Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=

# Hegemonic Competition with Carrots and Sticks

Authors: Timothy Meyer, Nicolas Wesseler  
Discussant: Alberto Martin  
Video: https://www.youtube.com/watch?v=U_ssYRB6uPs&t=4715s


## Talk (01:18:35 – 01:49:12)

[01:18:35] Can you please take a seat?  
[01:18:38] Alright everyone, we are very excited  
[01:18:43] to get started with the second paper here  
[01:18:45] and Tim is going to tell us about hegemonic competition  
[01:18:48] with carrots and sticks.  
[01:18:49] You have 30 minutes.  
[01:18:51] Okay, perfect.  
[01:18:52] Thanks a lot for having me.  
[01:18:54] So I'm Tim, I'm still at Keel and Vaughn  
[01:18:56] and moving to Columbia and UT Austin.  
[01:18:58] This is joint work with Nicholas  
[01:19:00] who is sitting over there at UBC  
[01:19:02] and is going to be on the job market this year.  
[01:19:04] So this is a paper about hegemonic competition  
[01:19:07] and that's of course the topic that's very much on the rise  
[01:19:10] in the States in 2018  
[01:19:12] great power competition  
[01:19:14] has become the primary focus of US national security.  
[01:19:17] What's meant by that is very clear.  
[01:19:20] It means competition with China is not seen  
[01:19:22] as the main geopolitical adversary of the US.  
[01:19:26] As all of us know in the world there's not just China  
[01:19:29] and there's not just the US  
[01:19:30] but there's many other countries who are deciding  
[01:19:32] on their geopolitical alignment  
[01:19:34] and the alignment that these countries choose matters.  
[01:19:37] Matters for flows of international trade  
[01:19:40] for flows of international finance  
[01:19:42] and of course ultimately,  
[01:19:43] alignment matters for questions of national security  
[01:19:46] for example in times of a war.  
[01:19:49] So perhaps therefore both the US and China  
[01:19:51] have been trying to actively influence  
[01:19:53] the alignment of other countries  
[01:19:55] using economic care assistance.  
[01:19:58] In terms of care, some of you might know  
[01:20:00] that China has become one of the world's largest  
[01:20:02] foreign aid donors  
[01:20:03] and there's not considerable concern  
[01:20:05] about the policy influence that Chinese foreign aid  
[01:20:07] might have at all.  
[01:20:08] In terms of sticks on the other hand  
[01:20:10] we're now seeing sort of administrations  
[01:20:12] repeatedly threatening other countries  
[01:20:14] with sanctions, export controls,  
[01:20:16] other forms of economic coercion  
[01:20:18] for geopolitical purposes.  
[01:20:20] So these are the policy instruments  
[01:20:22] we want to study in this paper.  
[01:20:24] We're going to ask how effective these are  
[01:20:26] at achieving their geopolitical objectives  
[01:20:28] and how they interact with one another  
[01:20:30] when two hedge amounts compete.  
[01:20:32] Now these are very challenging questions to answer  
[01:20:35] and they're especially challenging to answer  
[01:20:37] as this competition is unfolding in real time.  
[01:20:40] So we're reviewing the papers,  
[01:20:41] we take a step back to get a price in theory  
[01:20:43] with price of evidence that's based on the Cold War  
[01:20:45] and then at the end of the talk  
[01:20:47] I'm going to come back to the competition  
[01:20:49] between the US and China.  
[01:20:51] So we answer these questions then in four steps.  
[01:20:54] So at the start I'll have a very simple model  
[01:20:56] in which two competing hedge amounts  
[01:20:58] use both direct payments,  
[01:20:59] that's going to be the carrot,  
[01:21:01] I think a four and eight,  
[01:21:02] but also economic threats of sanctions  
[01:21:04] that's going to be the stick  
[01:21:06] to influence other countries.  
[01:21:08] The key object in that model  
[01:21:10] is going to be a reaction function  
[01:21:12] that describes how these small countries  
[01:21:14] choose their alignment in response to the comparison sticks  
[01:21:17] and we'll look by the hegemons  
[01:21:19] and that's what we'll try to estimate in the data.  
[01:21:22] In the second part of the paper  
[01:21:24] we then bring data to estimate this reaction function.  
[01:21:27] So we first measure the carrot  
[01:21:29] at US and Soviet economic and military aid  
[01:21:32] provided during the Cold War.  
[01:21:34] We're then going to measure the stick  
[01:21:36] using a model of international trade  
[01:21:38] where it can pathway how much each hegemon  
[01:21:40] can pressure other countries.  
[01:21:42] And we're going to measure alignment as a standard  
[01:21:44] the same data as Ernest was showing you  
[01:21:46] using boating at the United Nations.  
[01:21:49] In the third part of the paper  
[01:21:51] we then estimate this reaction function  
[01:21:53] that's at the heart of the model  
[01:21:54] and we estimate the crown of this reaction function.  
[01:21:57] We do so using an instrumental variable function.  
[01:22:00] What we find here is that these  
[01:22:02] geoeconomic tools work,  
[01:22:04] but they're relatively expensive.  
[01:22:06] So for example, if you take a country  
[01:22:08] that's initially neutral between the Americans  
[01:22:10] and the Soviets and you double for an aid  
[01:22:12] to that country that costs $120 million  
[01:22:14] an average for the US during the Cold War  
[01:22:17] you only move that country by 2.5%  
[01:22:19] of a standard deviation that's only 1%  
[01:22:21] of the geospatial distance between that country  
[01:22:24] and the NATO member.  
[01:22:26] So you can't move these countries  
[01:22:28] it just ends up being very expensive.  
[01:22:30] Finally at the end of the talk  
[01:22:32] I'm going to come back to the modern times  
[01:22:34] and I'll speak a bit to the competition  
[01:22:36] between US and China and in particular  
[01:22:38] since we estimate this reaction function  
[01:22:40] we can also do some counterfactual  
[01:22:42] like in Ernest's case.  
[01:22:44] What we'll do is we'll study  
[01:22:46] what might happen if the US cuts its foreign aid budget  
[01:22:48] as seems to be happening right now.  
[01:22:50] We study first the Chinese reaction  
[01:22:52] and second the reaction of the alignment  
[01:22:54] in the rest of them.  
[01:22:56] In terms of the Chinese we find that they're projected  
[01:22:58] to decrease their own aid in most countries  
[01:23:00] because influence becomes cheaper to buy.  
[01:23:02] And in terms of global alignment  
[01:23:04] we find that there's a small shift away from this  
[01:23:07] so we don't find these large effects  
[01:23:09] so it's not that the world order is changing  
[01:23:11] but potentially allies of the US  
[01:23:13] are moving more than that.  
[01:23:15] So the paper contributes to a number of strands of literature  
[01:23:19] mostly importantly of course  
[01:23:21] why we're here at Geo-economics  
[01:23:23] what we're trying to do with this paper  
[01:23:25] is we're trying to provide some quantitative evidence  
[01:23:27] on how effective these geoeconomic tools  
[01:23:29] that we're studying in many papers  
[01:23:31] actually are.  
[01:23:33] That allows us to answer empirically  
[01:23:35] as we see the world economy shifting  
[01:23:37] as hegemon's are taking different policies  
[01:23:39] what consequences might this have  
[01:23:41] or how the other hegemon's are responding  
[01:23:43] and what the consequences might this have  
[01:23:45] for geopolyms.  
[01:23:47] Of course there's also literature on foreign aid  
[01:23:49] that sort of thought about these questions before  
[01:23:51] there's a literature on the economics of the Civil War  
[01:23:53] and there's been a presented nice paper  
[01:23:55] similar to some data  
[01:23:57] sort of leveraging that setting  
[01:23:59] to understand how geoeconomics  
[01:24:01] is changing.  
[01:24:03] And understand how geoeconomic competition might play out.  
[01:24:07] So let me start with  
[01:24:09] sort of a simple model.  
[01:24:11] I don't have time to cover all the details  
[01:24:13] and sort of want to get more into the data  
[01:24:15] but I'll also give you the rough outline  
[01:24:17] of how we think of the world  
[01:24:19] and what we're going to take to the data later on.  
[01:24:21] So we're going to think of a world  
[01:24:23] with two types of values  
[01:24:25] hegemon's and small hegemon's  
[01:24:27] so there's going to be two hegemon's  
[01:24:29] the Soviets and the Americans  
[01:24:31] and using carrots and sticks when those are  
[01:24:33] comes on the next slide  
[01:24:35] all the variables with respect to the Soviets  
[01:24:37] are half a star.  
[01:24:39] There's also going to be a collection of small countries  
[01:24:41] indexed by I  
[01:24:43] and these countries are going to choose their  
[01:24:45] geopolitical alignment  
[01:24:47] on the human line so they can live all the way with Americans  
[01:24:49] at one, they can live with the Soviets at zero  
[01:24:51] and they can live anywhere in between those two.  
[01:24:53] In the background  
[01:24:55] we're going to think of a standard  
[01:24:57] trade model of following sort of Armington  
[01:24:59] and that's going to be  
[01:25:01] describing the economic environment  
[01:25:03] within that setting  
[01:25:05] we're going to think of a two stage date  
[01:25:07] so in the first stage hegemon's  
[01:25:09] make payments and threats  
[01:25:11] to the other countries what those are  
[01:25:13] comes in a second  
[01:25:15] in doing so what the hegemon's care about  
[01:25:17] we assume is they maximize the weight  
[01:25:19] and sum of the line in other countries  
[01:25:21] so they hear this is written out for the US  
[01:25:23] so they like it when other countries move close to that  
[01:25:25] and they don't like paying  
[01:25:27] other countries and they of course don't like  
[01:25:29] if they were working for sanctions  
[01:25:31] they also don't like imposing sanctions  
[01:25:33] so here these VI  
[01:25:35] are the importance of different countries  
[01:25:37] there's exogenous  
[01:25:39] and we assume that they're primitive  
[01:25:41] of the hegemon's preferences  
[01:25:43] similar in the second stage  
[01:25:45] countries move and they choose  
[01:25:47] their alignment given what's on the table  
[01:25:49] and they have some preference  
[01:25:51] high that they serve  
[01:25:53] balance against economic gains and losses  
[01:25:55] and then they're going to choose an alignment  
[01:25:57] between the Americans and the citizens  
[01:25:59] a bit more in detail how do these tools work  
[01:26:01] so for the  
[01:26:03] the threats  
[01:26:05] that's the stick  
[01:26:07] we're going to allow hegemon's to make tariff threats  
[01:26:09] that's the tariff threat for the Americans  
[01:26:11] and these tariff threats are going to specify  
[01:26:13] countries that if they move too far away  
[01:26:15] from the hegemon they're going to be hit  
[01:26:17] and they're going to get sanctions  
[01:26:19] so in this is a one shot game  
[01:26:21] so if you sanction you're sanctioned  
[01:26:23] an infinite tariff  
[01:26:25] so full exclusion of trade  
[01:26:27] with the hegemon no more imports  
[01:26:29] no more exports so for example like last week  
[01:26:31] Trump was saying you know he'll cut Spain  
[01:26:33] off of trade and this is that  
[01:26:35] so  
[01:26:37] it's important primitive  
[01:26:39] of this model is  
[01:26:41] how much these countries would suffer if they were hit by that  
[01:26:43] so  
[01:26:45] that's what we're going to find with the power  
[01:26:47] of a hegemon insist how much  
[01:26:49] countries would lose consumption if a hegemon were to punish that  
[01:26:51] if you can think of that as being the size  
[01:26:53] of the stick of the hegemon  
[01:26:55] we then saw  
[01:26:57] for the optimal threats  
[01:26:59] and we showed that the optimal threat depends on the difference  
[01:27:01] in the power between these two hegemon  
[01:27:03] so the way you can think about this is  
[01:27:05] in a game with a single hegemon  
[01:27:07] the hegemon takes away all the surplus  
[01:27:09] from trade from these countries  
[01:27:11] but now if there's another hegemon on the other side  
[01:27:13] you can't take away all the surplus  
[01:27:15] but only the surplus you have in excess  
[01:27:17] of the hegemon  
[01:27:20] and this is the best response to the  
[01:27:26] best response of the hegemon to one another  
[01:27:28] so this is the best response of the Americans  
[01:27:30] to what the Soviets are doing  
[01:27:32] so you can see if the Soviets are doing very little  
[01:27:34] it's optimal for the US to do very little  
[01:27:36] as the Soviets wrap it up  
[01:27:38] the Americans wrap it up as well  
[01:27:40] but at some point if the Soviets keep pushing  
[01:27:42] it comes too expensive to compete  
[01:27:44] and the US goes at the top  
[01:27:46] similar for the Soviets in this  
[01:27:48] of intersectionist equilibrium, here everything  
[01:27:50] is drawn up to be perfectly symmetric,  
[01:27:53] doesn't need to be the case in the paper, of course.  
[01:27:56] And what we're also doing in the paper,  
[01:27:57] I don't have time to go through this today,  
[01:27:59] is sort of testing the shape of these response functions  
[01:28:02] and how they align with the data.  
[01:28:06] Next, we set everything up so that we can solve  
[01:28:08] the reaction function of these countries in closed form  
[01:28:11] and we can write the alignment that these countries choose  
[01:28:13] as a function of their preference pi.  
[01:28:16] How much is hegemon is paying them,  
[01:28:18] where beta and beta stars, how much you like money  
[01:28:20] from the Americans and the Soviets  
[01:28:22] and the difference in the power of the Americans  
[01:28:25] and the Soviets.  
[01:28:27] So this is the reaction function,  
[01:28:28] it looks like a regression and that's the regression  
[01:28:30] we'll estimate in the data later on.  
[01:28:33] Now let me tell you how we get this through the data.  
[01:28:35] So we get this through the data  
[01:28:36] using the cold war as an example.  
[01:28:40] And the cold war is of course a nice setting  
[01:28:41] to think about these questions  
[01:28:43] because it's a setting where there was no much,  
[01:28:45] not much direct interaction between the two hegemon,  
[01:28:47] but there was direct interaction between the hegemon  
[01:28:50] and other small countries.  
[01:28:53] So it's a setting where the alignment  
[01:28:55] of these countries matter.  
[01:28:57] So ultimately of course on the side of both hegemon  
[01:29:00] there's a certain existential fear  
[01:29:02] that these countries would split to the other side  
[01:29:04] and there could be some guidance  
[01:29:06] pointed at you from these countries.  
[01:29:09] We also know that the main geoconomic instruments  
[01:29:13] that these countries were using during the Cold War  
[01:29:15] were aid and trade.  
[01:29:18] So this we know both by reading the classified reports  
[01:29:20] from the CA, but also by consulting  
[01:29:24] to come to various literature in these international relations  
[01:29:27] at that time that writes these as a main instruments  
[01:29:30] of economic state.  
[01:29:31] So this is what we're studying today.  
[01:29:33] So what historians call the real battleground  
[01:29:35] of the Cold War is the competition of these  
[01:29:38] two hegemon for influence in small countries.  
[01:29:41] Here's an example of what these countries were doing.  
[01:29:43] So this is the Aswan Dam which is this big dam  
[01:29:46] that the Americans built in Egypt during the 1960s  
[01:29:50] and sort of the Soviets built it  
[01:29:51] and they made very nice posters  
[01:29:53] of it trying to influence that country.  
[01:29:56] So in contrast the Americans don't spend pats.  
[01:29:59] So when the Soviets built this dam  
[01:30:01] it threatens to flood these temples at Abbas symbol.  
[01:30:04] So they have to move the temples need to be moved  
[01:30:06] and the Americans end up paying to move these temples  
[01:30:09] to sort of ward off the Soviet attempt heading.  
[01:30:12] So these are the kind of things we wanna capture  
[01:30:14] at least in the data.  
[01:30:17] So here's the data we're gonna use.  
[01:30:19] First we're gonna measure tariffs.  
[01:30:21] So this we're gonna measure as economic and military aid  
[01:30:24] provided by the US and the Soviet Union.  
[01:30:26] So for the Soviet Union the nice thing is that  
[01:30:29] during the Cold War the CIA was actually collecting  
[01:30:31] this data and the CIA was subtracting  
[01:30:33] in which country were the Soviets sending their money  
[01:30:36] and here's a snap out of this data.  
[01:30:38] For the US this is relatively straightforward to get  
[01:30:40] from what's called the Green Booking  
[01:30:41] and that's the standard source for US foreign aid.  
[01:30:45] So here's a snapshot of this data.  
[01:30:46] So this is for all the countries that the CIA calls  
[01:30:49] non-aligned countries.  
[01:30:50] They have a list of countries which are they're tracking.  
[01:30:53] And what I'm showing you is for every country  
[01:30:55] average German and Tyro-Cole war,  
[01:30:57] how much money comes from the Americans  
[01:30:59] relative to how much money comes from the Soviets.  
[01:31:02] You can see there's lots of variations.  
[01:31:04] Some countries are very blue,  
[01:31:05] you get most from the Americans.  
[01:31:07] Some countries are very red,  
[01:31:08] most come from the Soviets,  
[01:31:09] but there's also many countries  
[01:31:10] which are kind of pinkish.  
[01:31:11] So think of Egypt or India in this kind of chart.  
[01:31:15] They sort of get money from both sides  
[01:31:17] as I was showing you with this example.  
[01:31:20] Just in terms of the economic size of these flows,  
[01:31:23] I mean there's, we compare them here to GDP.  
[01:31:27] So there's many concerns you have in terms of valuation  
[01:31:31] on the measurement of GDP in these countries.  
[01:31:33] So take these numbers with the grain and salt.  
[01:31:36] But these are very large capital interests  
[01:31:37] into these countries.  
[01:31:38] So for the average recipient,  
[01:31:41] that this is around 2.5% of GDP.  
[01:31:44] So to first order,  
[01:31:45] this is likely the main capital inflow  
[01:31:47] into these countries at that point.  
[01:31:52] Next, we're gonna measure the stick  
[01:31:54] or the size of the stick.  
[01:31:55] So what happens in our models,  
[01:31:57] no country gets punished  
[01:31:58] because these punishments are gonna be off equilibrium,  
[01:32:01] but we can measure how much  
[01:32:02] each hegemon can threaten.  
[01:32:06] So this is what we do  
[01:32:07] using the model of international trade  
[01:32:09] where we can compute how much  
[01:32:10] each country would suffer  
[01:32:12] if they had to want to get punished with sanctions.  
[01:32:15] So this is like a trade dependence, measure in a sense.  
[01:32:18] And we're gonna use the trade model  
[01:32:20] to sort of take the under-faction consumption  
[01:32:23] these countries would have if they had to want to get punished.  
[01:32:27] So here's the power of the Americans.  
[01:32:29] You can see this, it looks kind of like  
[01:32:31] you would expect it to look.  
[01:32:33] The US has lots of power over countries in Americas  
[01:32:36] and Latin America more generally.  
[01:32:38] Less power if you get further away  
[01:32:40] just because there's less international trade  
[01:32:43] with these countries at that point in time.  
[01:32:46] The opposite kind of holds for the Soviets  
[01:32:48] where there's more power in the Middle East  
[01:32:51] and in Eastern Europe  
[01:32:53] whereas they're more closely connected  
[01:32:55] to the Soviet Union.  
[01:32:57] On average, the Soviets of course  
[01:32:59] have less economic power than the US  
[01:33:02] just because they're a smaller economic hub  
[01:33:04] than the Americans.  
[01:33:07] Finally, let me tell you how we measure alignment  
[01:33:09] so I don't have to say that much  
[01:33:11] because Ernest was using something similar.  
[01:33:13] So we measure alignment using voting at the UN  
[01:33:16] using the standard measure from Bailey's  
[01:33:18] to just enough voted so this is all off the shelf.  
[01:33:22] Which is just for those of you who don't know this measure  
[01:33:25] this is sort of a unit dimensional measure  
[01:33:27] out of all the votes cast at the UN.  
[01:33:30] So we think of one way of thinking the alignment  
[01:33:32] of these countries chooses.  
[01:33:34] There's many actions, some they vote yes,  
[01:33:36] some they vote no and this is sort of the average action  
[01:33:38] they take in this with the average alignment  
[01:33:40] of these countries too.  
[01:33:42] So this is a reasonable measure of alignment.  
[01:33:44] So in the paper we do some validation.  
[01:33:47] We do some qualitative validation going through case studies.  
[01:33:49] We do some quantitative validation comparing for example,  
[01:33:53] does it predict whether you attend the Olympics?  
[01:33:55] But I think the easiest way of validating  
[01:33:57] is this plotting.  
[01:33:58] So what you can see here is this measure  
[01:34:01] plotted for the Soviets and their allies in red,  
[01:34:04] the US and NATO in blue  
[01:34:07] and all the non-aligned countries from the CIA list  
[01:34:10] in a break and you can see it sort of recover  
[01:34:12] the central cleavages of the Cold War.  
[01:34:14] So you have the US block on one side,  
[01:34:16] the Soviet block on the other side.  
[01:34:18] You have the non-aligned countries sort of moving  
[01:34:20] in between those two which is sort of what you would want  
[01:34:23] from a measure of alignment at that point in time  
[01:34:25] and it sort of comes out of this UN voting.  
[01:34:27] So this is sort of relatively reassuring.  
[01:34:31] Now let me tell you how we estimate this reaction  
[01:34:33] which is sort of the heart of the model.  
[01:34:37] So we actually have a similar chance  
[01:34:38] to what we had in the last day.  
[01:34:41] So we have this reaction function  
[01:34:42] which describes the alignment that these countries choose  
[01:34:46] as a function of preferences,  
[01:34:47] how much each hegemon pays  
[01:34:49] and the relative power of each hegemon.  
[01:34:52] So like in Ernest's paper,  
[01:34:54] so we observe the alignment of these countries.  
[01:34:58] We observe how much hegemon are paying  
[01:35:00] and how much they can threaten.  
[01:35:02] That's what the data was showing you,  
[01:35:03] but we don't know sort of the preference.  
[01:35:05] So we don't know what Ernest was calling theta  
[01:35:07] and what's here pi.  
[01:35:09] So that's sort of the key unobserved.  
[01:35:11] That's key sort of threat to identification  
[01:35:14] and so this is what we'll try to tackle  
[01:35:16] using instrumental variables.  
[01:35:17] We have some strategies also actually similar  
[01:35:19] or similar to Ernest where we sort of control  
[01:35:21] for the ideology of the leader  
[01:35:23] where there's some left and right classifications  
[01:35:25] which you might take of a measure of pi.  
[01:35:28] But primarily I'll show you today  
[01:35:30] an instrumental variable.  
[01:35:32] There's of course other identification concerns  
[01:35:34] you might have, there's measurement error  
[01:35:35] and all these things.  
[01:35:37] This is not a complete description of geopolitics of course.  
[01:35:41] So there's other tools that are moving around simultaneously.  
[01:35:45] So we estimate this in the standard panel regression  
[01:35:49] using country year fixed effect and year fixed effects.  
[01:35:52] We're gonna measure aid in per capita units  
[01:35:54] so it could count for the size of these countries  
[01:35:57] and control for slow moving trends  
[01:35:58] because that's sort of gonna help  
[01:35:59] with the regional identification we're gonna pursue  
[01:36:02] and I'll show you the instruments in a few seconds.  
[01:36:05] Here we're gonna impose the same coefficient  
[01:36:07] on the power of the Americans in the Soviets  
[01:36:09] that sort of related to the instrument we'll use  
[01:36:11] that's gonna be cool in here  
[01:36:12] between the Americans and the Soviets.  
[01:36:14] So we cannot at least in the IV setting  
[01:36:16] identify the effects of US and Soviet power separately.  
[01:36:21] So this is just the OLS regression.  
[01:36:23] So here positive values mean more alignment with the US  
[01:36:27] and things sort of seem to go  
[01:36:29] the way you would think it would go.  
[01:36:30] The Americans give you more money,  
[01:36:32] you tend to move that way,  
[01:36:33] the Soviets give you more money,  
[01:36:35] you also tend to move that way.  
[01:36:37] And so similarly as the Americans  
[01:36:39] become relatively more economically powerful.  
[01:36:43] So these are associations you would expect.  
[01:36:45] You can also run this dynamically,  
[01:36:47] so this is like many of these models  
[01:36:49] as a static model,  
[01:36:50] but at least in the data we can sort of do dynamics.  
[01:36:53] So here we run this as a local projection  
[01:36:55] and what we find is that these sort of dynamics  
[01:36:58] are at least relatively short.  
[01:37:00] So you're not moving countries persistently over or up  
[01:37:03] and then they sort of fall into your camp forever,  
[01:37:05] but rather after three, four years,  
[01:37:07] they sort of go back to the original alignment.  
[01:37:10] So then let me tell you how we sort of try to tackle  
[01:37:13] these kind of dictation challenges,  
[01:37:14] which are of course very thorny  
[01:37:16] because it's very hard to get something  
[01:37:17] that sort of responds not to the preferences of these models.  
[01:37:22] So for the carrot,  
[01:37:25] we're gonna think through the lens of the model.  
[01:37:27] So of course in the model,  
[01:37:28] how much is country is paid is endogenous.  
[01:37:31] It's a choice by the data hegemons,  
[01:37:33] but the choice depends on how much hegemons value  
[01:37:36] different countries and the preferences of these models.  
[01:37:40] And so ideally what we would like is like some shock  
[01:37:44] to the preferences that hegemons have  
[01:37:46] for the alignment of other countries.  
[01:37:48] So we would like some sort of pull people turnover  
[01:37:51] in the hegemon that leads the hegemon to reallocate money,  
[01:37:54] not the money to do something that's happening  
[01:37:56] in the small countries.  
[01:37:59] So we're gonna try to approximate this  
[01:38:00] as best we can using a shifter instrument,  
[01:38:03] which leverages the countries  
[01:38:04] are differentially exposed to regional fluctuations in foreign aid.  
[01:38:08] So this instrument is constructed as the average flow  
[01:38:11] to a world region that's been going  
[01:38:13] to a particular country in previous years  
[01:38:16] and the total flow to that region.  
[01:38:18] So here the argument is sort of that hegemons  
[01:38:20] have some sort of budget allocation,  
[01:38:22] which sort of sends money across different places  
[01:38:24] of the world and then as these budgets are raised  
[01:38:27] or lowered some countries are treated more by that  
[01:38:30] than other countries and they're not doing  
[01:38:32] these budget reallocations in response  
[01:38:34] to shocks in these small countries.  
[01:38:37] So we've got through in the papers  
[01:38:39] with some checks on these shifter instruments.  
[01:38:41] So for one thing we do is we provide some narratives  
[01:38:44] at least for Latin America,  
[01:38:45] where we sort of show that the major shifts  
[01:38:47] are sort of linked to changes in the US administration.  
[01:38:50] So for example, 1969.  
[01:38:52] So a big foreign policy shift under Kissinger  
[01:38:54] which was like a big cut of foreign aid to Latin America  
[01:38:57] as US prior to shift towards the Pacific.  
[01:39:00] We also have like an event study for Jimmy Carter  
[01:39:02] where Jimmy Carter cut aid to Latin America  
[01:39:04] and then he compared that line of different countries.  
[01:39:08] So that's the instrument for foreign aid.  
[01:39:11] For the stick that's sort of the trade independence  
[01:39:14] of our different countries,  
[01:39:16] you know the reliant on the literature  
[01:39:18] and trading history of private Steven and co-authors  
[01:39:21] than similarly and using a time-varying gravity equation,  
[01:39:25] we're gonna predict flows of international trade  
[01:39:27] using different countries using changes  
[01:39:30] in the cost of air versus sea distance  
[01:39:32] and we're gonna use that predicted measure  
[01:39:34] of international trade that depends on technology  
[01:39:38] as an instrument for international trade.  
[01:39:40] You can see this is gonna be collinear  
[01:39:42] between the Americans and the Soviets  
[01:39:43] because as you move closer and predict the distance  
[01:39:45] to the Americans, you move further away  
[01:39:48] in predicted best difference to the Soviets.  
[01:39:51] So you might think, you know, there's other things  
[01:39:54] that sort of responding to these changing costs  
[01:39:57] of air travels like measure of cultural exchange.  
[01:40:00] So we've checked this event in the paper  
[01:40:02] so the CA was also collecting this data on, you know,  
[01:40:05] how many students to attend Soviet universities  
[01:40:08] send data for the US.  
[01:40:10] And so we find that these sort of respond  
[01:40:11] differentially to these measures of air  
[01:40:13] and sea distance costs.  
[01:40:14] The way we interpret this is that during the Cold War  
[01:40:17] the main way of exchanging across countries  
[01:40:20] was using international trade, not so much the movement.  
[01:40:25] So let me show you the OLS  
[01:40:27] and then we'll go through the instruments.  
[01:40:29] So here's the OLS regression  
[01:40:30] when we sort of instrument the US foreign aid,  
[01:40:34] we sort of find that the coefficient gets slightly larger,  
[01:40:36] doesn't move that much.  
[01:40:37] Similarly, as we instrument Soviet foreign aid,  
[01:40:41] as we instrument these power differences,  
[01:40:43] sort of these, there's a bigger change  
[01:40:45] of coefficient that's normal in this literature  
[01:40:48] and the sample shrinks a lot because you can of course  
[01:40:51] only do this for countries that are on a coastline.  
[01:40:54] So we also have some weak instrument for us  
[01:40:56] for insiporism.  
[01:40:58] If you care about that,  
[01:40:59] and finally we can instrument everything simultaneously  
[01:41:02] and this is sort of the coefficients we're gonna use.  
[01:41:05] So these estimates are not very precise,  
[01:41:07] they're relatively noisy.  
[01:41:09] We think of, and voting as like a noisy proxy  
[01:41:12] of the true alignment of these countries  
[01:41:14] and so you wouldn't expect these coefficients  
[01:41:16] to be particularly precise.  
[01:41:19] Going to sort of, to Ernest's paper,  
[01:41:21] if we have some sort of low tech version  
[01:41:23] of network regressions as well,  
[01:41:25] where we sort of impose that the network  
[01:41:26] is proportional to trade integration and the difference  
[01:41:30] and so we do find some network effects as well there  
[01:41:33] but they sort of, in order of regressions  
[01:41:35] tend to be smaller than these direct.  
[01:41:39] Now, what do these numbers mean?  
[01:41:41] These numbers means you can do alignment  
[01:41:43] using economic tools,  
[01:41:45] but it's relatively expensive to buy out.  
[01:41:48] So for example, we find that W for an H,  
[01:41:50] so that's $120 million on average for the US,  
[01:41:54] moves that country by only 2.5% of its standard deviation,  
[01:41:57] so that's kind of like raising the odds  
[01:41:59] of you bar-copping the Olympics by 5%,  
[01:42:02] it's 2% of a Pinochet coup or 1% of the geopolitical distance  
[01:42:06] between a neutral country and a negative one.  
[01:42:08] So you move these countries,  
[01:42:09] but it's very expensive to move these countries very fast.  
[01:42:12] Similarly, you can move these countries  
[01:42:14] by making them more dependent on you  
[01:42:16] relative to their dependence on your rival.  
[01:42:19] And so to get the same shift in geopolitical alignment,  
[01:42:22] you can move a country,  
[01:42:24] sort of you need to raise your power over that country  
[01:42:26] by about 5% of Mexico's dependence,  
[01:42:28] it's around 3% of total trade to GDP.  
[01:42:31] Of course, this is gonna be preferable  
[01:42:33] from the perspective of a hegemon  
[01:42:34] because these payments are sort of costly  
[01:42:37] whereas these threats are gonna be off-path,  
[01:42:39] so you can sort of move these countries  
[01:42:40] without actually having to carry out the same.  
[01:42:44] Now are these results reasonable?  
[01:42:46] So the first way of comparing this  
[01:42:48] is to literature and political science  
[01:42:49] that's estimated similar in regression to before.  
[01:42:52] And we sort of set the model up  
[01:42:54] to have the exact same sort of specification  
[01:42:57] that these people have been running  
[01:42:58] and sort of these values that we find  
[01:43:00] are very much in line with what political scientists  
[01:43:02] have found, at least in more modern data.  
[01:43:04] It's very relatively hard to shift alignment,  
[01:43:07] at least you measured using voting at the UN.  
[01:43:10] But it's also consistent with sort of narratives  
[01:43:12] of historians of the Cold War.  
[01:43:14] That are arguing that as the Cold War unfolds,  
[01:43:18] both the US and the Soviets,  
[01:43:20] they start these very expensive adventures in many countries  
[01:43:23] so there's massive amounts of money  
[01:43:24] according to Egypt for all these dams and so on.  
[01:43:27] But these countries sort of end up shifting relatively little.  
[01:43:30] So it's none of these times ever  
[01:43:31] becomes like a NATO member.  
[01:43:33] So even in the Cold War, according to historians,  
[01:43:35] these policies were relatively unsuccessful.  
[01:43:39] Finally, so if we've now set up the model,  
[01:43:42] we've just estimated the reaction function  
[01:43:44] of these countries.  
[01:43:45] And we can use that to speak a bit to the present.  
[01:43:49] Concretely, we're going to talk about the shut-down of the US  
[01:43:52] and the ID, which sort of happened last year in February  
[01:43:55] when sort of the US started cutting most of its development.  
[01:43:58] And of course, there's major concerns  
[01:44:00] about implications this might have  
[01:44:02] for global welfare, for global health,  
[01:44:04] but potentially also for global geopolitics.  
[01:44:08] What we're going to do is we're going to extend  
[01:44:10] all the measures we have to the present.  
[01:44:12] So that's become a lot easier to do now  
[01:44:15] than during the Cold War.  
[01:44:17] So for sticks, we can run the exact same trade model.  
[01:44:20] For carrots, we can rely on the same data for the US  
[01:44:23] and there's eight data for the Chinese.  
[01:44:25] And for alignment, we can use the same measure  
[01:44:28] of voting at the U.S.  
[01:44:30] We're then going to calibrate this reaction function  
[01:44:32] using the historical data.  
[01:44:35] So that's some extrapolation that of course  
[01:44:37] these values of the reaction function are the same,  
[01:44:40] but given an absence of better data  
[01:44:43] it's sort of the first place to start.  
[01:44:46] Given that, we're then going to evaluate the USA shut-down.  
[01:44:49] This involves two steps.  
[01:44:51] So in the first step, we're going to study  
[01:44:53] how does China react.  
[01:44:55] So remember we have these reaction functions,  
[01:44:57] which are a bit curvy,  
[01:44:59] and we're going to shift around the US function  
[01:45:01] and see how China reacts.  
[01:45:03] In doing so, we're going to have to hold  
[01:45:05] the preferences of these countries  
[01:45:07] that importance China places on different countries  
[01:45:09] and how it reacts.  
[01:45:11] You could let things vary as well,  
[01:45:13] but for baseline it's easiest to see  
[01:45:15] when you hold the specs.  
[01:45:17] And then finally, we're going to study how alignment reacts.  
[01:45:19] So first, when it comes to the Chinese reaction  
[01:45:22] that's probably the most interesting part.  
[01:45:24] So for China there's two considerations.  
[01:45:27] On the one hand, that the US is gone,  
[01:45:30] it's become a lot cheaper to buy influence  
[01:45:32] countries because you don't have a competitor.  
[01:45:34] On the other hand, of course,  
[01:45:36] now that the US is gone, you could set in  
[01:45:39] and you could call all these countries to your side  
[01:45:41] and at least directly after the USAID shut-down  
[01:45:44] there was lots of concern that this is what's going to happen.  
[01:45:47] So through the lens of our model,  
[01:45:49] this is not optimal for China to be in most countries.  
[01:45:52] So you can see here what's the predictive  
[01:45:54] Chinese reaction in most countries.  
[01:45:56] And you can see in most of the world,  
[01:45:58] especially in Africa, where most of Chinese  
[01:46:00] state is concentrated on the Chinese release,  
[01:46:03] China's actually predicted to decrease its 4.8  
[01:46:05] by around 10 to 15%.  
[01:46:07] So there are some points of the world  
[01:46:09] where they increase.  
[01:46:10] As usual, Latin America, Eastern Europe,  
[01:46:12] what's driving these differential responses  
[01:46:15] is that these countries in Africa  
[01:46:17] are initially already more aligned with the Chinese.  
[01:46:20] So there's this motive of you can push  
[01:46:23] these countries to your side, it's not very strong.  
[01:46:26] Is this reasonable?  
[01:46:28] So you can compare this to the end of the Cold War.  
[01:46:31] There's some evidence that at the end of the Cold War  
[01:46:33] that Americans also decreased their 4.8 budget,  
[01:46:36] especially due to strategic countries  
[01:46:38] because there was sort of this absence of a competitor.  
[01:46:41] And you can also, there's some preliminary work  
[01:46:43] from College of Heal where they studied  
[01:46:45] the Chinese reaction to the USAID shut-down  
[01:46:48] and they said find a very muted reaction.  
[01:46:50] And we read this as not being inconsistent  
[01:46:53] with what we find.  
[01:46:55] Finally, what happens to global geopolitical alignment,  
[01:46:58] we find that it goes down with the US,  
[01:47:00] that's not surprising, nothing else can happen in the model.  
[01:47:03] But it goes down relatively modestly.  
[01:47:05] So about only 50% of the distance between  
[01:47:08] France and Turkey in terms of geopolitics  
[01:47:10] for that.  
[01:47:12] The countries that move the most, of course,  
[01:47:14] are countries initially more aligned with the East.  
[01:47:17] Finally, you can ask the model,  
[01:47:19] can you offset this loss of influence  
[01:47:22] through 4.8 by tightening the screws a bit more  
[01:47:25] by increasing your geopolitical power?  
[01:47:28] At least through the lens of the model  
[01:47:30] we need very large increases in geopolitical power.  
[01:47:33] So we really need to tighten the screws a lot  
[01:47:36] in a sense to increase trade dependence of countries  
[01:47:40] to offset this.  
[01:47:42] Perhaps you can do this in Latin America  
[01:47:44] where countries are very integrated with you.  
[01:47:46] It's very hard to replicate this  
[01:47:48] in countries in Africa, which at least  
[01:47:50] in trade data are not that reliable.  
[01:47:53] So let me close here.  
[01:47:56] So this paper was the first attempt  
[01:47:58] to take these big geoeconomic models  
[01:48:00] and put them to the data to allow us to speak quantitatively  
[01:48:03] are the effects of these geoeconomic instruments  
[01:48:05] big or small, what can we expect  
[01:48:07] as countries are changing their geoeconomic instruments?  
[01:48:10] What we find is that from the perspective of a hegemon,  
[01:48:13] the geoeconomic tools that these hegemon are using work,  
[01:48:16] but they're relatively expensive  
[01:48:18] and it's hard to move these countries to your point.  
[01:48:20] From the perspective of small countries,  
[01:48:22] there's some results in this paper  
[01:48:24] that suggest that competition might not be that bad  
[01:48:27] so you can do less course  
[01:48:29] and perhaps you can extract some payment from both sides.  
[01:48:31] When it comes to the US and China,  
[01:48:33] at least what our model suggests  
[01:48:35] is that the rise of China in global trade  
[01:48:37] might be a lot more powerful  
[01:48:39] than China's rise as an aid donor  
[01:48:41] just because it's a lot cheaper  
[01:48:43] to turn to these countries by threatening them  
[01:48:45] than by having them keeping to pay them all the time.  
[01:48:48] And there's sort of these dynamics  
[01:48:50] of a competition might be a bit unexpected,  
[01:48:52] it's not the case that when one hegemon  
[01:48:54] backs away from competition,  
[01:48:56] the other hegemon may necessarily step in.  
[01:48:58] So with this, I close here  
[01:49:00] and I look forward to the discussion  
[01:49:02] and thanks a lot for your attention.  
[01:49:04] Our discussion is Alberta.  

## Discussion (01:49:12 – 02:04:38)

[01:49:13] Okay.  
[01:49:29] Thank you very much for the organizers  
[01:49:31] for inviting me to discuss  
[01:49:33] this very interesting paper by Tim  
[01:49:35] and Nicolas.  
[01:49:38] So let me start by saying this a paper about  
[01:49:40] how should non-aligned countries  
[01:49:42] navigate the world of hegemon  
[01:49:44] and how do hegemon influence  
[01:49:47] the behavior of these countries.  
[01:49:49] So let me start with a picture here.  
[01:49:52] This is Juan Perón,  
[01:49:54] he was president of Argentina  
[01:49:56] late 40s, early 50s.  
[01:49:58] And he was one of the founders  
[01:50:00] of these non-aligned movements  
[01:50:02] and he had a wonderful phrase,  
[01:50:04] understand it, he says  
[01:50:06] ni yanquis ni marxistas,  
[01:50:08] peronistas.  
[01:50:10] So basically how do we  
[01:50:12] stay away from these two poles  
[01:50:14] and how should we navigate  
[01:50:16] this world.  
[01:50:18] So this paper does exactly that,  
[01:50:20] it asks itself in this world  
[01:50:22] how do hegemon shape the behavior  
[01:50:24] of these third countries,  
[01:50:26] in particular how do we use  
[01:50:28] sticks and carrots or rewards  
[01:50:30] and punishments to shape  
[01:50:32] this world.  
[01:50:34] So this paper does basically  
[01:50:36] a couple of things.  
[01:50:38] Basically it has a simple conceptual framework.  
[01:50:40] It takes then the framework  
[01:50:42] to cold war data.  
[01:50:44] And what do we learn from this paper  
[01:50:46] where there's two key findings I took from it,  
[01:50:48] there's many findings,  
[01:50:50] but there's two key findings.  
[01:50:52] The first one is that yes,  
[01:50:54] it seems that this third country  
[01:50:56] alignment was affected by sticks  
[01:50:58] and carrots during the cold war.  
[01:51:00] So I think that's the first  
[01:51:02] thing that I learned from this  
[01:51:04] alignment highly.  
[01:51:06] So just to put the thing in context  
[01:51:08] because some of my comments have to do with that,  
[01:51:10] let me tell you quickly  
[01:51:12] what the setup is.  
[01:51:14] The setup is a country that chooses  
[01:51:16] a continuous alignment action  
[01:51:18] between zero and one, zero here  
[01:51:20] being the USSR, one being the  
[01:51:22] US, and I maximize  
[01:51:24] a welfare function like the one  
[01:51:26] that you see over here.  
[01:51:28] So I don't want to take actions too far  
[01:51:30] from my preferences, but on the other hand  
[01:51:32] I also derive utility from the carrots.  
[01:51:34] That's what you see there in blue.  
[01:51:36] They can come from the US or the USSR.  
[01:51:38] This is like a transfer.  
[01:51:40] And from the sticks, these are the tariffs  
[01:51:42] that you see there in red, also from the  
[01:51:44] US and the USSR.  
[01:51:48] Now, there's two assumptions  
[01:51:50] I want to point out.  
[01:51:52] I will come back to the first one is that  
[01:51:54] tariffs are either zero or maximum.  
[01:51:56] They are either in free trade  
[01:51:58] or they cut bilateral trade  
[01:52:00] with the hegemon and their definition  
[01:52:02] of power indeed is based on the  
[01:52:04] these sticks.  
[01:52:06] So power is measured as the difference  
[01:52:08] between what I would get in terms of consumption  
[01:52:10] in the baseline, say with no  
[01:52:12] tariffs versus what I would get  
[01:52:14] in terms of consumption.  
[01:52:16] If I was excluded  
[01:52:18] from trading with the hegemon.  
[01:52:20] The second one, second assumption  
[01:52:22] transfers, they give to me these are the carrots.  
[01:52:24] They are continuous.  
[01:52:26] And actually one thing they assume is that they affect my bliss point.  
[01:52:28] Remember, that's the thing I want to be close to.  
[01:52:30] So my bliss point depends on  
[01:52:32] some preferences that I have.  
[01:52:34] But they're affected by the carrots.  
[01:52:36] When I get money from the US,  
[01:52:38] my bliss point moves towards the US  
[01:52:40] and vice versa when I get money from the USSR.  
[01:52:42] Okay?  
[01:52:44] So that's the perspective of my country.  
[01:52:46] I'm getting these carrots and sticks and they're affecting my alignment.  
[01:52:48] What are the hegemon  
[01:52:50] doing? Well, you have it there.  
[01:52:52] And they're not doing it in the USSR.  
[01:52:54] So they're basically saying, I want to bring,  
[01:52:56] if I'm the US, the A is close to me,  
[01:52:58] I could have specific valuations  
[01:53:00] for each country. That's the new that you see there.  
[01:53:02] But giving the carrots is costly  
[01:53:04] and also using the stick is costly.  
[01:53:06] Because when I use the stick, I don't trade  
[01:53:08] and so also the hegemon loses something.  
[01:53:10] Okay?  
[01:53:12] So this is the game.  
[01:53:14] And what's the definition  
[01:53:16] of equilibrium here?  
[01:53:18] Well, the hegemon will choose carrots and sticks  
[01:53:20] and they will choose their alignment such that  
[01:53:22] we're maximizing all of these objects simultaneously.  
[01:53:24] One thing to note  
[01:53:26] is that the tariffs here, as I already mentioned,  
[01:53:28] they're going to be looking at kind of  
[01:53:30] threshold or cut-off strategies where they're either  
[01:53:32] zero if your alignment is close enough,  
[01:53:34] but they jump  
[01:53:36] to this outer key  
[01:53:38] if you don't align  
[01:53:40] beyond a certain threshold.  
[01:53:42] Okay?  
[01:53:44] So what are the key predictions  
[01:53:46] of this simple framework? Just to organize the data.  
[01:53:48] First of all, my alignment,  
[01:53:50] if I'm a country in the middle of this world,  
[01:53:52] should be increasing in the carrots I get  
[01:53:54] from either side and should be increasing  
[01:53:56] in the size of the stick that they can use.  
[01:53:58] Okay? So the bigger the sticks and the carrots,  
[01:54:00] the more I should align.  
[01:54:02] Second of all,  
[01:54:04] how many carrots should the superpowers give me?  
[01:54:06] Well, that will depend on how much they value my alignment  
[01:54:08] on the evaluation of my alignment.  
[01:54:10] Third of all,  
[01:54:12] the carrot that the U.S.  
[01:54:14] gives me is not monotonic on the carrot  
[01:54:16] that the U.S.S.R. gives me. Initially,  
[01:54:18] they increase, but at some point  
[01:54:20] if the U.S.S.R. is just giving me too many carrots,  
[01:54:22] too many transfers the U.S.  
[01:54:24] gives up and then the transfer begins  
[01:54:26] to fall. Okay? So that's the idea.  
[01:54:28] Now, they take the simple framework to the data.  
[01:54:30] They use cold war data.  
[01:54:32] What are the carrots here?  
[01:54:34] Team explained it. It's kind of official  
[01:54:36] flows. They could, they take them  
[01:54:38] from U.S. sources for the U.S.  
[01:54:40] They take them from CIA data for the  
[01:54:42] U.S.S.R. And actually an interesting thing  
[01:54:44] is that these flows were quite large.  
[01:54:46] So about one and a half percent of GDP  
[01:54:48] for the average country  
[01:54:50] during the Cold War, 55 and 85.  
[01:54:52] What are the sticks?  
[01:54:54] The sticks are model-based estimates  
[01:54:56] of gains from trade, essentially.  
[01:54:58] So what would happen if  
[01:55:00] the U.S. or the U.S.S.R.  
[01:55:02] cut me off completely from bilateral trade?  
[01:55:04] What would be the loss? So that's what they estimate.  
[01:55:06] These are larger for the U.S.  
[01:55:08] because they just traded more.  
[01:55:10] And then the alignment is basically U.N.  
[01:55:12] voting.  
[01:55:14] So what are the key findings?  
[01:55:16] They're here.  
[01:55:18] First of all, like the simple  
[01:55:20] conceptual framework would suggest,  
[01:55:22] alignment does seem to be correlated  
[01:55:24] with both carrots and sticks  
[01:55:26] during the Cold War.  
[01:55:28] Number two, you could think  
[01:55:30] it's this causation or not.  
[01:55:32] They have a kind of compelling  
[01:55:34] IV approach. So for the carrots,  
[01:55:36] they use shift-share type of instrument.  
[01:55:38] Imagine the U.S.  
[01:55:40] suddenly decides in this progress for  
[01:55:42] alliance to give a lot of money from  
[01:55:44] Latin America. I take  
[01:55:46] what was my previous share of  
[01:55:48] aid I received from the U.S.  
[01:55:50] in Latin America. If I'm say Uruguay,  
[01:55:52] and that's a kind of exogenous  
[01:55:54] increase in aid to Uruguay, I see how  
[01:55:56] that changes my alignment. For the sticks,  
[01:55:58] they use change in transport costs  
[01:56:00] as an exogenous shock to bilateral  
[01:56:02] gains from trade and therefore to the  
[01:56:04] size of the stick. Okay? And the  
[01:56:06] results survive.  
[01:56:08] Once we've established this, we can  
[01:56:10] use the model to back out edge and  
[01:56:12] evaluation of alignment of  
[01:56:14] different countries. I mean, given that  
[01:56:16] I'm giving these carrots and given that  
[01:56:18] I see the change in alignment, how much  
[01:56:20] am I valuing it in order to pay that  
[01:56:22] much? And finally, they have an  
[01:56:24] application which is the one team  
[01:56:26] discussed, which is what do I get  
[01:56:28] from eliminating U.S. aid if the  
[01:56:30] U.S. decides to eliminate U.S.  
[01:56:32] aid, how does China react  
[01:56:34] to that? Okay? So anyway,  
[01:56:36] it was a great paper. It was very  
[01:56:38] interesting to read.  
[01:56:40] Also, as I was telling team, I  
[01:56:42] prepared my discussion last  
[01:56:44] night. I've never been happier preparing  
[01:56:46] a discussion after the game  
[01:56:48] of Argentina, so I will forever  
[01:56:50] associate your paper  
[01:56:53] with  
[01:56:55] that wonderful victory. Okay?  
[01:56:57] So anyway, and I think  
[01:56:59] it makes a compelling case  
[01:57:01] that  
[01:57:03] I think it makes a compelling case that  
[01:57:05] the two teams, the carrots  
[01:57:07] and the carrots affect alignment.  
[01:57:09] So I'm going to have three general  
[01:57:11] comments. I actually have a fourth  
[01:57:13] one which I forgot last night in the  
[01:57:15] celebration, so I'm going to add it in  
[01:57:17] the end. But three general  
[01:57:19] comments. One about the conceptual  
[01:57:21] framework, how do we think about  
[01:57:23] carrots, sticks and powers, and  
[01:57:25] the two other one, they don't  
[01:57:27] really challenge the results which  
[01:57:29] I kind of believe, but I want to  
[01:57:31] think a little bit about the  
[01:57:33] simple framework, which is very  
[01:57:35] natural, is that it treats  
[01:57:37] carrots and sticks very asymmetrically.  
[01:57:39] Okay? So we think, for  
[01:57:41] instance, official flows  
[01:57:43] are carrots. Tariffs are  
[01:57:45] sticks, and you think why?  
[01:57:47] Why should an instrument be identified  
[01:57:49] as a carrot or a stick?  
[01:57:51] In the end, there are two sides of the  
[01:57:53] same coin, so I think I have a basic  
[01:57:55] tariff vis-à-vis the rest of the  
[01:57:57] world. If you misbehave, I'll raise  
[01:57:59] it, and if you, that's a stick,  
[01:58:01] I'll misbehave, I'll give you a better deal. That's a carrot.  
[01:58:03] Okay? Likewise with official flows.  
[01:58:05] I can give you official flows at some average  
[01:58:07] terms. If you behave, you'll get a  
[01:58:09] subsidy, and if you misbehave, I'll  
[01:58:11] whatever, tax it. So  
[01:58:13] I'm not so sure why you would want to  
[01:58:15] have this identification of  
[01:58:17] instrument to be either a stick or a carrot.  
[01:58:19] The second one  
[01:58:21] is they're also treated asymmetrically as  
[01:58:23] a source of power. As I told you in their  
[01:58:25] framework, when they talk about power,  
[01:58:27] they're really thinking only about the stick, only  
[01:58:29] about the tariff. They're saying,  
[01:58:31] what does it mean for me to be powerful  
[01:58:33] vis-à-vis any other country?  
[01:58:35] Well, if I push you to outer key,  
[01:58:37] what do you lose?  
[01:58:39] But you could take a broader definition  
[01:58:41] of power. If you look in the dictionary,  
[01:58:43] it tells you, you know, I have the control  
[01:58:45] or the ability to influence others.  
[01:58:47] I would add also at low  
[01:58:49] or reasonable cost, because  
[01:58:51] if I can influence you, but I need to  
[01:58:53] shoot myself to do that, well, maybe  
[01:58:55] I don't have that much power. If I can do  
[01:58:57] that, maybe I have more power.  
[01:58:59] So you could think transfers  
[01:59:01] are also a source of power. And I don't know  
[01:59:03] why you don't think  
[01:59:05] about them this way. So imagine I have two  
[01:59:07] superpowers. One is wealthier than the other one.  
[01:59:09] One is more constrained than the other one.  
[01:59:11] Presumably this happened during the Cold War.  
[01:59:13] Towards the end, the USSR  
[01:59:15] was suffering.  
[01:59:17] Well, it will be cheaper  
[01:59:19] for one superpower to make big transfers  
[01:59:21] relative to the other one. And so  
[01:59:23] you could argue it also has more power  
[01:59:25] even if the size of the stick were the same.  
[01:59:27] Okay.  
[01:59:29] And the last one  
[01:59:31] is the way they're introduced  
[01:59:33] in the framework.  
[01:59:35] Sticks and carrots also affect things  
[01:59:37] differently. Sticks affect incentives.  
[01:59:39] So I'm deciding should I  
[01:59:41] align or not? Well, I could be punished.  
[01:59:43] But carrots affect directly the preferences  
[01:59:45] the way they model it, because this bliss point,  
[01:59:47] that's my reference point I don't  
[01:59:49] want to be too far from, that's directly affected  
[01:59:51] by the carrot.  
[01:59:53] And that I found also a bit not very natural.  
[01:59:55] I'm not sure why you wouldn't just model  
[01:59:57] you know, I have my own preferences and both  
[01:59:59] sticks and carrots affect my choice.  
[02:00:01] Okay. So this doesn't matter much  
[02:00:03] for what they do empirically, but I'm just saying  
[02:00:05] as a conceptual organizing framework  
[02:00:07] I'm not sure sticks and carrots  
[02:00:09] should be treated so differently  
[02:00:11] when I think of this world.  
[02:00:13] Okay.  
[02:00:15] Now let me go to their  
[02:00:17] two of their main results.  
[02:00:19] The first one is that  
[02:00:21] the alignment changes with the sticks and the carrots  
[02:00:23] as I told you, but I want to focus here on the stick.  
[02:00:25] So the one  
[02:00:27] main finding is that when my bilateral  
[02:00:29] gains from trade vis-a-vis the edgaman  
[02:00:31] are high, well I tend to align  
[02:00:33] with the edgaman. The interpretation  
[02:00:35] of the paper is that this is like  
[02:00:37] brute force power.  
[02:00:39] I will sanction you somehow  
[02:00:41] unless you align  
[02:00:43] with me. Okay. In their framework  
[02:00:45] I guess these sanctions  
[02:00:47] never happened, they're all of equilibrium  
[02:00:49] but it's the shadow of the sanction  
[02:00:51] what's leading me to align.  
[02:00:53] Now that's one possible interpretation  
[02:00:55] but there's an alternative mechanism  
[02:00:57] which is, well, alignment  
[02:00:59] is a tool to reap part of the gains  
[02:01:01] from interaction or gains from trade.  
[02:01:03] So for instance we have this  
[02:01:05] paper with Christoph and Fernando  
[02:01:07] Brone and Josephine Meyer where  
[02:01:09] one of the forces in the  
[02:01:11] in the paper is that the more I align  
[02:01:13] with other countries, the more  
[02:01:15] our gains from interaction. You could think including trade.  
[02:01:17] And that's a world  
[02:01:19] where if my gains from trade  
[02:01:21] vis-a-vis the edgaman are very high,  
[02:01:23] naturally I will tend to gravitate to the  
[02:01:25] edgaman and I will tend to align  
[02:01:27] because that will reduce the  
[02:01:29] cross-border transaction  
[02:01:31] costs between us.  
[02:01:33] So you could think in this paper what they're finding  
[02:01:35] you could think of this even voting as  
[02:01:37] imperfect measure of alignment and so  
[02:01:39] what's happening is when our gains from trade are  
[02:01:41] large, we tend to align.  
[02:01:43] Now is this a stick in the background  
[02:01:45] or is this what I call here kind of soft  
[02:01:47] power in the sense that I just want to align  
[02:01:49] with countries with whom I  
[02:01:51] have large potential gains of  
[02:01:53] interacting. So once again this doesn't  
[02:01:55] question your result, it's just  
[02:01:57] a matter of interpretation.  
[02:01:59] Okay, the other cool  
[02:02:01] result they have in the paper is that  
[02:02:03] edgaman must value this  
[02:02:05] alignment a lot. Okay, in fact  
[02:02:07] they get that, one  
[02:02:09] of the figures they report in the paper is that  
[02:02:11] the US would be willing to spend 3%  
[02:02:13] of its GDP, so a huge sum  
[02:02:15] to shift Egypt from being  
[02:02:17] USSR aligned to being aligned with  
[02:02:19] the US. Now where does  
[02:02:21] this come from in the paper? Well, Tim explained  
[02:02:23] it kind of. The reason  
[02:02:25] where this comes from is that alignment  
[02:02:27] you know there's a significant  
[02:02:29] statistically significant correlation  
[02:02:31] between alignment and carrots but the  
[02:02:33] response is actually quite weak  
[02:02:35] and so the model tells you if I'm  
[02:02:37] spending all this money to move the alignment of Egypt  
[02:02:39] and Egypt is moving the U.N. voting kind  
[02:02:41] of very little it must be that I value  
[02:02:43] that a lot. Okay, and that's  
[02:02:46] where this number comes out of intuitively.  
[02:02:49] Now an alternative interpretation  
[02:02:51] is that of course U.N. voting is a very noisy  
[02:02:54] or not maybe the most relevant  
[02:02:57] part of alignment and actually  
[02:02:59] what I'm getting when I give money to Egypt is other things  
[02:03:02] military alliance, intelligence,  
[02:03:04] military bases, etc.  
[02:03:06] So U.N. voting is just a very  
[02:03:09] noisy measure of this and I'm capturing  
[02:03:12] only a small part of what I'm getting. So it could be that that valuation is actually quite lower  
[02:03:15] but I'm inflating it just by focusing on  
[02:03:18] U.N. voting and to some extent I think this  
[02:03:21] must be true because imagine a world  
[02:03:23] where I really all I cared about was U.N. voting.  
[02:03:26] Well in this world I don't care whether it's Egypt vote  
[02:03:29] or Fiji's vote because one vote is one vote  
[02:03:32] and so in that world I'm not sure it would make sense to think that some  
[02:03:35] countries are more strategic than others.  
[02:03:38] Okay and in fact since I have one minute  
[02:03:41] this brings me to my last point which you know I remembered  
[02:03:44] this morning which fits very well  
[02:03:47] with the previous paper which is the following.  
[02:03:50] Here everything is bilateral so I give money to Egypt  
[02:03:53] because I value Egypt's alignment and Egypt's  
[02:03:56] alignment is independent of Libya's alignment  
[02:03:59] and of Algeria's alignment so another measurement  
[02:04:02] thing that may be lurking here is that actually by moving Egypt  
[02:04:05] I'm affecting countries close to Egypt and that might also affect  
[02:04:08] how I back out this valuation.  
[02:04:11] Okay so bottom line  
[02:04:14] is you know I found it to be a very  
[02:04:17] interesting paper a great paper I think  
[02:04:20] makes a very compelling case once again that these sticks and carrots  
[02:04:23] matter that they moved alignment and as I told you here  
[02:04:26] you know one thing you know a few comments  
[02:04:29] is how would you think about these carrots and sticks in a more symmetric fashion  
[02:04:32] and how do you really think about interpreting these results  
[02:04:35] these empirical results. Thank you.  

## Q&A (02:04:38 – 02:16:19)

[02:04:38] Thanks so much Tim. Want to collect some questions here?  
[02:04:46] Steve? Let's go.  
[02:04:48] Kristin and Jaya.  
[02:07:38] Okay thanks also Alberto and to all the questions  
[02:07:40] that you have for me.  
[02:07:42] Thank you.  
[02:07:44] Okay.  
[02:07:46] Thanks also Alberto and to all the questions  
[02:07:48] that you have for me.  
[02:07:50] Thank you.  
[02:07:52] Okay, thank you.  
[02:07:54] Thank you.  
[02:07:56] Thank you very much.  
[02:07:58] Thank you.  
[02:08:00] Thank you.  
[02:08:02] Thank you.  
[02:08:04] Okay thanks also Alberto and to all the questions  
[02:08:08] I'll respond to first to some questions on on the aid  
[02:08:12] and also sort of to bit to Alberto's discussion.  
[02:08:15] So I have lots lots of sleep on how to model things and there are some asymmetries  
[02:08:19] which help us get this to the data in a very elegant way.  
[02:08:23] So if you do this like perfectly symmetric you have like a function  
[02:08:27] where if you move here you get money if you don't you don't get money.  
[02:08:30] What you end up having is in equilibrium only one side page  
[02:08:33] only one hegemon pays the country.  
[02:08:35] This is sort of like there's some papers which model like this  
[02:08:38] but it's not what we see in the data so that's what we went with in the end  
[02:08:42] and sort of it can be micro-founded different ways.  
[02:08:45] You can think about like we have this in the paper where you have multiple votes  
[02:08:48] you have zero or one you bid on every vote  
[02:08:51] and then these countries sometimes with the other one Americans  
[02:08:54] and sometimes with the Soviets.  
[02:08:56] On Steve's questions on sort of arms sales.  
[02:09:00] So these are in there and so far as they're in the green book  
[02:09:04] which would be a subsidized arms sale in some sense  
[02:09:07] which sort of you know there's these aid flows are a bit tricky  
[02:09:10] because you know there's not like little transfers  
[02:09:13] but they should be in there.  
[02:09:15] Effect on aid and growth it's well taken.  
[02:09:18] I mean if you just look at the allocation of foreign aid  
[02:09:20] if you would think this would support growth  
[02:09:22] it would all go to the poorest countries  
[02:09:24] but it looks very different  
[02:09:27] and similar to sort of the...  
[02:09:30] okay...  
[02:09:32] Kristen so China and so we've sort of  
[02:09:37] there's some data trickling in right now  
[02:09:39] so I was talking so there's a paper by Andreas Fuchs  
[02:09:41] Andreas Pespitero and co-authors  
[02:09:43] which looks at precisely this questions  
[02:09:45] so they only see a small slice of Chinese foreign aid  
[02:09:48] what they see is very little response  
[02:09:50] so they find like mostly which you know  
[02:09:53] one way of interpreting this is China doesn't care that much  
[02:09:55] about alignment but cares about you know other things  
[02:09:58] for example and other tools that restructuring  
[02:10:01] so it's a good suggestion we should have that  
[02:10:04] there's like a never ending toolbox of IMF loans  
[02:10:07] and you know you can put more and more into this reaction function  
[02:10:10] we have specifications where for example control  
[02:10:13] for lending by the IMF which gets a bit at this  
[02:10:16] but yeah we should get these other forms of making the transfers as well  
[02:10:19] and finally Jaya so yeah so yeah  
[02:10:24] if the other hedge amount responds very strongly  
[02:10:27] to the what the US and the Soviets are doing  
[02:10:30] you know this would attenuate our facts because we would see  
[02:10:33] oh they're not doing very much but really it's because the Chinese are pushing back  
[02:10:36] or the Soviets are pushing back so strongly  
[02:10:38] so it's been a while since I've run these regressions  
[02:10:41] but at least when you correlate just you know how much  
[02:10:44] like how where is the US spending with the Soviets  
[02:10:47] spending how correlated are they they are somewhat correlated  
[02:10:49] but not that much so it's not that we do you see one country  
[02:10:52] goes in the other country necessarily goes in super strongly  
[02:10:55] could be that they have lots of other things that they care about  
[02:10:58] other than alignment and should I go more questions  
[02:11:02] Giovanni?  
[02:11:25] Max?  
[02:11:54] Lena?  
[02:11:59] There's a lot I think of talking around about what it means  
[02:12:02] to vote similarly in the UN General Assembly  
[02:12:05] and many of us use this measure but if we're also frustrated  
[02:12:08] we may be not capturing the capturing alignment  
[02:12:11] and sort of costly actions in our life  
[02:12:14] there are some papers by Dreyer and Collins where they look at  
[02:12:18] what they call force trading in the UN Security Council  
[02:12:20] and they exploit right so the temporary rotating memberships  
[02:12:23] to see whether or not they're aligned to things like born age  
[02:12:26] but also world by bones they actually find the world by projects  
[02:12:29] that are out of the team but can for a member  
[02:12:31] have worse than their project ratings right  
[02:12:34] than non-personality one place so it could be  
[02:12:37] sticking with the UN and thinking more about security council voting  
[02:12:40] it may be costly  
[02:12:41] okay  
[02:12:47] Very quick one related to Lana's point  
[02:12:50] you're putting other variables on the left hand side as well  
[02:12:53] there's a bunch of new stuff there's that cool new measure  
[02:12:56] Bobby Tattoo Fun and John McGowan  
[02:12:59] the other thing we just saw state visits  
[02:13:03] in Ernest and Steve's paper I thought that was really cool  
[02:13:06] you could also look at military basing  
[02:13:08] I think that putting more things on the left hand side  
[02:13:11] variable separation probably  
[02:13:14] but that could be really compelling  
[02:13:19] Okay thanks a lot  
[02:13:21] so I know the paper I should  
[02:13:24] we're reading into it to see if we can take something from that  
[02:13:28] there's lots of questions on how do we think of alignment more generally  
[02:13:32] so do we think of this like being  
[02:13:35] you care about the change or maybe you just want some people to be in your camp  
[02:13:38] I mean it's very hard to tell this from  
[02:13:41] the scarce data that we have  
[02:13:43] we've sort of run sort of non-linearities like do you like countries  
[02:13:47] do you not care if they're like 50 or 55 but then if you really care about  
[02:13:51] being making them 90% aligned  
[02:13:53] you don't find that much but you find a bit of that  
[02:13:58] similarly for these other measures  
[02:14:02] we're working on incorporating them  
[02:14:04] the challenge we face is this like super data scarce  
[02:14:07] so like even in state visits the US is not undertaking many state visits to Zambia  
[02:14:12] in the 1950s or there's not many newspaper articles discussing  
[02:14:15] how the US thinks about Zambia at that point in time let alone the Soviet Union  
[02:14:20] so that's sort of the challenge we face is we need something continuous  
[02:14:23] which covers really many countries and which does not much data  
[02:14:27] and finally the UN Security Council  
[02:14:29] so we've thought a bit about how to interpret these papers  
[02:14:32] the lens of what we're doing  
[02:14:34] so of course these papers find that when you get onto the Security Council  
[02:14:38] there's this really large effect on how much aid you get  
[02:14:41] the way I think of it is consistent if you think of  
[02:14:44] it's very expensive to buy countries  
[02:14:46] you would think this is like a marginal shift in how geopolitically important you are  
[02:14:50] but you see this massive effect on aid flows  
[02:14:52] I think that's sort of consistent with the story we're sort of trying to tell  
[02:14:56] but yeah, we've thought about how do we sort of discuss this evidence in our paper  
[02:15:01] so thanks a lot  
[02:15:42] yeah, we should look  
[02:15:46] so we have some specifications where we control for realized sanctions  
[02:15:50] so that that would be in there  
[02:15:52] and so we find that these measures have predictive power over and above  
[02:15:56] when you actually exercise the stick  
[02:15:58] of course in this model you never do  
[02:16:00] but yeah, the data is very scarce  
[02:16:02] so it's kind of hard to get this sector level  
[02:16:05] you would ideally do something more sophisticated  
[02:16:08] but we sort of size of total trade flows  
[02:16:11] of what you can get for many countries  
[02:16:13] relatively straightforwardly  
[02:16:15] thank you so much and we're back to 11.30  
[02:16:17] alright, we're going to start in a second here