Technology Day 2022: A Celebration of MIT Economics – Dave Donaldson
Whitney Espich: Hi, I'm Whitney Espich, the CEO of the MIT Alumni Association and I hope you enjoy this digital production created for alumni and friends like you. So I know some of you are here today for your 50th. 50 year reunion. Others for your five year reunion, perhaps, and I think something we could all agree on regardless of our generation and exactly what that means is that we've lived through an unprecedented scope of globalization in recent times. And so it doesn't surprise us anymore to see pictures like this one where containers are lined up to receive the goods that this country is sending to the wider world. And the container the goods of this country is receiving from that wider world.
We're used to seeing pictures like this one, of the recently expanded Panama Canal, often called one of history's greatest engineering achievements. And we're used to seeing pictures like this one that remind us of the role that international trade and the things that it brings into our economy play in our daily lives. But of course, recent times have reminded us that many things disrupt trade as well. Globalization grows and also shrinks, sometimes there are blockages that get in the way. And sometimes, of course, there are actual blockades, like this one where to this day the Port of Odessa, which exports food grains that feed the mouths of millions of hungry people, poor people around the world, is currently being blocked.
So I began my studies as a student interested in physics. And I saw stories like these about the role that trade plays in our lives. And I have to admit they are kind of what pulled me into economics. And I have spent my time since trying to figure things out, and I'm going to tell you today a bit about some of that. I'm going to draw in particular on a kind of recently accepted paper that I've written with some colleagues, in particular with Arnaud Costinot, who's a fellow Professor here in the Economics Department. And we're going to kind of try to walk you through the way that economists, like the analysis I'll be showing you, sort of think about the impact of trade on people's livelihoods.
And I'm going to start with basics and basics would, in MIT speak, get called 1401. Some fans in the room, that's good. So principles of microeconomics, some of you may remember it fondly, and others may not remember anything. But I will try to remind you of some of the basics. The basics look like this.
There's a producer, and there's a consumer. The producer is a factory, the consumer is a dinner plate, and that's kind of where 1401 takes off and elaborates a great length about this kind of simple exchange between a producer and a consumer. 1402, maybe fewer fans in the room, but nevertheless. 1402, principles of macroeconomics, kind of reminds us about wider connections between firms and the economy. In particular would stress that for no firm stands alone.
A firm relies on the human and other resources that it draws on, and in particular it draws on the skills that some people provide. It draws on the ideas that maybe the entrepreneur or the innovator provides. It draws on the capital, the financing that investors provide without which the firm probably would be nowhere.
And so those people are involved in production and hence delivery to consumers via the fact that they provide the inputs that go into the firms, of course. So what about trade well that in MIT speak is called 1454, introduction to international trade, our undergraduate class in the field. And of course, this invites us to consider two countries, maybe one that you care most about because it's where you live or something, the United States, and a wider world, with a border between them. And of course, the interactions that get studied in 1454 are pretty simple. There's sort of one we start with which is importing.
So there's a foreign factory that provides food and other bounty for our domestic consumer. And of course, there's exporting too, and we could illustrate that this, where a firm or a farm in our country exports something directly to foreign consumers. And that's sort of the basics of 1454. So when we-- and of course, not to be forgotten, people are behind that farm just as much as they're behind a firm that makes things for the domestic consumer. OK, so when we turn to the post-graduate version of this course, we often stress kind of wider mechanisms that have the same roots.
So here I would talk about this phenomenon maybe most importantly, which is where a foreign factory provides inputs for a domestic producer, like this oil rig, which then sells things to the domestic consumer. So this is kind of a critical piece of any modern economy where much of what we import doesn't end up directly on our dinner plate, but it helps our firms to make things for us, the domestic consumer. Another thing we talk about is this one, which I-- is a little bit more nuanced, but it's where a domestic firm on the right of this picture sells something to another domestic firm, and that second domestic firm is the one that exports. So in some sense this first firm is not exporting. All it does is sell domestically.
But it's just as much of an exporter as the farm here. Everything it does is it sells to the farm, and then the farm does the actual shipping across the border. But without exporting, the first firm would be nothing in this picture. OK, so taking this further we can also talk, think about indirect forms of importing, the one shown here. This is where a foreign firm sells something to one domestic firm and then that second domestic firm sells something to another domestic firm, which then sells it to domestic consumers.
So there's a sort of intermediate firm here that's not directly importing anything, but they themselves would be nothing without the fact that a fellow domestic firm is the one doing the importing. And we call that indirect importing. Finally, as you can imagine as in fact, even features on some Seinfeld episodes, some firms do both. They import and they export. And that's what this last firm is doing over here. They're just importing something to then send it back to foreigners in a transformed way.
So this is a quick kind of summary of the way we think about the way trade can impact people. The people at the top, inside your domestic economy. And just to summarize all that, there are people in green who are directly involved in exporting.
There are people in light green who are indirectly exporting. There are people in orange who are directly importing. There are people in light orange, who are indirectly importing, and there are people in blue who are doing a bit of both. Just like in the real world. Many people are engaged in both. So this picture, of course, is embarrassingly oversimplified.
The real world would have more links. We can imagine the firms do multiple things. They don't just sell domestically, they also export. Some of these links are showing up as I talk.
Sometimes firms even sell and back and forth to the same firm, and goods go in cycles. We see that in our data. But the links can compound, and can complexify the way we think about things. Until eventually we end up something that maybe looks like this, that might be a snapshot of the way a real economy works. And the red lines would summarize all the interconnections in a networked economy. So this is the way economists think about trade.
The major problem with the way we think about trade is that there is no data on these red links. The way the data comes to us as economists is far more aggregated and far more coarse than our ability to actually see these red links. But what I'm going to talk about now is kind of a burgeoning attempt by economists, myself and my colleagues included, to try to start to chip away at that gap in our data knowledge. And this is going to draw in particular on one country where we've just kind of figured out how to make this work, and now we're kind of eager to scale it up to other countries where we can do it, too. And I'm going to be doing this in Ecuador, a country whose tax system and tax transparency works in a certain way that allows researchers to fill in all those red gaps that I talked about earlier. So Ecuador is going to be trading with the rest of the world.
Lots. All the other countries in the rest of the world, in this example. And basically the way we see the Ecuadorian economy starts with seeing kind of each of roughly 3 million individuals who work at Ecuador's firms, and half a million individuals who own Ecuador's firms, or at least a share thereof. We can then connect those individuals to the firms, at 1.5 businesses that are in our Ecuadorean tax data.
1.5 million businesses that are in our Ecuadorian tax data. And we can see how much those businesses trade with each other, domestically. The crucial part of a modern interlinked network economy. And then of course, we can actually connect those firms to the domestic consumers, to the foreign consumers. That's the exporting. And back and forth, importing and exporting with foreign firms.
So I'm excited about the fact that we can finally do this. It took many years of hard work to kind of get access to the data, to be able to put it together, to be able to make sense of it. But we can actually now start to kind of learn some lessons from it.
So I'll start with some basic lessons, just some fairly descriptive lessons. I'm going to start with what I'm going to call export participation. And so in principle, we could take any one of those 2.9 million people, whether there are a worker or a business owner or a bit of both, and tell you how much that person is participating in exporting. Of course, I don't have-- I don't know people's names or anything. It's all anonymized.
But far more interesting is to project all the people onto some common characteristic, and the characteristic I'll focus on today is the income distribution. So consider that the x-axis here. Percentiles of the income distribution. And I'll plot on the y-axis the extent of export participation, which you can interpret as the share of people's income that they derive from exporting, both directly and indirectly. So it looks like this.
With the red-- the X's being the raw data at the percentile level, the red line being a particular sense of smoothing across those 100 data points. And I have to be honest. This completely shocked me.
Why? Well, because what I expected was that it would be upward sloping. That was just my prior, my guess based on existing knowledge and my own understanding of the way the world worked. What would upward sloping mean? It would mean that Ecuador's relatively rich people are the ones who are most export engaged. Most export participating. In contrast, what actually we learned from the data is that Ecuador is kind of middle class, that are the ones who are most involved, most participating in exporting once you actually get the data on both direct and indirect export linkages throughout the Ecuadorian economy.
The second thing we can talk about is import connections. Here I'm going to stress what I'm going to call import exposure, which is actually going to be interpreted as going to be defined in a way that we think it's actually economically bad or relatively bad to be import exposed. This is sort of-- think about this as you might work in a firm that sort of directly competes with foreign imports.
So that sense-- that form of import exposure would be bad. Some of these import exposures, however, are good. That would be where you work at a firm that is importing a part that is kind of complementary to you as a human being providing human capital or capital to that firm.
So this is going to be the net of all those things, but it's defined in such a way that kind of having high import exposure is bad. And this one looks like this. So this one surprised me, too.
But I can tell you what it means. It means that Ecuador is poorest are the ones who are most import exposed. Ecuador's richest are the ones who are least import exposed, in this sense. So that in essence means that while exporting was pro middle class, importing is actually relatively pro-poor. So pro rich in this example, because the harm done by import exposure is less bad for the rich.
OK, so given that I've sort of shown you some rough snapshots of how this looks, just the kind of first time we've been able to get access to data like these and actually see what it looks like in a real economy, the next question is, OK, well what would be the impact of changing the amount of trade that Ecuador does? Those were just snapshots based on the amount of trade it does. So what we're going to do is kind of simulate, according to our analysis, according to our model, our understanding of the way the economy works, what life would be like if you took Ecuador sort of from no trade, to the amount of trade they're currently doing. So let's call that a globalization impact. They're going from lack of trade to a lot of trade. And when I talk about such impacts I'm going to kind of sum-- it's just a simple sum-- two forms of such impact.
The impact due to export participation that I've already shown you, and the impact due to import exposure that I've already shown you. And we'll just add them up. And it looks like this. Going to focus for a second on what I'm income-- sort of think about it as the income you earn as employees.
So this kind of omits business profit earning for all people. It only counts their employee earnings. But what this basically shows us is that sort of export effect that was relatively pro middle class is overshadowed, overweighed by the importing effect that was relatively pro rich. And the net effect is by and large pro rich. When you add capital earnings to that, which is shown here in blue, it gets even stronger. The no surprise, the Equadorians that are the richest are the ones that tend to own Ecuador's businesses.
According to our analysis the effect on those businesses, those business profits, is very much positively affected on net by the access to trade, the access to exporting, et cetera. And so the gap between the red and blue tells you the sort of differential effect that's enjoyed by people who own businesses and provide capital to businesses, but it doesn't really overturn the story. The net picture is that trading and the impact of trading in a country like Ecuador could be good for everyone. But it would be particularly good for the rich, relatively good for the rich, by the tune of 10% more than the median income earner. So that's kind of the main lesson that we've extracted from these data so far. The impact of a simulation like that.
There's a lot more that we have done and are interested in doing. Just give you a kind of very brief snapshot of some of that. So the first thing we obviously want to do is ask how this generalizes. We've got the numbers we believe in for Ecuador, but what about the rest of the world? And so there's about 10 countries nowadays by our estimates for which the data exists that one could kind of do exactly this kind of right now.
And we're in the process of trying to do that along with collaborators around the world. The next thing I'd add is our wish list, our to do list is incorporating other sorts of flows. I've stressed international flows and involve the exchange of goods and services. But we know that countries interact in lots of other ways. So they, of course, interact via people who migrate, either temporarily or permanently. By the flow of capital and other kind of know how that we think the foreign firms, when they invest in a foreign-- and domestic firm invest in a foreign country they might bring expertise and know-how that spills over to the wider recipient country, and that form of interaction or flow is missing here, too.
The third thing we'd like to do is kind apply the same data, the same analysis, the same machinery to kind of new phenomena. And here I could just start with, for example, the phenomena that my colleague Nancy was just talking about, the study of competition. The study of how firms are competing and whether that's a net good for efficiency involves the interconnectedness of all the firms in the economy. And that's exactly the kind of thing that economists are now able to start to incorporate. And we in fact, Nancy I we're talking about some students that are trying to sort of take that baton and run with it.
The final thing I'd say is kind of new sources of data. So Ecuador's analysis and the analysis for other countries around the world that's possible due to their tax systems, we think that generalizes. Just to get the basic idea, it works on the fact that these countries have a value added tax. And so all transactions are taxed when they take place.
So whether they're from firm to firm or firm to consumer. And so that allows a tax agency to sort of record them, and share them with researchers. But there are, of course, lots of other ways that economists and other analysts could start to understand the network economy. There are sort of digital payment systems, or just good old fashioned credit cards and bank accounts that allow us to understand who from the consumer side is interacting with which firms via the purchases they make.
There are data from the tracking of physical goods, the movement of goods, the internet of things that we're optimistic might be enabled to study this sort of thing. Satellite imagery is allowing us to see the movement of container ships in ways that was never possible before. I'm sure there's a gazillion other ideas for new sources of data, too, that we have yet to learn about, but we're kind of eager to learn. And that really brings me to this.
We tend to sort of have this help-wanted view of research in our department, that whether it's insights on countries where policies or policymakers need lessons from analysis like this, or interested in cooperating with researchers on their data to enable better science and better policymaking. Whether it's the insights about new types of flows, new phenomena that you think are important in the world economy, or even new sources of data from the kinds of private firms and government agencies that you know of, we're eager to kind of hear about all that and get your thoughts. So I just end with my email address. Please contact me if you would like to. I cherish the opportunity.
And my website there is the link to the paper where you can download the analysis that I talked about today. Thank you. I'll stop there.
[APPLAUSE] What would you like [INAUDIBLE]? Yes. So Dave, thank you very much. I was impressed when I hear that amazing amount of work you've done there. And so can I ask, I'm gonna start with you put together those many sources of data in Ecuador to get a view of impact of trade in the Ecuadorian economy. You start talking about more countries.
Is that something that would be feasible to do for a country like the United States? Or what are the sort of obstacles that have led you to work in the Ecuadorian context? So unfortunately, right now in the United States there is no source of data that one could imagine as a census, or the universe of all of these flows. The fundamental reason is that the US doesn't have a value added tax. 75% of countries in the world do. And so I'm optimistic that the US might be an outlier for which we don't-- the hope of having exactly that right tax data that would help us do this is not yet on the table. However, the US does-- the Census Bureau does conduct a survey of flows of goods and services throughout the economy that doesn't give us this kind of universe view, but with the right approach to the math and statistics that I think some people still have to work out, like could be the right way to sort of treat it as a survey, a sample, a sub sample with which to inform the right picture of the universe network.
So I'm optimistic. But we're not there yet. Yeah.
OK. So another question I got I got a couple of questions related to tariffs and tariff policies as a way to protect workers. So does your work on Ecuador let you sort of say something about what would happen if Ecuador tried to use tariffs to protect its workers? And are you willing to draw any lessons for thinking about the role of tariffs in protecting American workers? Absolutely. So Ecuador has tariffs in place.
We in fact in the paper study some of those tariffs. The analysis I showed you here could be thought of as kind of imposing like a prohibitive tariff, an extreme tariff on everything. Or the flip of that. But yeah, so the basic way I think about it is that I think let's start by thinking about the import direction. When we impose a tariff on a good that is directly coming into our consumer markets, that is kind of a tariff on import competition. That's protectionism, the classic example.
The could protect workers. But the danger that happens is that we tax those kind of goods that are coming in to help our firms to produce what they make. In my example, that was the foreign firm that's making like a part that my oil rig really needs, you know? And when we put tariffs on those things, we of course probably-- we certainly harm our firms. We probably harm on average the workers and the owners of those firms too. Countries tend not to tax their exports. The US typically doesn't do that, but equally no country can export without importing.
We know that's a basic lesson of 1402, so whenever a country is taxing its imports, they're always taxing exports too. So we're always in that grander sense kind of whenever we're sort of protecting some workers on the import side, if that's our goal, we're always harming some workers on the export side, too. And that's something I would sort of issue as a major caveat to thinking about protecting workers. And sticking with the Ecuador topic for a second. You said you were surprised or your expectations were that it wasn't going to be the low income workers who gain more from the imports in Ecuador and vise versa.
And then that net effects went the way they did. After having seen the results and having looked through the data, do you have intuition for why the winners and losers in Ecuador are who they are? Yeah, we think that a very important phenomenon is that thing I just stressed, which is sort of the foreign firm, foreign economy, supplying let's say parts for the domestic economy that make domestic firms more productive, or more competitive. And it just so happens that Ecuador's rich people tend to own and work at those companies that are doing that importing of parts. And when you think that mechanism tends to trump all else. But the other surprise I mentioned, however, was on the export side.
I had this image that it's the big firms, it's the superstar firms that are able to successfully export to and kind of compete successfully in foreign markets. And as I had this prior that it's the rich who work at and own those kind of superstar firms, and that's why I expected the export effect to be so pro rich, whereas I was surprised. It's not. That's not the way things work in Ecuador, evidently. OK, so can I ask a question? Do you see in light of recent events, we've had a lot of talk about globalization in the last decade or several decades. Do you see the world potentially de-globalizing? And if the world does de-globalizing, any thoughts on what the effects of that would be? I think on net, the force for globalization has been largely technological.
Not so policy driven. The digital exchange of stuff, not just goods and services, but ideas and culture is immense and probably unstoppable, is my own prediction. And so I don't see much that could revert that tide. I could see changes along the way. I think maybe rightly many countries are waking up to a heightened realization of risk, that their all eggs in one basket kind of view of how we source those critical foreign parts is starting to get challenged, and that's probably on large a good thing for individual firms to recognize, that they need to diversify their supply chains in the height of risk they didn't expect. And for individual countries, especially small countries, possibly to equally have that diversified view.
But that doesn't have to be de-globalizing. Diversification could involve more foreign production, just more different types of foreign sourcing, as well as potentially more domestic sourcing. OK.
In terms of work you've done, any insights on or could you do similar work to think about effects of carbon exporting and importing, and the effects of energy and environmental effects of this importing and exporting? Yeah, that's-- I should have. That's a great idea. I should have stressed that. So the way that current proposals for carbon taxes, carbon markets, even just consumers knowing how carbon intensive is their consumption footprint rely on the kind of network view that I sketched. So you would want to know of what's ending up on my dinner plate, how much carbon is in it, and that depends on all those kind of indirect and indirect and indirect possibly cross-border ways the carbon or whatever else that we're trying to avoid.
It could be ethical supply chains that we might be interested in. It's exactly this network view that would allow one to do that. And so for example there are researchers in Brazil that are trying to make it easy for consumers to understand when you're buying beef that has a particular thing, a particular number on the back of its packet, a package of beef jerky or something, that you kind of know what slaughterhouse it was sourced from, and whether that sort of slaughterhouse tends to source its beef from land that was illegally deforested. And so that gives kind of-- that supply chain visibility is just another example of the type of thing that this connects to.
OK. I guess I actually have a few educational questions. I should say people probably don't know this. I don't there's the bio that Dave is actually currently also the head of our undergraduate economics program.
But so I had some questions that are of the form, what is going on right now at MIT? Can you say anything going on at MIT right now to sort of get insights on international trade and globalization to the potential entrepreneurs in course six and elsewhere, and where people are learning about this kind of material? Right, so the first thing I'd mentioned is the new major. Joint major we've created with core six. Many of you already know about that.
But it's called 614, it's a joint major program that's about economics, computer science, and data science, and it is very much in its genesis about enabling students, MIT graduates to understand how to think about policy and business problems with the help of huge data sets, and with the help of mathematical and statistical tools. And an example of which would be this. And so yeah, one of the class I mentioned, 1454, is part of that joint major program. And one final question.
I sort of emphasized earlier, it started my introductory remarks, the sort of credibility revolution and how we emphasize causality in economics. Can you give us any sense of you're showing this sort of your affects of international trade on people at different parts of the income distribution. So what is it that you're doing, or how is it that you're trying to get at what are causal effects of international trade on people. Can you? Yeah, it's very important question. So our leverage on that is that we have six years of data.
And everything I showed you with snapshots from the middle of year, but we have six years of data. And Ecuador had a stream of foreign events that shocked, in some sense, the extent to which different products, different firms in Ecuador are able to sort of export and compete with imports. And so we use kind of what we hope are kind of as good as random kind of foreign determine components of those events. Foreign countries changing their tariffs. Foreign countries exchange rates, exchange rates changing unexpectedly. Foreign countries going through recessions or booms.
Commodity discoveries, things like that. And we sort of look at as the foreign shocks go up and down, we sort of actually try to sort of trace out what's the impact of that inside kind of firms in Ecuador, inside supply chains in Ecuador, and inside kind of connections between workers and owners and firms in Ecuador. So that's kind of our window. It's not a perfect window.
It's just this kind of six years. And admittedly the variation that takes place in those six years is smaller than the variation that I showed you, which is like a dramatic reduction in trade. But still, we think it's critical to kind of discipline the device by so that we make sure that our analysis is kind of estimated so that everything in the model is calibrated exactly to match how the economy, the real economy, responded to those actual trade shocks that happened in sample. It's very good. And so I guess I take it that sometimes maybe there the fact that Ecuador is a relatively small country helps you in having those sort of small shocks be exogenous to Ecuador.
I think that's a plausible view, yeah. Yeah. There are-- yeah. I don't think it's hopeless for large countries to find such events. Obviously this is a great segue to the work of our colleague Josh, who really pioneered the way economists think about those problems in general. And so they have been applied to very micro problems, and also increasingly to very macro problems, like this one.
The impact of globalization. And it's all due to the work of people like him. OK. Thank you very, very much. Thank you. [APPLAUSE] Whitney Espich: Thanks for joining us and for more information on how to connect with the MIT Alumni Association please visit our website.
2022-06-24 06:13
