This activity comes after students are to have listened to a lecture (slides) on political economy, a considerable focus of which was on identifying the winners and losers from economic cooperation.

In that lecture, I argued that trade tends to enrich the owners of, employees of, and investors in sectors in which that country has a comparative advantage, while it harms the owners of, employees of, and investors in all other sectors. One implication of this is that a shift from relatively low to high levels of global trade is that a small number of people in highly developed states should benefit while most workers in the developed world will be harmed, while the majority of the developing world should benefit.

That’s undeniably a simplification of a very complex phenomenon, and there are many other factors at work, but to a first approximation, this is clear prediction of a lot of work in political economy, and while there’s more debate among economists about the impact of economic globalization than I discussed with my students, there’s some evidence that this is indeed what has happened. The following graph (from a paper by Branko Milanovic which has gotten a lot of attention), shows the percentage change in income from 1988 to 2008 across the global income scale.

Of course, the graph doesn’t tell us who lives in what country, and it can’t tell us how much of what we’re seeing is due to trade as opposed to other factors, but given that only the poorest individuals in wealthy countries have incomes that are below those of the top income brackets in poor countries, it’s pretty reasonable to interpret this graph as telling us that the highest earners in the highest income countries did very well in that twenty year period, the rest of the population in high income countries did not do too well, but a very substantial portion of the rest of the world did quite well. The poorest of the poor saw no gain, but from the 5th percentile to the 65th percentile, huge progress was made. The story of globalization is not one of the rich getting richer while the poor get poorer, but of the richest in the rich world conspiring with most of the developing world against the middle and working classes of the developed world. We have seen within country inequality expand while inequality between countries has declined. Or, to really put it in terms students can understand, the American middle class was hit hard, and cities across the Rust Belt (which includes Buffalo, where I teach) were absolutely devastated, but the forces responsible for this did not just enrich mega-corporations in the US—they also lifted billions of people out of poverty, mostly in rural China and India.

I give you that background because the latest classroom activity was designed make my students somewhat complicit in that outcome so they might better understand it. (After all, your typical undergraduate is much more comfortable adopting a black-and-white view of decisions made by others than those they’ve, even in a fairly loose sense, made themselves.)

The decisions they faced were pretty simple, at least for those who’ve realized at some point that they probably need to know how to calculate expected utilities if they want to do well in my course.

Here, our hypothetical business expects a loss of $9.7 billion dollars if it opens no new factories, but a net gain of $4.25 billion if it opens one and a net gain $2.65 if it opens two. So the optimal number of factories is one. This, of course, tells us nothing about globalization yet, but it gives us an important baseline.

The second part is equally straightforward, and helps move us toward illustrating the stylized fact discussed above.

As before, if our hypothetical business builds no new factories, it expects to lose $9.7 billion. And if it’s going to build any new factories, it’s going to do so in location 2, because there’s no reason at all building at location 1 is simply throwing money away (given the implicit utility function of this business). So let’s compare the expected net profits for one versus two new factories at location 2: the former is $9.25 billion, the latter $12.65. So the optimal strategy is to build two new factories at location 2.

What does that have to do with globalization and inequality? Well, if we assume that location 2 is a developing country and location 1 a developed country, then when we move from the first part of the activity (where our hypothetical business didn’t even have the option of opening up new factories abroad) to the second (where they did), we expect to see: job opportunities for relatively unskilled workers in the developed world disappear; job opportunities for relatively unskilled workers in the developing world to expand; the latter effect to exceed the former, since twice as many jobs will be created in the developing world as destroyed in the developed world; and profits for the business to increase substantially. In short, we expect to see growth overall, benefiting workers in the developing world and businesses in the developed world, but a complete disappearance of manufacturing jobs in the developed world. And, again, this all very simplified (it is an introductory course after all), but that’s roughly what we’ve seen happen in the real world. That is, more or less, why the city of Buffalo, like Detroit and Rochester, Cincinnati, Flint, Cleveland and so many others, is shrinking, and its remaining population struggling to get by. And a bunch of students at the University at Buffalo just helped reproduce that outcome.¹

^{1. Well, sorta. In the past I made students play the part of the business more directly, and gave them points that reflected the outcomes the business would experience. I didn’t do that this time for two reasons: first, too much is left open to chance, and students are too prone to thinking that if a gamble didn’t pay off, it must have been a bad bet, no matter what the expected utility was ex ante. Second, the university was closed today due to a huge snowstorm, and next week is Spring Break. So I either had to make every one come to class on Friday for what should have been the second run of the activity, allow the two days of the activity to straddle Spring Break, or allow them to submit their answers electronically for a change. I went with the latter. And since I’m having them answer by email, and they wouldn’t see me draw a random number live in front of them, they’d undoubtedly be suspicious of the outcome. Therefore I figured it was best to just ask them to identify the optimal strategy, even though that means they didn’t really shut down all the factories in Buffalo but simply observe that there was money to be made in doing so.↩}