.
03:23 Baldwin,
03:24 who has really shaped our thinking in this whole area,
03:28 Richard is professor of international economics at the Graduate Institute.
03:33 He's been the editor in chief of VoxEU since June 2007.
03:39 I could say many things about his
03:42 multiple accomplishments,
03:43 but let me focus on
03:45 his 2016 book The Great Convergence
03:48 Information Technology and the New Globalization.
03:52 Lawrence Summers called it one of the 5 most important books on globalization ever.
03:59 And his more recent book,
04:02 The Globotics Upheaval,
04:04 is
04:06 been published in 2019
04:08 and has been almost as influential in helping us think about what's happening,
04:13 so
04:13 I was tempted to use Richard's original title of his Vox EU column,
04:18 which has many medical metaphors.
04:21 It was called
04:22 Supply Chain Contagion Waves.
04:25 Manufacturing contagion
04:28 and reinfection
04:29 from the COVID concussion,
04:31 but he has chosen today
04:34 a less dramatic title which is Thinking Ahead on COVID-19 and GVCs.
04:39 So Richard,
04:39 why don't we begin with you,
04:41 and I'll introduce the other speakers after you finish.
04:51 Richard,
04:51 you're muted.
04:57 It's a pleasure to be here with all of you.
04:58 What I want to do in the next 15 minutes is
05:02 run through a few points,
05:04 um,
05:05 not so much to inform you as to sort of inspire thinking and discussion
05:09 about what's going on.
05:11 I think there are big implications of COVID for global value chains,
05:15 but I think a lot of people are,
05:16 are misthinking them and,
05:18 uh,
05:19 missing a few points.
05:20 So,
05:20 let me just get started with
05:23 What I'm hoping to do today.
05:25 So,
05:25 the,
05:26 if I can get this.
05:29 There we go.
05:30 The outline of the talk is I'm gonna have 3 parts.
05:32 The first is the nature and the duration of the shock.
05:35 Now,
05:35 this is gonna involve talking a little bit about medical things,
05:39 but one of the first aspects of misthinking,
05:42 which I think is systemic is
05:44 People not realizing how long and how deep and how
05:48 how
05:49 uncoordinated this shock is going to be.
05:52 And
05:52 the second part is
05:54 why this shock is so different from the 2008,
05:57 2009 1.
05:58 And then the last,
05:59 I'm going to talk about the supply contagion
06:02 and re-contagion.
06:04 And if I have time,
06:05 I'll say a few words about
06:06 whether there's going to be long-term damage to global value chains or not.
06:10 So,
06:11 let me just start getting into the medical imperative.
06:14 So,
06:14 we've all heard
06:15 that we have to flatten the epi curve or the epidemiological curve.
06:20 Um,
06:20 and let me just go through that because I think it's
06:23 important to understanding the length and the nature of this thing.
06:26 So,
06:27 what this blue curve here is the number of new cases,
06:31 uh,
06:31 and when it,
06:32 and this is uh on the horizontal axis week,
06:35 so the,
06:35 the vertical axis is the number of people.
06:38 And in the beginning,
06:39 it doesn't look like very much.
06:40 That's the discovery and initiation phase,
06:43 but then we go into the acceleration phase
06:45 where people are
06:47 getting other people sick who are getting other people sick,
06:50 and the number of new cases can
06:52 double
06:53 every 2 or 3 days.
06:54 So it tends to overwhelm government's ability
06:57 to think ahead or plan
06:59 because by the time they form the committee,
07:01 it's already doubled or tripled or quadrupled what it was before.
07:05 But um,
07:06 as the number of people getting
07:09 the new infection ramps up,
07:11 with about a 2-week lag,
07:14 the number of people who are recovering is also ramping up.
07:17 And I'll,
07:17 we'll,
07:17 we'll follow that for a second.
07:18 We'll,
07:19 we'll show that,
07:19 that that actually has a,
07:21 a big implication.
07:22 But the point about this is,
07:24 these are the new infections,
07:25 and one of the interesting things,
07:27 one of the important things about this disease
07:29 is that although it's
07:31 explosively contagious,
07:33 it's not very deadly.
07:34 So the number of hospitalizations
07:37 coming from these new cases is about 11/10 or 15% of what they are.
07:41 But each one of these new hospital admissions stays
07:44 in there for 2 weeks or 3 weeks sometimes.
07:46 So the cumulative number of people in the hospital
07:49 climbs
07:51 much higher than the than the number of hospitalizations,
07:55 and it peaks a little bit later.
07:57 So if the number of hospitalizations overwhelms the hospital capacity,
08:01 we get this overload and excess deaths.
08:03 In fact,
08:04 it can be quite traumatic,
08:05 the things you've seen going on in Italy
08:07 and in New York and in Spain.
08:10 What went on in Wuhan,
08:12 these are things that really
08:13 make
08:15 people very,
08:15 very upset,
08:16 and governments
08:17 take extreme measures to avoid this kind of overwhelming things.
08:21 So that's essentially where the imperative comes,
08:24 because more people will die than is necessary,
08:27 and the way they die
08:28 is very
08:30 traumatic,
08:30 and people
08:31 are,
08:31 are view it as absolute imperative not
08:34 to,
08:35 not to do that.
08:36 So,
08:36 avoiding this yellow is not a balancing between GDP and jobs or
08:41 dollars and debts,
08:42 it's an imperative.
08:45 Now,
08:45 what I want to talk about
08:46 is,
08:48 uh,
08:49 next,
08:50 how this is connected with the economic shock,
08:53 the recession,
08:53 which then hits GDPs.
08:55 So,
08:56 if you take this yellow,
08:57 this red curve as what the epidemiological
09:00 curve would look like without containment policies,
09:02 in order to avoid that overloading,
09:05 we have containment policies,
09:06 lockdowns and things like that,
09:08 that flatten and spread out the the epi curve.
09:12 Now,
09:12 those containment policies
09:14 also make the recession,
09:16 which would be rather
09:17 short and not too,
09:19 not too deep if it,
09:19 if there were no containment policies.
09:22 It makes the recession much deeper.
09:24 And in rich countries,
09:26 at least,
09:26 governments use economic policy to flatten the recession curve
09:30 by spending truly
09:31 historic sums of money to try and protect the economy and also
09:36 to
09:38 share income and redistribute the cost a little bit.
09:41 So that's how we have these containment policies are a medical imperative,
09:45 and at least in rich countries,
09:47 they're using the economic policy
09:49 to counter them.
09:50 But it's worthwhile saying that
09:53 the
09:54 Economic
09:55 recession
09:56 is really a health,
09:59 a health measure.
10:00 Now,
10:00 the containment policies
10:02 are inducing a rather unusual recession.
10:05 Most recessions are just a lack of spending,
10:07 and central banks or governments can
10:10 undo the recession by stimulating spending,
10:13 but this containment policy is hitting the supply side and hitting the
10:17 The supply chains and the financial sector,
10:21 all of this all at the same time,
10:23 and it's done on purpose,
10:24 so it's difficult to just fix one bit or another bit.
10:27 It's shutting it down quite systematically.
10:30 So as a consequence,
10:31 this recession
10:32 is Complex supply shock
10:35 and a demand shock all at the same time.
10:37 Moreover,
10:38 it's happening
10:39 at different times in different countries around the world.
10:42 And because they're linked by global value chains and also just regular trade,
10:46 the shock is being transmitted
10:49 internationally.
10:51 So the next thing I want to talk about is how long will the shock last?
10:55 Are we talking about a couple of quarters,
10:57 as it seems like many of the
10:59 investment banks and some of the international organizations
11:02 are
11:03 projecting that,
11:04 that the second half of this year
11:06 will all be positive and will start to come back,
11:08 or are we talking about a couple of years?
11:11 And my opinion is I think people are way underestimating how long
11:14 this shock is going to last.
11:16 And I want to convince you that
11:18 by looking at a little bit of epidemiology.
11:21 And we've all had to become a little bit uh epidemiologists,
11:24 uh,
11:25 in,
11:25 in,
11:25 in the,
11:26 uh,
11:26 the last couple of months.
11:28 The good thing is epidemiologists have had to
11:30 learn a little bit of economics as well,
11:31 so.
11:33 So,
11:34 the model that,
11:35 that,
11:35 this is a,
11:36 this is like the supply and demand model of epidemiology.
11:39 It's called the
11:40 SIR model.
11:42 So,
11:43 I for infectious,
11:45 S for susceptible.
11:46 I,
11:46 I made a mistake,
11:47 that should be an S,
11:48 and R for recovered,
11:49 that's where the model name comes from,
11:51 SIR.
11:52 Now,
11:52 here's how it works.
11:53 You divide the population into three bins,
11:56 infectious people,
11:57 susceptible people,
11:58 and recovered.
11:59 People who are infected can give it.
12:01 People who
12:02 are susceptible can get it,
12:04 and people who are recovered
12:05 are immune.
12:06 Now,
12:06 the new cases is what we were talking about before,
12:09 that's the epidemiological.
12:10 This is the explosive contagion in the first place,
12:13 and then there's a deceleration,
12:15 and there's a steady stream of people who are recovering,
12:18 going from infectious
12:19 to recovered.
12:21 Now,
12:22 uh,
12:22 when you want to look at the dynamics of that whole thing,
12:25 It's useful to first of all point out that this is exactly what China,
12:30 China did.
12:30 They went through the,
12:31 this is the daily cases.
12:33 This is the epi curve.
12:34 They went through the accelerating phase,
12:36 the decelerating phase,
12:37 and now they seem to have at least
12:39 put it to bed,
12:40 although they're getting a second wave.
12:42 We'll talk about that.
12:44 But to think about how long this is gonna last,
12:46 I,
12:47 I did some illus illustrative dynamics,
12:49 so don't take any of this serious.
12:51 I'm not an epidemiologist,
12:53 and lots of people are doing real simulations.
12:55 I just want to illustrate the dynamics
12:57 so we can get our mind around how like,
13:00 likely,
13:00 how long this thing is likely to happen.
13:02 So this is when there is no
13:04 reduction in the,
13:06 um,
13:06 in the containment policies or anything.
13:08 So this orange thing,
13:10 the num,
13:10 the stock of infected people rises very quickly at peak,
13:14 and then it comes down.
13:15 And what that's consisting of is this red curve,
13:18 which is the new cases.
13:19 So that's the epi curve over here,
13:21 it looks like that.
13:22 And the green is the recovered cases,
13:24 which looks a lot like the new cases,
13:26 but just delayed because of the recovery.
13:29 Now,
13:29 when the recover number of
13:31 new cases equal the number of recovered,
13:33 that's when we're at the peak,
13:34 and once we get past the peak,
13:36 we can start to get better.
13:37 And what the black curve is showing is the number of people who've had it,
13:41 and that starts rising and then it sort of levels off
13:44 as the stock comes down.
13:46 OK.
13:47 So,
13:48 what I'd like to do is put some orders of magnitude
13:50 on this to argue this is way far from over.
13:54 So I'm just gonna do some simple math with this model.
13:57 Beta is the number of infectious contacts per period
14:01 per infected person.
14:03 So that's a,
14:03 a,
14:04 a medical kind of parameter,
14:05 but it's also connected
14:07 with,
14:07 or can be influenced by social distancing.
14:10 So the number of new cases is this beta parameter
14:13 times the number of people who are infected
14:16 times
14:17 the fraction of the people they meet
14:20 that are
14:21 susceptible.
14:22 So if,
14:23 if half the population has it,
14:25 so n is not the size of the population,
14:28 s is the number of people susceptible,
14:30 so if this ratio is 15,
14:32 then only half of those contacts by infectious people will actually lead
14:36 to new infections.
14:38 The new recover,
14:39 recoveries is even easier.
14:41 It's just gamma,
14:42 which is a recovery rate,
14:43 which is medically given,
14:45 times
14:46 i.
14:47 So the top of the infection is when the new cases equal the new recoveries,
14:51 and if you just cancel out the i,
14:52 you get this extremely simple formula
14:55 that the fraction of the population who has not had it,
14:59 that,
14:59 that are still susceptible
15:00 is gamma over beta.
15:02 Now,
15:02 gamma over beta
15:04 is equal to 1 over R0,
15:06 and R0 is this very famous statistic that gets estimated all over the time.
15:10 And the estimates for COVID is that R0 is between 2 and 3.
15:14 So that means
15:16 when R0 is not modified by containment policies,
15:20 something like
15:21 between half
15:22 and
15:23 a third of the people will not get it,
15:25 and everybody else will get it.
15:27 Now,
15:27 in,
15:28 we don't know exactly how many people have got it,
15:30 but it's at least an order of magnitude less than would need to be
15:35 before this thing goes into
15:37 herd immunity and getting rid of it.
15:39 Now,
15:39 the point of that is
15:41 it,
15:41 a couple of points.
15:42 One,
15:42 if you reduce
15:44 beta
15:45 by
15:45 50%,
15:46 let's say through containment policies or mask or whatever you do,
15:50 then you're going to double the number of people who do not get it,
15:52 so you save a huge number of lives.
15:55 But as soon as you let beta go back up to where it was before
15:59 by relaxing containment policies,
16:01 this dynamic will come again.
16:03 So there almost surely will be 2nd and 3rd waves.
16:06 As we start to relax the restraints,
16:08 we will get more and more waves,
16:10 and they're also,
16:11 importantly for global value chains,
16:13 they're not synchronized across the world.
16:16 OK.
16:18 So,
16:18 now what I want to do is argue that this is different.
16:21 So that
16:22 in 2008,
16:23 2009,
16:24 it was sudden,
16:25 severe,
16:25 and synchronized,
16:26 but it was basically all coming from the demand side.
16:29 So,
16:30 apart from the United States,
16:31 the UK and a few other very,
16:33 very clever countries
16:34 who
16:35 got into this financial engineering,
16:37 most countries in the world had nothing to do with the subprime crisis.
16:41 So their financial sector was not directly hit by this.
16:45 For most of the world,
16:45 it was a great big demand shock which led to a great big trade collapse,
16:49 and all the global value chains shut down at the same time.
16:53 So the global value chain contagion
16:55 wasn't really that important.
16:57 During the 2009,
16:58 because everybody shut down at the same time and people started coming back up
17:02 more or less at the same time.
17:04 And in any case,
17:05 the supply chains weren't damaged.
17:07 What happened was there was just a lack of aggregate demand.
17:09 And once the demand was restored,
17:11 we were good to go.
17:13 This time,
17:14 as
17:14 Addita mentioned that Daria has calculated,
17:17 it hit all the biggest traders and manufacturers within 2 months,
17:20 more or less simultaneously,
17:22 but on the supply side and the demand side,
17:25 so there was,
17:26 uh,
17:27 there has been some damage in terms of bankruptcy,
17:30 but in any case,
17:31 the supply chain was directly shut down either by transportation limitations
17:36 or by lockdowns of factories.
17:39 The 2nd,
17:39 the 3rd point
17:40 is that global value chain integration is
17:43 much bigger than it was back then.
17:45 And here you can see
17:46 a slide that I put together.
17:48 This is from a,
17:49 a Vox column that I posted a few weeks ago and was calculation by Rebecca Friedman.
17:56 So the exposure,
17:57 what this shows is a higher exposure to inputs.
18:00 Now,
18:00 what this table does,
18:01 it shows the total exposure
18:03 of the row nation's manufacturing
18:06 to the column nation's manufacturing inputs in 2015.
18:10 So you can see the
18:12 inputs from the US,
18:14 from Germany,
18:15 and for China are are important for basically everybody,
18:19 and Japan and Korea are also basically important for everybody.
18:22 And countries,
18:24 their sus susceptibility to the inputs from other countries
18:28 um
18:28 are shown here and you can see there's a regionalization of these.
18:31 This is factory North America,
18:33 this is factory Europe,
18:35 and this is factory Asia,
18:36 but Asia is important for everybody and so is Germany.
18:40 So the idea here is
18:41 if
18:42 factory China shuts down,
18:43 for example,
18:44 Mexico
18:46 is
18:46 Mexico's value of manufacturing is heavily dependent upon China.
18:50 So when China's shut down,
18:52 that's going to affect Mexico.
18:54 And when Mexico shuts down,
18:56 that's going to affect the United States because
18:58 the United States is,
19:00 is
19:01 reliant on
19:02 Mexico for the,
19:03 this is 1.6.
19:05 So,
19:05 you get these echoes going through,
19:07 and these numbers are larger
19:09 than they were before.
19:10 So,
19:10 it,
19:10 it's,
19:11 it's a,
19:11 it's um a different crisis and it's uh gonna affect through global value chains,
19:16 uh,
19:17 more importantly.
19:19 So the,
19:19 the
19:20 last point I want to talk about is this contagion.
19:23 So,
19:23 I'm just going to take the example of the United States
19:26 and uh the US
19:28 and
19:28 those two countries,
19:29 uh,
19:30 manufacturing sectors are,
19:31 are pretty integrated.
19:33 So,
19:33 when,
19:34 when China shut down,
19:36 that
19:38 impacted manufacturing in the United States and China also
19:42 um
19:42 was a source of the contagion to the United States.
19:45 And
19:46 now,
19:47 China's getting better,
19:48 so their manufacturing sector is back on its feet,
19:52 but the US is experiencing a huge outbreak of COVID and therefore,
19:56 its manufacturing is shutting down
19:58 either because of plant closures or difficult of transportation.
20:02 Turns out there's a lot of air transportation and sea transportation problems.
20:06 So the supply chain problems
20:08 that
20:10 China is now facing is in part because
20:12 of their supply chain dependence on the United States
20:15 and the US and other people are reinfecting China.
20:18 So the idea is that this contagion and reinfection
20:21 could continue
20:23 for quite a while as long as
20:26 we don't have a vaccine or we don't acquire
20:29 herd immunity.
20:31 So,
20:31 um,
20:32 I don't know how am I doing on time,
20:34 Adita?
20:34 Do,
20:34 should I,
20:35 I have a couple more slides I don't want to talk about the future,
20:38 or should I just stop here?
20:39 I think,
20:40 I think I started,
20:41 uh,
20:42 Richard,
20:42 I don't you take another minute because that's gonna be interesting to hear
20:45 about how you see things evolving.
20:48 OK.
20:49 So,
20:49 um,
20:51 my argument is that
20:53 the,
20:53 the value chains are not gonna be affected,
20:55 and I think there's a great deal of wooly and wishful thinking.
20:59 And instead,
20:59 we should think about sources of hysteresis.
21:02 So there's all sorts of people who said,
21:03 oh my goodness,
21:04 we can't rely again
21:06 on X,
21:06 Y,
21:06 and Z country.
21:07 We should have it all at home.
21:09 First of all,
21:09 that's just wrong because you should be diversifying risk,
21:12 not concentrating it.
21:13 But the second
21:15 is,
21:15 did anything really change?
21:17 When the European people and the European Commission or the European Council
21:21 say we should think about not being so reliant on China,
21:25 well,
21:25 unless they actually changed anything,
21:27 it's going to go back.
21:28 Because
21:29 you need a permanent change in global
21:30 value chain that will require government subsidies,
21:33 government-mandated stockpiles,
21:35 diversification,
21:36 or massive tariffs on inputs and final goods to move anything.
21:40 So,
21:40 I'm quite willing to accept that when it comes to medical equipment,
21:43 we'll see some changes.
21:45 But are people really going to be willing to pay
21:48 10,
21:49 20% more for their cars,
21:50 50% or 100% more for their electronics,
21:53 for all sorts of gadgets that are made around the world?
21:56 Are they going to do that on the abstract
21:59 possibility
22:00 that something's going to happen in the future?
22:02 Because after all,
22:03 the
22:04 Global value chain allocation,
22:06 what we actually have now is really balancing reward cost savings against risk.
22:11 It's just a portfolio problem.
22:13 And unless something moves this indifference curve,
22:15 which is set by private firms interested in a profit
22:18 this quarter
22:20 or something moves the risk reward frontier.
22:23 We're gonna go back as soon as the shock passes to the way it was before.
22:27 So,
22:27 I think a lot of the discussion about this is
22:29 finally the time we're gonna unwind global value chains,
22:32 I think that's wishful and wooly thinking.
22:34 So I'll stop right there.
22:36 Thank you very much.
22:39 Thank you,
22:39 Richard,
22:39 for that excellent introduction.
22:42 Um,
22:43 after Richard,
22:44 we have,
22:45 uh,
22:46 Vasco Carvalho.
22:48 He's a professor of economics at the University of Cambridge,
22:53 professor of macroeconomics,
22:54 and also the senior Keynes Fellow
22:57 in Cambridge.
22:59 Again,
22:59 he has done a lot to shape our
23:02 understanding of how production networks affect macroeconomic propagation.
23:08 He was awarded the 2014 Wiley Prize in Economics by the British
23:12 Academy for achievement in research by an outstanding early career economist.
23:19 He's the principal investigator of the European Research
23:22 Council grant Macronets Production Networks in Macroeconomics.
23:27 And also a recipient of the Lehoon Prize Fellowship.
23:32 Um,
23:34 Vasko did not share with me,
23:36 as far as I know,
23:37 his presentation,
23:38 but he
23:40 did
23:40 remarkable work
23:42 helping us to understand
23:44 how the Great East Japan earthquake.
23:48 Propagated through
23:51 Japan.
23:52 And how input output linkages were a mechanism for the propagation
23:58 and amplification of that shock,
24:00 but I see
24:02 Vasco is going to make a more general
24:05 presentation which will nicely complement Richard's introduction.
24:10 And international shock propagation.
24:13 Vasco's past work that I'm familiar with
24:15 looked at shock propagation within countries.
24:18 So I'm going to assume that this will continue
24:20 in the same way unless Vasco has internationalized his
24:26 analysis.
24:27 Vasco,
24:27 over to you.
24:29 Thank you very much,
24:30 Richard.
24:30 Thank you for the invitation.
24:31 This certainly beats uh teaching fractions to my kids,
24:35 which is what I'm otherwise doing.
24:37 Um.
24:39 So,
24:40 yes,
24:40 that's what the first thing I was going to say is that
24:43 you will not see international trade or global in my slides.
24:48 Um,
24:48 I do think
25:11 The thing I would like to do
25:13 is to
25:15 just take a step back and and notice that
25:19 Um,
25:20 in economics,
25:21 the role of supply chain linkages as a
25:24 source of risk and as a source of synchronization
25:27 is almost as old a dis discipline.
25:29 I'm not even gonna go back to Quesna and the Tau Economique,
25:33 um,
25:34 but it's been there.
25:35 The first general equilibrium model by von Neumann is a multi-sector model by,
25:40 by,
25:40 by the liberation of the author.
25:42 The beginning of growth theory is a multi-sector models.
25:46 The beginning of modern business cycle theory with real business cycles
25:50 is multi-sector.
25:51 And of course there's Leon TF writing much more beautifully than
25:55 anyone writes now.
25:57 reminding us back in the 30s and 40s how
26:01 laymen and professional economists and practical planners
26:04 are all equally aware of the existence
26:05 of some kind of interconnection between even
26:08 the remotest parts of the national economy.
26:11 And he goes on to say that the presence of
26:13 this invisible but nevertheless very real ties can be observed
26:17 in the synchronization of say,
26:19 demand in New York City
26:21 and how that translates
26:23 into
26:24 groceries in Detroit or mines in Pennsylvania to
26:28 textile mills in New England.
26:30 And then reassert itself over the business cycle,
26:33 right?
26:33 And so this is a very
26:36 There's nothing new here.
26:39 There is an increasing policy awareness.
26:42 So
26:43 my understanding is that a national strategy for the
26:46 US is a step just below a national agency.
26:49 The US has a national strategy.
26:51 Uh,
26:51 for global supply chain security,
26:53 acknowledging again
26:55 the disruption to supply chains as a source of global economic risk,
27:03 right,
27:04 be it by natural disasters,
27:05 be it by
27:07 criminals targeting
27:08 the
27:09 system.
27:11 Um,
27:12 and of course,
27:13 as Leontiev told us,
27:16 uh,
27:17 these supply chain risk aspects
27:21 tend to reassert themselves with relentless regularity.
27:25 So this is just a cut from the Financial Times a month ago or so.
27:30 Uh,
27:30 which was exactly what Richard was saying about the,
27:34 the
27:35 The first
27:38 from economic contagion from China early on in this process.
27:42 So having set the stage,
27:43 I was kind of asked to do two things.
27:45 I want,
27:46 I was asked to review a little bit evidence on
27:49 that the supply chain,
27:51 uh,
27:51 linkages do propagate risk across uh otherwise largely independent units.
27:57 Uh,
27:58 and then I was asked to do some kind of,
28:00 um,
28:01 material for discussion.
28:02 So let's try to do both.
28:05 Uh,
28:05 uh,
28:05 I'll start with,
28:06 um,
28:08 The review of paper on
28:10 the Great East Japan earthquake.
28:12 I want to name my co-authors here,
28:14 which is Makoto Niere,
28:15 Yukiko Saito in Japan,
28:16 and then Alireza Taba Sale in northwestern.
28:20 Um,
28:21 and kind of
28:23 There's the
28:24 the starting point for this paper was there was lots
28:26 and lots of evidence out there about how supply chains
28:30 may be a source of macro risk,
28:32 and they're assumed so by policymakers,
28:34 but there's very little systematic evidence.
28:37 From the point of view of kind of causal strictures that we impose ourselves,
28:42 we're talking about two fundamental challenges.
28:44 The first would be,
28:45 can you identify
28:48 a supply chain shock,
28:50 an individual localized shock.
28:54 Uh,
28:54 that's number 1.
28:55 And can you then,
28:57 number 2,
28:59 having had such shock or plausibly exogenous shock,
29:02 can you trace the shock as it,
29:04 as it spreads to the network?
29:05 So we're going to look at data from Japan in 2011 and try to get some basic magnitude.
29:10 And then depending on time,
29:12 I'll I'll go for for a more general discussion of what we learned a little
29:15 bit in Japan or the discussions within Japan
29:17 at the time that are relevant more globally
29:20 now.
29:21 So if I can take you back
29:23 uh nine years now,
29:25 uh,
29:25 in March 11,
29:26 2000,
29:27 so,
29:27 so basically 99 and 9 years and 10 days now.
29:31 It was a triple disaster.
29:32 There was a
29:34 a very large earthquake off the Pacific
29:38 northeast coast of Japan
29:40 that started a
29:42 tsunami
29:43 which
29:43 then ended up
29:45 hitting a nuclear power plant in Fukusushima.
29:48 Uh,
29:49 what you have there is a graph of fatalities and essentially human loss in red and
29:55 physical capital loss in green,
29:58 and what you see is that,
30:00 you know,
30:00 despite being a horrible
30:02 shock,
30:03 it is relatively localized in Japan.
30:05 It affected.
30:07 It's northeast prefectures
30:09 and
30:11 in particular the seaside prefectures because of the
30:15 The tsunami.
30:17 So that we're going to exploit that as our
30:20 first uh dictum that we needed,
30:22 the exogenous large localized disruption.
30:26 Um,
30:26 just to remind you,
30:27 you know,
30:28 I'm gonna very briefly go over this.
30:30 This is not the data that we'll be using,
30:32 but just as a reminder of what happened in Japan then,
30:35 there's very little high frequency data.
30:37 This is one snippet of high frequency data that we can show to you,
30:41 which is a year on year monthly production industry
30:45 index by prefecture.
30:47 And what you see in the red line is a 40%
30:49 year on year decline of industrial production in the affected prefectures.
30:53 And uh
30:55 a relatively strong recovery throughout
30:57 the months of the year that follows.
31:00 And in a way,
31:01 the question of this paper of this research is,
31:05 uh,
31:05 if we take a look at the blue line,
31:07 that's the entire Japan industrial production and
31:10 how it evolves throughout the year.
31:12 And what we know is that the disaster area by simple
31:16 national accounting by adding up is too small to account for
31:20 a 50% decline of
31:23 industrial production in Japan on impact
31:26 and it,
31:27 you know,
31:27 recovered throughout the year.
31:28 So we are
31:29 going to try to source firm level data
31:32 to see if this can come about,
31:34 if the translating the red line to the blue line is an aggregate outcome.
31:39 This can come about by supply chain propagation of this disruptions.
31:45 How are we going to do this?
31:46 We're going to access a very large firm level database.
31:49 So
31:50 essentially we're going to look at which firms are behind this red line.
31:54 What are the firms that are producing,
31:57 not only in services,
31:58 the database is very broad and covers
32:01 the entire
32:03 landscape of the economy very thoroughly.
32:05 And then we're going to overlay this with
32:08 3 governmental decrees essentially establishing a disaster area,
32:12 an emergency area following the earthquake,
32:16 and we're going to take advantage of the fact that we know
32:19 the address of the firms to geolocate them inside these disaster municipalities,
32:27 which is a subunit of these prefectures that we use.
32:31 Now,
32:32 having said this,
32:33 and you can see them again plotted against the northeast coast of Japan,
32:36 these like little reddish dots are the ferns and their location.
32:41 Having said that,
32:42 we don't care about the disaster area in this paper or the fate of these firms.
32:48 What we care about is the supply chain
32:52 link through to these firms,
32:54 and what you're seeing there is the same database as,
32:59 which is something very valuable and still somewhat unusual
33:02 for about a million firms in Japan.
33:04 That's about 2/3 of all the firms in Japan.
33:08 It has information of suppliers and customers of each firm on the identity,
33:12 not the quantity of trade.
33:15 So in the topmost graph,
33:16 I am plotting again the disaster firms that
33:19 we just identified previously in the graph.
33:22 And then I have 4 degrees of separation over
33:26 the production network or over the supply chain.
33:29 So what are the reds,
33:30 the second
33:32 up,
33:32 the 2nd graph.
33:35 Uh
33:37 coming from top to bottom.
33:40 Those are all direct suppliers or direct customers of affected area firms.
33:47 These are firms who are headquartered in
33:50 away from the disaster prefectures,
33:52 right?
33:54 But to have direct input-output linkages either by being suppliers
33:58 or customers to disaster area firms that we just.
34:03 And then I can go on.
34:04 I can ask,
34:05 OK,
34:06 who are the suppliers
34:08 and customers of the red firms?
34:10 So orange firms
34:12 are firms that
34:14 um
34:16 are what we call network distance to or supply chain distance
34:20 to from the disaster areas that they are not directly linked.
34:23 They are not direct customers or suppliers,
34:26 but they're either suppliers of the suppliers or
34:28 customers of the customers of disaster areas.
34:32 And then we go,
34:33 uh,
34:34 and,
34:34 and we continue going,
34:35 uh,
34:36 for the network buffs out there,
34:38 the diameter is 7,
34:39 which means we'd run out of firms at distance 7.
34:42 Uh,
34:42 for the popular culture buffs,
34:45 this is very much like the Kevin Bacon
34:48 degrees of separation.
34:49 The average distance is small.
34:51 That is economically relevant because it means that
34:54 within
34:56 In between,
34:57 somewhere in between the orange and the green ferns,
34:59 you cover 50% of the firms in Japan.
35:04 Which means modern supply chains are small worlds in their technical definition,
35:08 meaning
35:09 Uh,
35:11 they are sparse in that most people,
35:13 most firms are not trading directly with each other.
35:17 But they are nevertheless indirectly very close.
35:20 They are 2 or 3 links away from the disaster area.
35:24 And this is nothing in particular that is special to this disaster area.
35:28 This is a general feature of modern production networks in the data
35:32 that we are that is coming online and we can see.
35:35 Now,
35:36 having that,
35:36 what I want to do,
35:37 the outcome of interest now is to understand what was the impact
35:41 of being
35:43 Uh,
35:44 say,
35:45 Of the earthquake on non-disaster firms,
35:49 right,
35:49 firms that were linked
35:51 to disaster area only by their supply chain linkages,
35:55 either directly,
35:57 and then I would call that
35:59 distance one,
36:00 either upstream or downstream,
36:02 or indirectly,
36:03 which would be distance 23,
36:05 and 4.
36:06 So,
36:07 what we're trying to do here is basically saying,
36:11 Relative to our control group,
36:13 and our control group here to be clear,
36:16 is firms that are distance 5 and above from disaster area firms,
36:20 firms that are very remote in a supply chain sense
36:24 from the disaster area.
36:26 The firms that were closer in the supply chain distance,
36:30 um,
36:30 either downstream or upstream,
36:33 uh,
36:33 uh,
36:35 uh,
36:35 can we see an impact on their,
36:37 the level of their sales in the following year,
36:40 and this is the yearly frequency.
36:43 I'm just gonna now show you the main headline
36:45 result at this kind of the micro reduced form.
36:48 The,
36:48 in the paper,
36:49 you will find lots of robustness checks and and more discussion.
36:53 But this would be the key outcome.
36:55 So to read this graph,
36:56 let's just focus on the blue
36:59 bars,
37:00 and the circles are the point estimates.
37:03 And so this is saying that in the year following the earthquake,
37:08 a firm that was a direct customer,
37:10 Downstream One.
37:12 Of,
37:12 of,
37:13 of disaster area firms,
37:15 um,
37:18 Declined its sales by 3.6%
37:22 relative to firms that were distance 5 or above.
37:27 What we can see is that then you can go for
37:30 only indirect customers,
37:32 so downstream 2 or customers of customers of customers,
37:35 downstream 3,
37:36 etc.
37:38 and we see that this,
37:40 um,
37:40 there was significant propagation,
37:42 indirect propagation going downstream
37:45 all the way to firms
37:47 being with 4 degrees away.
37:50 Going downstream from Disaster River.
37:53 On the right side of the graph,
37:54 we also see that there was significant upstream propagation.
37:57 So from disaster area firms to their suppliers,
38:00 upstream one,
38:02 they lost 3% following the earthquake.
38:04 The supplier suppliers upstream too,
38:07 they lost 2%.
38:09 And so on and so forth.
38:11 OK.
38:12 What we're seeing here is a large,
38:15 uh,
38:16 spread,
38:16 large contagion or cascading behavior of this localized supply chain disruption,
38:22 with effects decaying with network distance as models imply.
38:26 Uh,
38:26 and very briefly,
38:27 I'll note what,
38:28 what the red dots tell you is that there is no evidence of pre-trend.
38:32 So these firms did not differ
38:35 already,
38:35 uh,
38:36 in their performance
38:37 already,
38:38 uh,
38:38 previous to the earthquake.
38:39 Again,
38:40 the,
38:40 the paper has a bunch of uh
38:43 Of um
38:44 robustness checks,
38:45 I found interesting.
38:47 But now I just want to take a step back from this very micro level impact,
38:50 and the first thing I want to notice is
38:53 for any of you that has looked at firm level data and firm
38:56 level growth rates and the standard deviation of the firm level growth rate,
39:00 even the largest of these numbers,
39:02 3.6% negative,
39:04 is not very large.
39:06 So it's not that the effects for individual firms are large.
39:11 Rather,
39:11 the effects may matter in the aggregate because it affects many firms.
39:17 So,
39:17 if you recall,
39:18 I told you that about 50% of the firms in Japan are within two steps
39:22 of the,
39:23 of disaster area firms.
39:25 And what we can see from the previous thing
39:27 is that though there are again relatively small effects,
39:30 right,
39:31 you lost 2%,
39:32 3% at most.
39:35 Um,
39:35 the fact there are,
39:36 you are shaving off 2% for 50% of the first.
39:41 And that's what generates an aggregate shock,
39:43 an aggregate movement,
39:45 right.
39:46 It's the fact
39:48 that there is some propagation
39:50 and the fact that the production network or the supply chain is wired in such a way.
39:55 That everyone,
39:57 if you hit a note at random,
39:59 a production site at random,
40:01 A lot of people are within a neighborhood of two
40:05 supply chain linkages of the way,
40:07 right?
40:07 This small world aspect of supply chain.
40:11 You could also do formal aggregation in GE rather than back
40:14 of the envelope aggregation of the effects I told you,
40:16 and what we're getting now,
40:18 and we're just computing these as we speak,
40:20 is that Japan lost about 0.5% of GDP that year due to the Earth.
40:25 And that's not very large,
40:27 but it's in line with empirical studies on the effects of disasters.
40:32 It's a threefold magnification from the simple adding up,
40:35 you know,
40:36 traditional accounting adding up of,
40:38 of,
40:38 uh,
40:39 say,
40:39 the real value-added weight of this.
40:42 And it's significant for Japan because you need to keep in mind that
40:45 Japan's growth rate average growth rate for the last 20 years is 1%.
40:50 OK
40:52 Now having,
40:53 so this kind of gives you
40:55 that these effects are real,
40:57 they matter,
40:58 and they show up in aggregate.
41:00 And
41:01 they are arguably causal
41:03 at a large scale of looking at a million firms,
41:07 um,
41:07 firms that are apparently very distant,
41:10 both geographically in a sector manner,
41:13 and even in supply chain distance still get affected.
41:18 Now I want to just take you a little
41:20 bit over the discussion that happened in Japan and,
41:22 and,
41:23 and mostly around business schools,
41:25 much less in economics,
41:27 after the Japan earthquake,
41:29 and thinking about what our strategies to deal with.
41:33 Uh,
41:33 this kind of supply chain risk.
41:36 Um,
41:36 and,
41:37 and this is my own summary.
41:38 Um.
41:40 But the way I see this,
41:42 there,
41:42 you know,
41:42 at the firm level,
41:44 I'm thinking first about private sector
41:46 responses on recognizing supply chain risk,
41:49 there was the obvious,
41:50 the obvious first thing that came about was the lack of information.
41:54 Even very large firms claim
41:57 that they lost weeks understanding what to
41:59 do with production because they didn't even know
42:02 why their input supply stopped.
42:07 Uh,
42:08 their direct supplier was up and running.
42:11 So they did not even know indirectly who are,
42:14 who they linked,
42:15 and this led to a number of programs led in particular by big firms in Japan
42:20 of recursively imposing on suppliers.
42:23 So if you want to be my supplier,
42:25 you need to tell me who your suppliers are
42:27 and the suppliers of the suppliers and the suppliers of the suppliers.
42:31 Large automakers in Japan try to,
42:35 try to get
42:35 gather information on supply chain.
42:39 The second
42:41 that I think Richard already kind of
42:42 briefly touched upon is diversification strategy.
42:45 This is both of suppliers
42:47 and about production sites,
42:49 my own production sites,
42:50 or demanding,
42:51 for example,
42:52 that my supplier,
42:53 if I have one,
42:54 has two distinct productions.
42:57 This is to ensure short term substitution possibilities.
43:00 And the other,
43:02 um,
43:02 which
43:03 I think we have very little evidence whether it's gone on in practice or not,
43:07 but there was a discussion certainly about this,
43:09 is the standardization of components,
43:11 in particular,
43:12 you know,
43:12 one reason why the supply chains are fragile is the myriad of
43:17 um
43:18 You know,
43:19 customer demand
43:21 spec of I want my
43:23 headlights to have a shade of blue or a shade of pink,
43:27 and I want the color of the car to be this and not that.
43:31 And that
43:32 um
43:33 variety
43:34 that is,
43:35 we think is welfare enhancing,
43:37 um,
43:37 is,
43:38 brings about complexity and generates over ring type
43:41 problems where everything is ready to go,
43:43 but
43:44 the controller of the headlights is not available.
43:47 And that that that leads to a little bit of less choice,
43:50 less variety and standardization of components.
43:53 And one reason why these lessons may have
43:56 not been learned more globally and even without
44:00 outside of Japan is that these are very costly,
44:02 as I think Richard was already
44:04 Pointing out information is costly and introduces a gap between
44:08 large firms that can gather information and small firms that do not.
44:13 Diversification is costly,
44:14 in particular when you think about relation-specific investment.
44:18 If I now have two suppliers
44:20 for the same goods,
44:22 there is much less investment in a particular relation,
44:24 and it's costly to maintain all these relationships.
44:29 And,
44:29 and,
44:30 and the last point is standardization or less variety.
44:34 And either by direct higher cost or lower variety,
44:38 what this means is if you want to build robustness.
44:42 This comes at a cost.
44:43 How large is this cost?
44:44 It's not clear.
44:46 The way I think about it in the first instance is
44:49 firms who participation in the supply chain or
44:52 in this case in the global supply chain.
44:54 Um,
44:55 the marginal firms are affected,
44:57 firms that are basically,
44:59 uh,
45:00 You know,
45:02 Close to breaking even
45:03 if they face a larger cost,
45:05 they will either not implement these changes or remove themselves from.
45:18 Large firms it seems to be is uh what this presents for research,
45:23 so this area of production networks has expanded enormously over the last 10 years,
45:28 um.
45:30 And has definitely put on the table the notion that there are systemic
45:35 production firms or sectors.
45:38 Narrowly defined production.
45:40 Or actors
45:41 Roscoe and systemic risk.
45:44 Roscoe,
45:44 another couple of minutes please.
45:45 Thank you.
45:46 Yes,
45:46 I will,
45:47 I'll finish in one slide.
45:48 Um
45:51 And,
45:51 and
45:53 And while it considers that it makes it very clear,
45:55 it also considers that the supply chains are efficiently,
45:58 efficient,
45:59 and that there is no room for policy intervention.
46:01 So
46:02 friction on the environment is something that
46:03 is just now being looked at seriously.
46:06 The other thing that is not that there is not a lot of work on,
46:09 although there is some great work on supply chain formation.
46:13 But we understand
46:14 how firms rewire after firms exit.
46:18 Or how firms make these linkages to begin with,
46:21 and microdata being a large bug.
46:26 I'll finish
46:27 Uh,
46:28 for policymakers,
46:29 right,
46:29 in the situation where we are now,
46:31 where
46:32 we've identified clearly some key goods that we need to have them operational.
46:37 We need ventilators,
46:38 we need masks,
46:39 we need tests,
46:40 or here in the UK you also need flour.
46:42 Apparently flour has disappeared.
46:46 Now,
46:46 this presents incredible challenges to policymakers,
46:49 which is
46:50 Uh,
46:51 typically it's not because the ventilators cannot be done at the final producer.
46:56 This is because somewhere in the supply chain there's a
46:58 bottleneck that cannot be overcome in a very short run.
47:02 This requires the policymakers to know in real time
47:04 what are bottlenecks to the production of individual goods.
47:09 And then we would require them to act,
47:10 and it's not obvious again we've seen that we
47:12 don't have great theoretical justification for how to ring fence
47:17 what are deemed to be key supply chains or mandate diversification and capacity.
47:24 So this presents lots of challenges for policymakers in
47:26 the sense that information being one of them,
47:29 most countries in the world do not have this supply chain data
47:33 that allow them to trace
47:35 which are the key firms to ensure the production of certain goods.
47:40 Uh,
47:41 we do not know,
47:43 and when I talk to people exactly what we mean by ringfencing.
47:48 Uh,
47:48 and what we,
47:48 we don't know the,
47:49 the pernicious effect of though.
47:51 And we don't have a great justification to
47:54 enact
47:55 policy measures that ring fence or ensure
47:58 diversification or national champions as Richard.
48:03 So,
48:03 I'm happy to take further discussion in the Q&A.
48:07 Thank you very much.
48:09 Thank you very much,
48:10 uh,
48:11 Vasco.
48:11 That was
48:12 an excellent,
48:14 careful,
48:14 thoughtful presentation.
48:16 Um,
48:17 after Richard and Vasco,
48:20 uh,
48:20 who have talked about
48:23 God-made shocks
48:25 like disease and natural disasters,
48:28 we will now turn
48:29 to how
48:30 these
48:32 shocks are amplified by man-made shocks.
48:35 In the form of trade restrictions.
48:38 And for that we have Simon Enet,
48:40 who's a professor of international trade and economic development.
48:44 And MBA director at the University of St.
48:47 Gallen.
48:48 Simon
48:49 has specialized in
48:53 Understanding,
48:54 monitoring,
48:56 highlighting,
48:57 protection.
48:59 At the start of the global financial crisis,
49:01 Simon created the Global Trade Alert Initiative,
49:05 and this has become the leading independent
49:09 conscience.
49:11 of the world,
49:12 and he has highlighted
49:15 some of the most egregious
49:17 measures taken by governments and perhaps
49:21 Through shining his light on them
49:24 has disciplined them.
49:26 Simon,
49:26 apart from his many other accomplishments,
49:29 has been a staff member
49:32 of the World Bank not once but twice,
49:36 and he has been
49:38 a friend for many of us for a long,
49:41 long time.
49:42 So Simon,
49:43 over to you.
49:49 Thank you.
50:05 Uh,
50:06 in light of,
50:06 uh,
50:07 what we've known about
50:08 response to previous global economic shocks
50:11 and of course the academic literature and the research literature
50:15 which is
50:15 being prepared
50:17 now if I know how to
50:19 use this
50:20 technology,
50:21 I'll try and move through the slides.
50:23 Great.
50:23 I'm gonna start by discussing.
50:27 The uh
50:28 uh changes that we have seen this year compared to previous years,
50:32 uh,
50:32 in the overall global stance
50:34 now
50:35 monitoring trade policies is at some level quite an unsatisfactory,
50:39 um,
50:39 uh,
50:40 endeavor
50:41 because there are.
50:42 There are lots of ways that theory would tell us we would like to be able to
50:45 assess policy stance,
50:47 ideally in terms of some ad valorum equivalent or the
50:50 amount of trade which is covered and the like,
50:52 but much of the information we need to be
50:54 able to make those types of meaningful calculations.
50:59 Isn't available at the time when we,
51:01 when we need to be able to assess what governments are doing.
51:04 So
51:05 what has tended to happen is
51:06 that there have been
51:09 essentially
51:10 statistics based on counts of measures
51:12 that governments have undertaken.
51:14 Now
51:14 we can all think of
51:16 many reasons why those counts are potentially misleading.
51:19 Wrong,
51:20 and I will show you some export coverage data later,
51:24 but this is the metric which is typically
51:27 used by the international organizations
51:29 and one which my team at Stallen has been using
51:34 for 10 years now.
51:36 So I should say in our
51:37 work and reports we do try and report other measures too,
51:41 but if you want to know what's happening
51:43 right now,
51:44 then I'm afraid you're stuck with counts.
51:46 That's pretty much the only game in town.
51:49 Well,
51:49 this first slide that I have here
51:51 on
51:52 developments this year shows just how different this
51:55 year has been compared to previous years.
51:57 If you look back over
51:59 the years 2009 to 2019
52:01 and you look at the measures which were put in place from 1 January
52:05 to 17 April,
52:07 you'll see
52:08 that your traditional anti-dumping,
52:10 anti-subsidy
52:11 and other measures
52:12 are
52:14 are the
52:16 largest share.
52:17 They're typically half of shares,
52:19 half of the measures which are
52:21 found in the first few months of the year,
52:23 and then
52:24 subsidy trade distorting subsidies are the second largest grouping
52:28 and then tariff increases.
52:29 Now of course,
52:30 there's variation across the years and
52:32 you can imagine
52:33 that this pie chart looked a little bit different during the trade war years.
52:37 If you're looking at some type of average to benchmark this year against,
52:42 then this is the starting point on the right hand side.
52:45 On the left hand side,
52:46 we have information on what governments have actually been doing this year,
52:50 and you can see that the most prevalent form
52:52 of
52:53 trade policy intervention,
52:54 which is harmful to foreign commercial interests
52:57 are measures which relate to exports,
52:59 in
53:00 particular,
53:01 measures which are restricting exports,
53:03 about which I'll say more later.
53:05 So we have a shift in the
53:08 mix of
53:09 trade distortions that we've observed.
53:11 We can see also that there's a far greater number,
53:14 proportionally
53:15 of
53:16 subsidies which have been introduced.
53:18 These are subsidies to
53:19 import competing firms,
53:21 and we can see that contingent protection measures,
53:23 which were the largest category in previous years,
53:26 is now the 3rd largest category this year.
53:28 So there does appear to be a trade policy shift
53:32 which has occurred,
53:33 and of course,
53:33 one question we will want to
53:35 look into over time
53:36 and in research is whether that
53:39 had meaningful effects.
53:41 Now these summary statistics here
53:44 are drawn from
53:45 an initiative where we essentially monitor
53:48 very carefully what is published on government websites
53:51 around the world and systematically collect that information.
53:55 If you're talking about the overall quantum of the number of
53:58 Harmful measures introduced this year.
54:00 It is below what it was last year.
54:02 This year we've had 173
54:04 harmful measures which have been documented,
54:06 which is below the 315 found last year,
54:10 but still
54:10 this year's numbers are relatively high compared to the 10 year average.
54:15 So this is the first finding.
54:18 Now if we can go to the next slide.
54:27 Actually,
54:28 if I could have some help getting to the next slide,
54:29 I'm find
54:30 finding this rather difficult,
54:32 um.
54:35 Ryan or
54:37 Ale?
54:40 Um,
54:41 Ryan?
54:44 Hi Simon,
54:45 if you just,
54:45 have you tried
54:46 pressing a
54:48 The director,
54:49 the arrow key,
54:50 did that
54:51 do it?
54:52 Uh,
54:52 no.
54:54 Or pressing down
54:55 might also.
54:58 Hold on.
55:00 Neither of those.
55:03 Neither of those works,
55:04 so can I just ask you to?
55:05 Sure,
55:05 sure,
55:06 I'm happy to.
55:07 That's great.
55:07 So if we can go to the next slide,
55:08 I can
55:09 try and catch up a little bit.
55:10 Um,
55:10 the next slide slips,
55:13 looks at
55:14 not the trade harm,
55:15 the harmful measures which have been put in place,
55:17 but the liberalizing measures,
55:19 and hopefully Ryan will bring this up in a second.
55:21 And
55:22 here we have
55:23 a different pattern which is
55:26 worth
55:27 noting.
55:28 Normally between 2009 and 2019,
55:31 by now we have seen about
55:34 half of the measures which are introduced,
55:35 liberalizing measures are
55:37 tariff measures.
55:39 About 45% of them,
55:40 and this proportion has increased a lot,
55:44 or at least somewhat I should say,
55:45 and then
55:46 we also tend to see
55:48 another breakdown in terms of
55:50 export related measures,
55:51 normally the removal of export restraints,
55:54 but the other big change if you compare.
55:56 In 2020
55:56 with earlier years,
55:58 is
55:59 that we see far fewer FDI liberalizing measures this year,
56:03 and this is actually consistent with the
56:05 work that we've seen on the ground,
56:07 where more governments,
56:08 if anything,
56:08 will be moving in
56:09 a more restrictive way in terms of how
56:11 they're treating foreign direct investment this year.
56:13 So overall,
56:14 then,
56:14 what we have seen
56:16 is
56:17 on the harmful side we see
56:19 more export
56:20 related measures,
56:21 and on the liberalizing side we see more tariff cuts and fewer FDI reforms.
56:26 So this is the overall picture
56:28 for this
56:30 the particular year,
56:31 looking globally across.
56:32 All of the
56:34 countries of the world.
56:35 Now what I'd like to do,
56:36 since the purpose is to focus a little bit on COVID-19,
56:39 is to look
56:40 at
56:41 the developments inside the sector,
56:44 which covers
56:45 medical equipment,
56:46 medical supplies
56:47 and medicines,
56:49 and if we could go to the next slide please,
56:50 that would be great.
56:52 Oh,
56:53 I should have added him.
56:55 Actually,
56:55 uh,
56:56 there's,
56:56 uh,
56:56 one other point I,
56:57 I'm,
56:58 now I can see I've
57:00 missed the order here.
57:01 Um,
57:02 before we get to the,
57:03 uh,
57:03 export incentives and the export restrictions in medicines and the like,
57:07 let me,
57:07 um,
57:08 Uh,
57:08 just quickly say one other thing,
57:10 um,
57:10 although,
57:11 uh,
57:12 we earlier mentioned the export restrictions being prominent,
57:15 if you're actually looking at trade affected,
57:17 the biggest issues this year to date
57:19 have actually been export incentives which have been offered by countries
57:23 to their exporters,
57:24 and so,
57:25 uh,
57:26 we have seen three such measures.
57:27 Perhaps the largest such measure
57:29 was one introduced by China on the 20th of March,
57:32 which,
57:33 where they offered higher export incentives through their.
57:36 The tax system
57:37 for
57:38 products which were
57:40 exported
57:41 from that country
57:42 and we can see that worldwide,
57:44 the total estimate so far
57:47 of the amount of trade affected by this
57:49 year's trade distortions is just under $0.5 trillion
57:52 and of that $3 trillion
57:54 comes from the export incentives
57:57 which the Chinese and others have offered.
58:00 And so you can see that a large share of the trade distortions in terms of.
58:03 Coverage
58:04 are actually
58:05 unrelated to the export restrictions,
58:07 which I'll spend a bit more time looking on.
58:09 Please be aware that
58:10 the vertical axis on this chart
58:12 is in log form.
58:14 I've also put in place the
58:16 share of trade which is affected by import tariffs,
58:18 and
58:20 these are the most salient forms of
58:22 import restrictions,
58:24 and you can see they pale
58:25 in significance to
58:27 the impact or sorry,
58:29 the scale of export incentives.
58:33 So then just to summarize the
58:34 what the the the changes so far this year,
58:37 we've seen lots of export restrictions
58:39 but
58:39 the uh policy which um which is probably affecting the most trade
58:43 are actually export incentives.
58:47 These export incentives,
58:48 by the way,
58:48 are typically very poorly documented by either the international organizations
58:54 or in economic research.
58:57 We now have more research on the VAT
58:59 rebates done by China.
59:01 There is a growing literature on this.
59:03 But
59:04 I should say
59:05 that these are
59:06 the principal form
59:08 of trade distortion that we saw
59:10 at the
59:11 turn of the last global economic crisis
59:14 and also it seems to be today as well.
59:17 If we can move on to the next slide,
59:19 please,
59:20 then I'll turn now to the export curbs with respect to
59:23 medical products,
59:25 medical products and medicines.
59:27 What
59:28 we have found this year,
59:29 and this has received,
59:30 of course,
59:30 a lot of attention,
59:31 given the salience of the pandemic,
59:34 is
59:34 that over time more and more countries have imposed
59:38 different types of Export curbs on medical products and medicines.
59:42 At the moment we are counting that 76 nations
59:45 have limited
59:47 the
59:48 exportation of different types of medical supply.
59:51 Not all countries have done it the same way and they have not all
59:54 covered the same degree of products,
59:56 but we do see
59:58 a spread over time
59:59 in the measures which have been taken.
1:00:02 And uh
1:00:03 this uh spread has occurred from
1:00:05 uh the source of the virus
1:00:08 uh westward over time as we have seen,
1:00:10 um,
1:00:10 as,
1:00:11 as the virus has spreads to the export
1:00:14 constraints have spread.
1:00:15 I should say that if you're thinking of a product like medical ventilators,
1:00:19 then
1:00:20 countries like Australia,
1:00:23 Mexico,
1:00:24 Canada and Japan are now pretty much the only suppliers which are available to send
1:00:30 these important pieces of medical equipment
1:00:32 to the rest of the world.
1:00:33 All of the other suppliers have some type of
1:00:36 export ban in place.
1:00:38 So this is,
1:00:39 I think,
1:00:39 a highly salient topic.
1:00:41 The literature,
1:00:41 as we know,
1:00:42 on export curbs
1:00:43 is uniformly damning.
1:00:45 A lot of the great work done on this
1:00:47 was done on the export curbs associated with the
1:00:49 agricultural
1:00:51 sector
1:00:52 and in particular with respect to the 2008 to 2010 commodity price spike.
1:00:58 Much of it was done actually at the World Bank and by World Bank affiliates,
1:01:02 and of course,
1:01:02 what makes that research so much easier is that we have
1:01:05 World prices
1:01:06 or national market prices
1:01:07 for these commodities.
1:01:09 Most medical products,
1:01:10 on the other hand,
1:01:11 are so differentiated
1:01:12 that I suspect we will not have the same,
1:01:15 we will not have the same rich data sets to be able to do
1:01:18 that type of research.
1:01:19 We'll probably have
1:01:20 more narrower focused research,
1:01:22 but still,
1:01:23 the work that has been done by Michele Rutte and colleagues
1:01:26 has shown that
1:01:27 the estimated impact on prices
1:01:29 of these export curves can be significant.
1:01:32 If we could move on to the next slide,
1:01:34 please,
1:01:35 I'd like to
1:01:36 talk about something which has received a lot of attention as well,
1:01:39 which are export curbs on food.
1:01:42 These have begun to receive a lot of attention from policymakers
1:01:46 and the press.
1:01:47 I think it's important to point out,
1:01:48 clearly this is a
1:01:50 potentially important matter which needs to be tracked over time,
1:01:53 but we are nowhere near the same
1:01:55 degree of
1:01:57 prevalence of export curbs in food that we have seen
1:01:59 in medical products,
1:02:01 really in terms of the major producers.
1:02:03 There's only one which has curbed
1:02:05 exports in a serious way,
1:02:07 and that's Russia
1:02:08 so far.
1:02:08 This may change,
1:02:09 of course,
1:02:10 and we'll have to watch it carefully,
1:02:12 but we do see
1:02:14 a much more limited exportation curbs
1:02:17 in this particular sector,
1:02:19 and we have also seen tracking countries here
1:02:22 that the political economy of these curbs are quite interesting.
1:02:24 Interesting.
1:02:25 A number of countries put in place curbs
1:02:27 which are extremely aggressive,
1:02:28 often bans,
1:02:31 and then they face pressure from farmers and from internal
1:02:35 forces within a government
1:02:37 which end up
1:02:39 causing the export measure to be relaxed.
1:02:42 We've seen a bit of that
1:02:43 going on in countries like Vietnam and Romania.
1:02:46 So there's a very interesting political economy story unfolding there as well.
1:02:50 So this is with respect to food.
1:02:53 Yep,
1:02:53 and let me just,
1:02:54 uh,
1:02:54 quickly,
1:02:55 uh,
1:02:55 move on to the next slide,
1:02:56 and I think that's,
1:02:57 uh,
1:02:58 my takeaway slide,
1:02:59 so I can wrap up.
1:03:01 So what are the main takeaways?
1:03:02 The first is we do see
1:03:05 some differences between this year and previous years.
1:03:10 The previous decade
1:03:11 saw a much more
1:03:13 traditional pattern of trade policy intervention
1:03:15 with resort to subsidies and to contingent protectionism.
1:03:19 This year we see
1:03:21 a lot of subsidies,
1:03:22 but we also see
1:03:23 many more export restraints.
1:03:26 It's unclear
1:03:29 what will happen next,
1:03:30 but if we do have a
1:03:31 collapse in world trade,
1:03:32 which has been forecast,
1:03:34 then there will be concerns that this may lead to
1:03:37 an unleashing of lots of protectionism,
1:03:39 and we'll have to track that
1:03:41 very carefully and see what format protectionism will take.
1:03:45 When it comes to the impact on development sensitive areas,
1:03:48 clearly
1:03:49 the impact on medical supplies and food are a concern,
1:03:52 but we see many more
1:03:54 impacts and many more effects in the
1:03:57 more
1:03:58 interventions in the area of medical supplies and food
1:04:01 at the moment.
1:04:02 And the last point I would make is that
1:04:04 whilst export restrictions get a lot of attention,
1:04:07 Uh,
1:04:07 from the media and the like,
1:04:09 and they are very,
1:04:10 very salient for good reasons,
1:04:12 uh,
1:04:12 often the most pervasive,
1:04:14 uh,
1:04:15 forms of protectionism or trade distortion are actually
1:04:19 measures to increase exports,
1:04:21 uh,
1:04:21 which governments take,
1:04:22 which amount to
1:04:24 market share stealing or market share reallocation
1:04:27 away from trading partners towards subsidized firms,
1:04:31 and this is what we should be watching as well
1:04:33 as this
1:04:34 COVID pandemic unfolds.
1:04:36 Thank you very much.
1:04:39 Thank you,
1:04:40 Simon,
1:04:40 for that important and um
1:04:43 careful presentation.
1:04:45 Um,
1:04:48 Before we open up for questions,
1:04:49 I,
1:04:50 uh,
1:04:51 have been asked to take stock and make
1:04:54 some,
1:04:54 um,
1:04:55 uh,
1:04:56 observations.
1:04:58 The 3 questions we could ask at this juncture.
1:05:02 The first is how
1:05:04 production interdependence
1:05:07 leads to a transmission of shocks.
1:05:11 And that's a question that Richard and Vasco
1:05:14 addressed.
1:05:16 The second question is how the shock
1:05:20 that we are witnessing today is leading to
1:05:22 a disruption of global value chains today.
1:05:26 And the third question is
1:05:28 how
1:05:29 does the shock
1:05:30 and
1:05:32 the
1:05:33 Policy response to the shock
1:05:35 lead to a longer term reorganization of global value chains.
1:05:43 I think in the discussion now it would be great
1:05:45 to hear from all of you and from the speakers
1:05:48 is how is this shock different,
1:05:50 you know,
1:05:50 as I said at the beginning,
1:05:51 this is not like the Japanese tsunami,
1:05:53 it's not like the American mortgage crisis.
1:05:56 It is a shock that's virtually simultaneously affecting
1:06:00 many different parts of the world.
1:06:02 It's also unusual
1:06:03 that it's not so much
1:06:05 a shock that is disrupting manufacturing capacity.
1:06:10 Some of the biggest changes we are seeing are in the Services sectors.
1:06:15 So what does that mean when a large part of the economy,
1:06:19 which is relatively less dependent on backward value chains,
1:06:23 but perhaps more important in forward value chains?
1:06:26 How should we think in terms of the nature of this shock as different
1:06:31 and how do we expect it to be transmitted differently
1:06:35 when we think of the disruption,
1:06:38 how is it the types of stages of production that are being affected?
1:06:42 Uh,
1:06:44 you know,
1:06:44 high density collective consumption,
1:06:47 high density collective production,
1:06:50 face to face interaction,
1:06:52 how do we expect those
1:06:54 also to lead to a disruption in global value chains?
1:06:58 And then when we think about the longer term,
1:07:00 I think Vasco very nicely set out
1:07:02 the kind of issues,
1:07:04 but there are,
1:07:04 as he said,
1:07:05 a tension between acquiring more information and the big benefits of delegation.
1:07:11 The
1:07:13 Tension between diversifying
1:07:17 versus efficiency precisely because of relationship specific
1:07:22 investments that these global value chains require
1:07:25 and also a tension between the standardization
1:07:28 which would make it easy for you to switch
1:07:31 and the customization,
1:07:33 which is increasingly a feature of these international production relations
1:07:37 and I do think that even though Simon said it's not just export restrictions.
1:07:41 The fact that we are seeing export restrictions,
1:07:44 I think,
1:07:45 could have a particularly pernicious effect,
1:07:47 not just on global value chains,
1:07:49 but on trade.
1:07:50 Because
1:07:51 they
1:07:52 suggest
1:07:53 that trade is not
1:07:55 a reliable way of dealing with difficult circumstances.
1:08:00 The whole idea
1:08:02 that
1:08:02 if in bad states of the world countries
1:08:05 import export restrictions on food and medical supplies,
1:08:08 then it does legitimize
1:08:12 greater
1:08:13 self-sufficiency even in good states of the world.
1:08:17 So how should we think about that and especially
1:08:20 When all our existing multilateral institutions,
1:08:24 even in their
1:08:25 prime,
1:08:26 they were not able to discipline
1:08:28 these export restrictions.
1:08:30 How is it
1:08:31 that we can think of new international cooperative
1:08:34 arrangements which preserve the huge gains from this
1:08:39 fragmentation of production,
1:08:41 both
1:08:42 in terms of productivity and diffusion of technology?
1:08:45 So I think these are the questions that are important,
1:08:49 you know,
1:08:49 we We are doing some work here with Caroline Freund,
1:08:52 Daria Taglioni,
1:08:53 Michele Ruta
1:08:55 to try and understand
1:08:56 both the short run and the long run implications of these changes,
1:09:00 but it'd be great to hear now from all of you.
1:09:02 But let me open up
1:09:04 for questions.
1:09:06 Uh,
1:09:07 one,
1:09:09 you can use YouTube,
1:09:11 you can use Webex,
1:09:13 and let me begin with one question
1:09:16 which has come from Sabrina Holman.
1:09:19 Who's asking
1:09:21 Any thoughts,
1:09:22 views on the impact of these disruptions on
1:09:26 current
1:09:27 and potential trade agreements,
1:09:29 regional agreements.
1:09:31 So perhaps Richard,
1:09:32 you can begin and then the others can step in if they wish.
1:09:36 And after that I'll try and keep track of questions and bundle them together.
1:09:49 Richard,
1:09:49 are you there?
1:09:50 Let me,
1:09:51 oh,
1:09:51 sorry,
1:09:51 yeah,
1:09:51 I,
1:09:52 I hadn't thought about it,
1:09:53 but let,
1:09:53 let me,
1:09:54 um,
1:09:55 talk about,
1:09:56 uh,
1:09:57 what I think happened,
1:09:58 for example,
1:09:58 in NAFTA.
1:09:59 I think there's been less disruption
1:10:01 and less export restrictions within NAFTA,
1:10:04 and I think within Europe also there was an initial push
1:10:08 that blocked a lot of stuff within Europe,
1:10:10 but then
1:10:11 it was
1:10:12 allowed inside Europe but not outside Europe.
1:10:15 So I think to my,
1:10:16 my,
1:10:16 my instinct would be that what we're going to,
1:10:19 especially in things like medical,
1:10:20 is uh
1:10:21 part of the regional
1:10:23 agreements would be not to block each other's exports
1:10:26 in this,
1:10:27 in this time of need.
1:10:28 So I think it may,
1:10:29 may come to have clubs of people who
1:10:32 promise not to block food or
1:10:34 medical equipment or something like that.
1:10:37 But I have to say that I,
1:10:39 I think a lot of,
1:10:40 of this will,
1:10:41 apart from the medi medical equipment,
1:10:42 I think a lot of it
1:10:44 will pass and
1:10:45 the vast majority of trade in
1:10:47 chemicals and vehicles and electrical equipment and
1:10:51 electronics,
1:10:52 that stuff is not going to be changed because the,
1:10:54 the basic political economy which led to the
1:10:57 diverse,
1:10:58 led to the global value chains isn't gonna change.
1:11:01 And uh so I,
1:11:01 I think a lot of this discussion,
1:11:03 oh my goodness,
1:11:04 now we have to change everything,
1:11:05 it's not gonna go through.
1:11:07 But anyways,
1:11:07 uh,
1:11:08 with respect to regional,
1:11:09 um,
1:11:10 regional trade agreements,
1:11:11 I think that's where
1:11:12 the restrictions have sort of been not,
1:11:14 not extended.
1:11:15 So,
1:11:15 NAFTA is trading with each other.
1:11:18 Europe is trading with each other.
1:11:19 I don't know the situation in Asia.
1:11:25 Simon,
1:11:25 do you want to come in on that?
1:11:29 Yes,
1:11:30 if I may,
1:11:33 I think there's a very interesting question which is raised by
1:11:35 your question and Sabine's question about
1:11:37 reciprocity,
1:11:39 because we typically think that if a country
1:11:41 liberalizes
1:11:42 its import regime in the context of a trade agreement,
1:11:45 the payoff
1:11:46 comes not only to the benefits from importing
1:11:49 but also for the benefit of.
1:11:50 Benefits from uh
1:11:51 being able to export to other countries,
1:11:53 but that calculus gets disturbed
1:11:56 in if uh
1:11:57 uh it turns out that if you liberalize your import tariffs let in
1:12:01 um imports,
1:12:02 but if there are no imports to buy because the other party is using export curbs,
1:12:06 especially on essential
1:12:08 items when you need them,
1:12:09 then a lot of the benefits.
1:12:10 From liberalizing your trade your trade regime
1:12:13 has been um
1:12:14 has been lost and so
1:12:16 what we
1:12:16 learned then is that the benefits from reciprocal
1:12:20 trade reform I think are conditional
1:12:22 on being able to somehow qualify or limit the use of these
1:12:26 um export restrictions and here I just build a bridge to what Richard said.
1:12:30 There are a number of
1:12:31 Trade agreements which have tried
1:12:33 to qualify the use of export restrictions.
1:12:36 Interestingly,
1:12:37 some are in North America
1:12:38 and some involve North American countries with Latin American countries,
1:12:43 and it would be interesting to see
1:12:45 if the disruption to supply chains there or the threat of such disruption
1:12:49 was perceived as being much lower as a result.
1:12:53 Thank you,
1:12:54 Simon.
1:12:55 Uh,
1:12:56 Vasco,
1:12:57 2 questions for you.
1:12:59 Uh,
1:13:00 one from Tristan Reed.
1:13:04 and I think from key,
1:13:05 uh,
1:13:06 is what are the most promising proposals to enhance the resilience of
1:13:13 supply chains and especially medical supply chains?
1:13:17 And here,
1:13:18 if I can paraphrase the question,
1:13:20 that would it be a good idea if firms don't do it themselves
1:13:25 to mandate greater holding of inventories or essential inputs.
1:13:38 Rasco,
1:13:39 you have to unmute yourself.
1:13:40 Here we go.
1:13:41 Sorry.
1:13:42 Um,
1:13:43 thanks a lot for the question.
1:13:44 I think that they're good questions.
1:13:45 Um,
1:13:50 I think what I am seeing from many conversations and from,
1:13:55 uh,
1:13:56 I think Richard was discussing this ahead of the,
1:13:58 you know,
1:13:59 you know,
1:13:59 we had a small pre-discussion to the conference.
1:14:03 I think the most promising thing or the most
1:14:06 positive thing that I'm seeing around the world.
1:14:09 Is the deployment of massive data.
1:14:14 That is on the information side if you want,
1:14:16 what we were discussing before.
1:14:19 So
1:14:19 even countries where an understanding of
1:14:23 what is your supply chain,
1:14:25 what are your supply chain risks,
1:14:26 can you map out the supply chain with some detail?
1:14:30 Can you tell me
1:14:31 what you require to produce ventilators?
1:14:34 Which was a kind of a pipe dream,
1:14:37 I would say 20 years ago.
1:14:39 Um,
1:14:39 if you asked me 2 years ago to do that on a systematic level,
1:14:44 uh,
1:14:44 at least nation by nation,
1:14:46 it was still a pipe dream.
1:14:48 And what I am seeing in this crisis is governments and policymakers and advisors.
1:14:55 Um,
1:14:55 actually implementing very interesting solutions with payment system data,
1:15:00 with transaction data,
1:15:02 with banking data.
1:15:04 Which
1:15:04 is a first step in understanding
1:15:08 what can be done
1:15:09 in understanding what are bottlenecks in your supply chains.
1:15:14 Where do you stand as a macro risk if
1:15:17 plant X goes down?
1:15:19 A couple of years ago,
1:15:20 we had this,
1:15:21 we found,
1:15:22 or I found out that half of the world's
1:15:25 hard disk drives were
1:15:26 were produced in a valley in Thailand,
1:15:28 I believe.
1:15:29 And then the valley flooded and the hard disk drives
1:15:32 price doubled overnight.
1:15:35 Um,
1:15:36 I think we,
1:15:36 we are now living in a world where
1:15:40 we actually are,
1:15:41 are,
1:15:42 are much more aware and can map out these risks,
1:15:44 and that's the first step for policy.
1:15:47 Um,
1:15:48 so I think that's been an exciting development.
1:15:50 Now,
1:15:50 to go from there to what can actually governments do in terms of shoring up
1:15:56 our,
1:15:56 in terms of,
1:15:58 um,
1:15:59 ring-fencing,
1:16:03 what national interest,
1:16:04 entire supply chains around national interest,
1:16:07 uh,
1:16:08 is something that I think we are far from.
1:16:12 I,
1:16:12 I,
1:16:12 I,
1:16:12 I suspect that the most we will be able to deliver the capacity
1:16:16 to start thinking about these things.
1:16:18 Um,
1:16:19 I can tell you that,
1:16:20 you know,
1:16:20 the UK and Spain and
1:16:22 different governments that I've been in touch are
1:16:25 thinking very,
1:16:26 thinking ambitiously about these things,
1:16:28 trying to figure out.
1:16:30 Exactly what do I need to
1:16:33 ensure
1:16:35 production at the current level of demand of good X?
1:16:40 And I think we are very close to having,
1:16:42 you know,
1:16:42 relatively scientific answers to those questions,
1:16:46 but I think politically,
1:16:47 in terms of the political economy of that
1:16:49 and all that can go wrong
1:16:51 along that path,
1:16:53 something that we haven't thought,
1:16:54 and that I would like to think a little bit more,
1:16:57 um.
1:16:59 I,
1:16:59 I know that this is,
1:17:00 is,
1:17:00 is for sure happened with,
1:17:01 with medical supplies to go back to you.
1:17:06 Uh,
1:17:07 Bhaskar,
1:17:07 thank you.
1:17:08 There's another question,
1:17:09 if again I can paraphrase from Hatem
1:17:12 Jabani.
1:17:14 And
1:17:15 he is or she is asking about,
1:17:19 you know,
1:17:19 are you taking into account
1:17:22 How,
1:17:23 how are you thinking about switching costs on both the consumer
1:17:27 and the supply side when you think of disasters,
1:17:30 number one?
1:17:31 And secondly,
1:17:33 uh,
1:17:34 is it not possible that
1:17:37 the disasters or value chains transmit shocks,
1:17:41 they're also able to diffuse shocks.
1:17:43 Isn't
1:17:44 a global value chain itself.
1:17:47 An exercise in diversification,
1:17:50 even though,
1:17:51 you know,
1:17:51 the Leonti,
1:17:52 the interdependence makes you vulnerable to these oing kind of
1:17:57 problems,
1:17:59 should we not see that as a positive side of global value chains?
1:18:05 Um,
1:18:11 Right.
1:18:12 So,
1:18:14 Let me give you an experience of.
1:18:17 Of an automaker in Japan.
1:18:20 Uh,
1:18:21 so they turned out that after a few days,
1:18:23 they realized
1:18:25 they didn't have a pigment to paint the car,
1:18:27 to,
1:18:27 to form the paint to paint the car.
1:18:30 And that's a relation to specific investment because the automaker makes
1:18:34 um
1:18:37 Guarantees to the final consumer
1:18:39 on,
1:18:40 you know,
1:18:41 10 years,
1:18:42 or we'll paint the car for free for you,
1:18:44 that there will be no rust.
1:18:45 And
1:18:47 again,
1:18:47 it goes according to spec of what is the color,
1:18:50 exact color that the consumer asked for.
1:18:53 Um,
1:18:56 That,
1:18:57 so switching was so costly in that sense,
1:18:59 and it was not that,
1:19:00 it was not their car point in the world,
1:19:02 is that,
1:19:02 that's a relation specific investment for the automaker.
1:19:05 That the automaker shut down his plant and send
1:19:08 his workers to rebuild the plant of the supplier.
1:19:13 Um
1:19:15 Now,
1:19:15 what that tells you is that switching costs are high,
1:19:18 and even if you
1:19:20 plan for them ex ante,
1:19:22 um,
1:19:22 you can have,
1:19:24 uh,
1:19:25 you,
1:19:25 you can destroy the value of relation-specific investment.
1:19:29 Remember,
1:19:29 what we think by relation-specific investment
1:19:32 is,
1:19:32 uh,
1:19:33 that I share,
1:19:35 there is some kind of private good.
1:19:37 That is that I cannot go just to the market on
1:19:39 eBay and find another supplier is I create value with you,
1:19:43 my supplier,
1:19:45 and we share the rent of that.
1:19:47 This uh the notion of bilateral monopoly over that.
1:19:51 When you destroy that bilateral monopoly,
1:19:53 say,
1:19:53 by diversifying,
1:19:54 say by maintaining multiple suppliers at the same time.
1:19:59 You destroy the incentives
1:20:00 that underlie these
1:20:02 relations and therefore you destroy value.
1:20:06 That's number.
1:20:07 So the other thing that you've talked about,
1:20:09 and this will follow all the way to
1:20:12 final consumers and costs.
1:20:14 That are probably passed through to the final.
1:20:18 Um,
1:20:20 The other thing you said is,
1:20:21 can they be um
1:20:23 Are we speaking too negatively as supply chains as
1:20:27 just a conductor of shocks rather than dissipator of shocks?
1:20:30 And of course you're right.
1:20:33 There is just kind of
1:20:34 the clean thinking there is
1:20:36 what is the counterfactual.
1:20:37 The counterfactual is
1:20:39 uh to have everything,
1:20:41 all my eggs in one basket and have all my production
1:20:45 in,
1:20:45 in my country.
1:20:47 Well,
1:20:47 that is,
1:20:48 first,
1:20:49 you're gonna lose
1:20:50 large efficiency gains
1:20:52 from comparative advantage reasons.
1:20:54 You're also gonna be
1:20:56 subjecting yourself to much more risk.
1:20:58 So it depends on what your counterfactual is.
1:21:00 Yes,
1:21:01 supply chain,
1:21:02 global supply chains pose transmission of risk,
1:21:05 but your counterfactual,
1:21:07 I would venture to say,
1:21:09 is more risk and more concentrated precisely for the reasons you.
1:21:13 Bye.
1:21:15 I,
1:21:15 I do think this dissipation is there and that's
1:21:18 one of the reasons apart from efficiency reasons,
1:21:20 the dissipation of of risk relative to
1:21:24 having all the production concentrated domestically and all your supply chains
1:21:28 that
1:21:28 that leads us where we are now.
1:21:32 Thank you,
1:21:32 Vasco.
1:21:33 Richard,
1:21:34 uh,
1:21:34 you were
1:21:36 relatively sanguine about what might happen to global value chains,
1:21:40 but here's a question from
1:21:42 Roberto E Chandi.
1:21:43 Who says
1:21:44 that discussions on vulnerability of GVCs may suggest that,
1:21:50 you know,
1:21:51 the
1:21:51 private sector is not sufficiently sensitive
1:21:54 and it may justify protectionist policies
1:21:58 and at a time when developing countries had finally embraced trade
1:22:02 and benefited from GVCs,
1:22:04 what could we now be entering a slippery slope of
1:22:10 protectionism.
1:22:11 And a related question from
1:22:16 Someone else
1:22:17 is that
1:22:18 given that
1:22:20 countries are simplifying
1:22:23 their,
1:22:23 I think it's key,
1:22:27 their documentation and various things to help their traders,
1:22:31 could this be a time where we can actually,
1:22:34 you know,
1:22:36 aim for greater harmonization or simplification
1:22:39 and trade facilitation.
1:22:43 Good.
1:22:43 So,
1:22:43 let,
1:22:44 let me address those one at a time.
1:22:46 So,
1:22:46 I don't think we're heading towards an increase in protectionism here.
1:22:50 And I think uh Simon's chart showed that actually,
1:22:54 even the Trump administration is liberalizing tariffs because they realize
1:22:58 there are pluses to uh imports,
1:23:00 not just minuses to imports.
1:23:03 And what it is,
1:23:04 is we're worried about uh the slippery slope to
1:23:07 export protectionism.
1:23:09 So let,
1:23:09 let me just take that in,
1:23:11 in,
1:23:11 um,
1:23:11 in,
1:23:12 in two bits.
1:23:13 So one,
1:23:14 when the GAATT was set up and when the
1:23:18 European Union was set up.
1:23:20 Wartime
1:23:21 embargoes and shortages were front of mind of the people who wrote these documents.
1:23:25 So all this kind of export restrictions are legal
1:23:29 on purpose,
1:23:30 but it was never really an issue
1:23:32 in the in the post-war world because supply was never a constraint.
1:23:36 All you needed was demand,
1:23:38 so it was all about restraining people from protectionism,
1:23:42 not from export restraints.
1:23:44 So I suspect we may see a greater realization.
1:23:46 That we also need to coordinate on restraining export restraints as well,
1:23:51 especially in food,
1:23:52 and we've seen this movie when it comes to food before.
1:23:54 There's,
1:23:55 there's nothing new about export restrictions
1:23:57 tumbling out of control on food.
1:23:58 It happened in
1:23:59 2006,
1:24:00 2007,
1:24:01 or
1:24:02 2007,
1:24:02 2008,
1:24:03 I think was the last time it happened.
1:24:05 So maybe we'll have more um
1:24:08 liberalization on that.
1:24:09 In terms of the second question about the harmonization,
1:24:12 we have seen some incredible things where it's,
1:24:15 it's difficult for countries to import
1:24:17 masks that are made to a Chinese standard,
1:24:19 even though they're basically the same as an American standard,
1:24:22 but the American standard was put in there more or less as a protectionist device.
1:24:26 So,
1:24:26 this may again be one of these realizations that,
1:24:29 that it is a moment where we can reflect.
1:24:32 But I,
1:24:32 I would again say,
1:24:33 I think most of this reflection
1:24:35 is being spun out too generally.
1:24:36 So this is in
1:24:38 medical equipment and a few things which are
1:24:40 actually,
1:24:41 uh,
1:24:42 I think the total is 7% of world trade is in these things.
1:24:45 And the vast majority of all these lessons we're learning.
1:24:47 are not going to go on
1:24:49 to the big volume trade things
1:24:50 because there hasn't been a change.
1:24:52 But let me come back to something you asked before about stockholding.
1:24:56 And so in Switzerland,
1:24:57 we're very vulnerable to imported food,
1:25:00 imported medical,
1:25:01 imported all sorts of things.
1:25:03 And so the Swiss government mandates stockholding
1:25:06 that's done in cooperation
1:25:08 with the private sector and subsidized,
1:25:10 and that's the way we deal with the supply shocks.
1:25:13 And I think many countries in terms of food,
1:25:16 where this is realized,
1:25:17 have food stocks.
1:25:18 So I think one of the solutions instead of trying to bring all the production home,
1:25:22 you just have a buffer stock
1:25:24 and you cooperate with that and may have to subsidize that,
1:25:27 but I think that's a better solution.
1:25:28 And I think in the end,
1:25:29 when we're done with this,
1:25:31 people will say that these global value chains were actually a savior,
1:25:35 because,
1:25:35 for example,
1:25:36 China can now make masks when they couldn't before
1:25:38 and uh when the US industry is getting shut down,
1:25:41 they can import masks and tests and things like that.
1:25:44 So,
1:25:45 at the end of this whole story,
1:25:46 which is going to go on for another couple of years,
1:25:48 I think trade will be viewed as a massive insurance policy,
1:25:51 not as a source of,
1:25:52 of uh contagion.
1:25:56 Thank you,
1:25:57 Richard.
1:25:58 Simon,
1:25:59 a question for you.
1:26:01 Have you seen any evidence on how these different measures that you described,
1:26:06 both the export
1:26:09 promotion measures or import restriction measures,
1:26:12 are having an effect on actual trade patterns
1:26:15 or on global value chains?
1:26:19 Right,
1:26:19 so,
1:26:20 the,
1:26:20 the developments this year are just simply too,
1:26:22 um,
1:26:23 too fresh for us to have any
1:26:24 serious empirical work,
1:26:25 but as I said to you
1:26:27 earlier,
1:26:28 with respect to export restraints,
1:26:30 and we have a,
1:26:30 you know,
1:26:30 fantastic research which has been done on,
1:26:33 mainly on the agri-food sector.
1:26:35 which I think has a clear set of conclusions
1:26:38 about the counterproductive nature of these export restraints.
1:26:42 Whether that carries over to
1:26:44 medical goods we will see,
1:26:46 but I think the benchmark,
1:26:47 the starting point is
1:26:49 quite well established in the literature.
1:26:51 With respect to export promotion measures,
1:26:54 there is a growing literature
1:26:56 which uses quite sophisticated gravity equation work.
1:27:00 Which looks to see if export growth into
1:27:04 third markets is faster for countries
1:27:07 where the products have been
1:27:08 subsidized compared to other
1:27:10 non-subsidized
1:27:12 exports from other countries,
1:27:13 of course,
1:27:14 controlling for all other factors.
1:27:16 And we do find
1:27:17 indeed quite strong effects for
1:27:20 market share reallocation.
1:27:22 So I think
1:27:22 the finding is
1:27:24 that trade actually gets increased by
1:27:28 these export promotion measures,
1:27:29 but also gets redistributed
1:27:32 as well and reshuffled between
1:27:34 trading partners,
1:27:34 and of course this is where it becomes
1:27:36 both a trade policy and a resource misallocation
1:27:39 problem.
1:27:41 Let me just quickly summarize the last few questions that have come in.
1:27:45 One,
1:27:47 Could it be that the worst is behind us,
1:27:50 and now it's only a question of stimulating demand?
1:27:55 How does the oil price
1:27:57 decline figure in all your
1:28:00 um
1:28:00 assessments
1:28:02 and
1:28:04 Is there a chance
1:28:05 that China's early recovery
1:28:09 Is going to lead also to a
1:28:12 reassessment and a reconfiguration of value chains.
1:28:16 Now,
1:28:16 I'll leave it to each of you
1:28:18 to decide whether or which question to respond to,
1:28:22 uh,
1:28:23 but also please.
1:28:24 Take a minute to sum up and conclude
1:28:27 and what has been the most illuminating discussion.
1:28:32 So why don't we go
1:28:34 in the order of Simon,
1:28:35 you go first and then Bhaskar stays second and Richard concludes.
1:28:40 Thank you and thanks for organizing
1:28:43 this particular session.
1:28:44 Let me just address the question,
1:28:46 is the worst over?
1:28:47 I strongly doubt it.
1:28:49 I think
1:28:49 even if one does not believe that there will be
1:28:52 substantial across the board tariff increases,
1:28:54 I do think
1:28:55 we will have to watch very carefully
1:28:58 whether the
1:28:59 measures which governments are using to subsidize firms
1:29:02 go beyond just traditional liquidity support
1:29:05 into.
1:29:06 Something a little bit more
1:29:07 sinister in terms of
1:29:09 either bailing out failing firms
1:29:11 or
1:29:12 firms which were failing before the pandemic that way,
1:29:15 or
1:29:17 attempts to try and
1:29:18 foster new industries or indeed to repatriate
1:29:21 supply chains.
1:29:22 We're already seeing some of that.
1:29:25 Perhaps the most prominent example is
1:29:27 the Japanese government putting a lot of money on the table
1:29:29 for Japanese firms to repatriate.
1:29:31 Our supply chains out of China,
1:29:34 uh,
1:29:34 and so we'll have to watch that carefully to see
1:29:36 if,
1:29:37 uh,
1:29:37 the trade distortions of the future
1:29:40 are going to be anything like the ones we've seen in the past.
1:29:43 Thank you.
1:29:48 Thank you,
1:29:49 Simon Bosco.
1:29:52 So
1:29:54 To,
1:29:55 to try to wrap up,
1:29:56 um,
1:29:59 I think there was a question there regarding is the
1:30:02 worst over and are we gonna go out of this.
1:30:05 I think it's a little bit,
1:30:06 uh,
1:30:07 related to the,
1:30:08 your your own initial question of what is the nature of this shock and uh.
1:30:13 Mm
1:30:14 What is awaiting us both now versus the future,
1:30:17 right?
1:30:18 Um,
1:30:19 so I,
1:30:20 it's just a set of,
1:30:21 of,
1:30:22 of factors.
1:30:23 One is
1:30:24 related to the length of the shock and,
1:30:26 and how do we expect,
1:30:27 and I think that will play a role in the sense
1:30:30 that as Richard said early on,
1:30:32 there was not a productive capacity was not affected.
1:30:36 There was no fundamental reason why we should
1:30:39 uh rewire global value chains or domestic chains.
1:30:43 Um,
1:30:44 As
1:30:46 the
1:30:47 The recession lengthened in time,
1:30:50 when you start factoring in um inefficient exit.
1:30:54 We,
1:30:55 you know,
1:30:55 and and have this piles up over time,
1:30:57 there may be larger effects in the sense of
1:31:00 um
1:31:02 You know,
1:31:03 the macro effects
1:31:04 done by inefficient liquidation of large parts of the productive,
1:31:08 productive,
1:31:08 uh,
1:31:09 side of the economy.
1:31:10 And if we go along that scenario,
1:31:13 um,
1:31:14 and there may be very large effects on,
1:31:16 on,
1:31:16 on,
1:31:16 on trade and,
1:31:17 and,
1:31:17 and longer lasting.
1:31:19 Um,
1:31:21 so I think that,
1:31:21 that's kind of,
1:31:22 it shines a light on the,
1:31:23 the,
1:31:23 the,
1:31:24 the need for,
1:31:25 uh,
1:31:25 you know,
1:31:25 all of these national power uh
1:31:27 policies of liquidity and some,
1:31:30 some sort of,
1:31:30 of buy.
1:31:32 Um,
1:31:33 reassurance of of kind of reassurance of the production side maintained.
1:31:37 Um
1:31:39 There's really a trade-off between what is,
1:31:42 uh,
1:31:42 protectionism versus what is
1:31:45 protecting inefficient liquidation on a massive scale if this continues going on.
1:31:49 Second point is,
1:31:50 you know,
1:31:51 what is the effect generally on,
1:31:53 if,
1:31:54 if there is,
1:31:55 what I try to say is
1:31:57 there may be
1:31:59 uh
1:31:59 costly manners of robustifying a little bit,
1:32:03 be it
1:32:04 Uh,
1:32:04 and,
1:32:04 and I've kind of covered a few ways.
1:32:06 Richard has just said,
1:32:07 stockpiling in key areas would be another one.
1:32:09 I believe all of these are costly,
1:32:11 but I believe all of these are costly at the margin.
1:32:14 So they are,
1:32:15 they,
1:32:15 they will,
1:32:16 you know,
1:32:17 at worst,
1:32:18 diversity a variety will go down a little bit or costs will go up a bit.
1:32:24 Given
1:32:25 large benefits of global trade,
1:32:27 what you,
1:32:28 what economics,
1:32:29 what,
1:32:29 uh,
1:32:29 you know,
1:32:29 the last 20 years looking at Virginia firms would say is
1:32:34 it is
1:32:35 the,
1:32:35 the firms at the threshold of participating that we will see disappearing.
1:32:39 But as we know,
1:32:40 these firms,
1:32:41 the bulk of global trade is accounted by very large firms.
1:32:46 To the extent that the,
1:32:47 the surplus that those are generating
1:32:50 is only,
1:32:51 only somewhat affected by any robust,
1:32:55 there,
1:32:55 there is a path that is feasible in the sense of robustifying the system
1:33:00 without
1:33:01 large,
1:33:02 uh,
1:33:03 large costs and large disruption in global value chain.
1:33:06 And the final point that is a little bit too far away from here is,
1:33:10 of course,
1:33:10 this all depends on political economy.
1:33:13 So,
1:33:15 The way we manage this as societies and as citizens and,
1:33:18 uh,
1:33:18 and politicians,
1:33:20 um,
1:33:22 you know,
1:33:22 turns to protectionism,
1:33:23 turns to disintegration,
1:33:25 uh,
1:33:26 may,
1:33:26 may be on the cards,
1:33:28 but that is less,
1:33:29 and it's,
1:33:29 um,
1:33:30 so I think there's a lot about how,
1:33:32 uh,
1:33:34 Rather than just thinking about it as strict demand or supply side policies,
1:33:38 but as policies that
1:33:40 try to uh keep
1:33:42 Negative political economy developments from,
1:33:45 from,
1:33:46 from,
1:33:47 from gaining foothold,
1:33:49 which would indeed send the system into very nonlinear
1:33:52 dynamics where crisis can happen and disintegration can happen.
1:33:58 Thank you,
1:33:59 Vaskar Richard,
1:34:00 final minute.
1:34:02 OK,
1:34:02 let,
1:34:03 let me just uh talk a little bit about the oil,
1:34:05 uh,
1:34:06 and this wraps back also into your question about trade and services.
1:34:10 So the oil price normally would be a stimulus to
1:34:13 having bigger
1:34:15 global value chains and,
1:34:16 and,
1:34:16 uh,
1:34:17 and getting things getting cheaper.
1:34:18 But as um uh an e-book that Simon Evannett and I are editing on,
1:34:23 on export restrictions that'll come out on Monday or Tuesday next week.
1:34:27 We have a couple
1:34:29 chapters on
1:34:30 sea transportation and air cargo,
1:34:33 and it turns out that both of those have been heavily disrupted by
1:34:37 the COVID restrictions directly.
1:34:39 So the price of oil isn't what's holding up
1:34:42 or hindering global value chains from moving.
1:34:44 It's actually direct
1:34:46 disruptions to sea transportation
1:34:48 and air cargo.
1:34:50 So let me go
1:34:51 to my,
1:34:52 the wrap up point.
1:34:53 The,
1:34:53 the thing I would say is this movie is not even near the end.
1:34:56 We don't even know what this movie is really about yet,
1:34:59 and the lessons that people are trying to draw from the early,
1:35:02 early stages
1:35:03 of this,
1:35:03 I think,
1:35:04 are slippery,
1:35:06 and not very
1:35:07 useful.
1:35:08 So
1:35:09 at the end,
1:35:09 I think people will be very happy that they had access
1:35:13 to the manufacturing sectors of other countries and the food producing countries,
1:35:17 sectors of other countries.
1:35:19 And,
1:35:19 and really at the end of this,
1:35:20 what we may come out of it is that we may want to all agree.
1:35:24 Not to restrict things
1:35:25 except in disciplined ways.
1:35:27 So we may end up with rounds of limiting export restriction,
1:35:31 and that's sort of
1:35:32 a longer thing,
1:35:33 the longer takeaway is that
1:35:35 cascading
1:35:36 or spirals of export restrictions are as
1:35:39 dangerous as cascading or spiraling protectionism,
1:35:42 and therefore we need international negotiations to do that.
1:35:46 Now,
1:35:46 of course,
1:35:46 all that requires the administration in the United States to change,
1:35:49 so.
1:35:50 Um,
1:35:51 it,
1:35:51 it,
1:35:51 it will depend for what happens in November,
1:35:54 but if
1:35:55 the administration changes,
1:35:57 I think there is a real chance
1:35:59 that
1:35:59 we will have a moment of cooperation to reshape
1:36:02 things,
1:36:03 to remove some of the tariffs that were put up and remove some of the restrictions
1:36:07 in a kind of a big talk.
1:36:08 Maybe you could call it the Biden round or something,
1:36:11 in which case,
1:36:12 discussing export restrictions could be part of it.
1:36:14 So I'll,
1:36:15 I'll just stop there.
1:36:18 Thank you very much,
1:36:19 Richard,
1:36:20 Simon,
1:36:20 Vasco for a
1:36:22 deeply insightful and richly stimulating discussion.
1:36:26 Thank you.
1:36:26 Bye-bye.
- add-style
- lp-body-content