 by a tendency to be understated. So over the last few weeks I've been thinking about how to give an introduction for Professor Orly Aschenfelter that allows you to understand the importance of his contributions to economics over his career. So I've started out this way. Professor Orly Aschenfelter is the Joseph Douglas Green 1895 Professor of Economics and Director of the Industrial Relations Section at Princeton University. My inner Minnesotan says, that's pretty good. I hear Princeton's a pretty fine school too. But just telling you about Professor Aschenfelter's many accomplishments and awards. His position of leadership among economists and the volume and breadth of his research and publications didn't seem to convey the weight that his contributions deserve. So I tried this, I don't know, maybe a little over the top. But then it came to me. Professor Aschenfelter's most important contributions have had an impact on every one of us and every one of our lives. In 1972, Orly Aschenfelter was selected by the United States Department of Labor as the Director of its Office of Evaluation. It was here he began his work to use econometric techniques to evaluate government retraining programs. Over the next 40 years, his quantitative program evaluation has helped federal, state, and local governments determine the most cost-efficient and effective ways to deliver everything from government retraining programs to, in some places, mosquito nets, early childhood education, life-saving vaccinations to people around the world. Orly Aschenfelter's work has had a positive impact on anyone who pays taxes. So most of us, I hope most of us here. Because quantitative program evaluation is used and often required to identify the most cost-effective program before tax dollars are spent. Professor Aschenfelter is also credited with the development of a natural experiment method which involved clever selection of data to deal with inherent problems of identifying the direction of causation and a control for many influences that might confound a simple direct line of causation from A to B. For example, how can we control for differences in the talent or innate abilities that allow some people to be more productive using education to increase their income? Orly Aschenfelter's natural experiment addressed these problems through a very clever selection of data. He gathered his data by interviewing twins who attended the National Twins Festival in Twinsburg, Ohio. And an example of a natural experiment in the data collection. Professor Aschenfelter has used natural experiments to study the statistical value of a human life, age discrimination, and comparing real wages across countries in terms of Big Macs, among other things. There we go. My favorite Dr. Aschenfelter's natural experiments though, and I may be using that term a little liberally here, but involves the value of expert opinion. In expert ratings measure quality, the case of restaurant wine lists, Orly Aschenfelter attempted to determine the informational value of expert ratings to wine customers and to end the financial rating, or financial value, of the rated restaurants. For one experiment, Orly Aschenfelter and his co-authors chose to test the value of the Wine Spectator Award of Excellence, given by Wine Spectator magazine to the world's finest wine restaurants. For the test, the authors invented a restaurant in Milan, Italy, Osteriola, and Treppito. I'm not even trying to pronounce that in Italian. The only thing that was real about this restaurant was the pure invention, except it had a working telephone number. The menu was ordinary. And the wine list consisted of Italian wines that Wine Spectator itself had given its lowest ratings to over the past over the previous years. The authors submitted their entry, and along with a $250 fee to Wine Spectator, and waited to learn their fate. I guess it'd be nice to have a drum roll, but not having a drum roll. We'll just say that the fate, of course, was that they won a 2008 Wine Spectator Award of Excellence for Osterio El Treppito. Remember, this is fictional, right? OK. Demonstrating that the Wine Spectator Award of Excellence had no informational value to customers, but great financial value to assume about $4 million in entry fees to Wine Spectator magazine. It's my honor and my privilege, and that is indeed an understatement, to introduce to you Dr. Orly Aschenfelter. Thank you very much. Well, you can see what I'm going to talk about. It's a good thing it's after lunch, so you won't feel hungry. It's my special privilege to be here. There are several reasons for that, not least the fact that my mother grew up in Hope, North Dakota, so I always like to be close to North Dakota. It's a little easier to be outside of it and close than inside and close. In a town that reminded me today, in fact, when Dan Ariole was talking about how I was always told it had 500 people in it, it was only when I went to graduate school and learned how to use the US census that I found out there were only 300. I think that's the little white lie, I guess, they used to call it in Hope, North Dakota. And then to meet your president, the president of this university, who is a Princeton engineer, a graduate of the Princeton College of Engineering, a woman who entered Princeton at a time when the university had just become coeducational. I asked her if she felt lonely and she said yes. So it was a brave woman. And then of course the high point, without any doubt, due thanks to Paul, was visiting Jim's Apple stand, if I may call it that, out here on the road into St. Peter, the gigantic yellow thing. I have never seen anything quite like that and they couldn't get me out of there. I have never ever been to a place where not only could I find a knee-high orange. I haven't drunk one of those since I was a kid. They have an entire section of that store devoted to orange sodas. It was without a doubt unique. I'm going to talk today. I put up here comparing rural wages to the wage project. I'm going to use quite a few slides. A few of them are going to be technical. I think they're pretty well labeled. So I think that if you pay attention, you'll learn something from it. But I'm going to try to do this in a format that gives you a little bit of history, explains what I'm trying to measure, gives you a little bit of history, as well as explaining what it is that I'm trying to measure, then explains to you how this project came about. It has become quite massive, a project to try to measure rural wage rates around the world on a uniform basis. And then I want to show you how the measurements I've made seem to be consistent with other measurements. So I think probably they're useful tools for figuring out some interesting things about economic development around the world. And then finally, I want to show you some results from it on what's happened to inequality and economic development in the last 15 years. And I'll end with a very brief example of the reason why in part the data were collected. And that is I'll show you an application to a measurement problem in economics. And hopefully you'll find it a kind of a convincing way of thinking about it. Now, first of all, what is a real wage? Real wage is a measure of the amount of consumption that you can earn from an hour of work. If you could think of it as you get paid a certain amount of dollars per hour, if I divide that by a certain amount by the price of some goods, that's dollars per good, if I divide that, I get goods per hour. So the real wage is a measure of what people earn or produce even in the form of goods per hour. It's real, it doesn't involve money. Why do we do this? The main reason is to compare living standards around the world and over time. The value of that is that we can then try to evaluate what happens when there's a public policy program. We can ask ourselves, is it really true, for example, that some innovation has increased real wage rates in one place or another? And I think it does one more thing. It's an indirect way of measuring productivity in an economy. Generally speaking, real wage rates are what they are, and I'll show you this, it's pretty elaborate, are what they are mainly because of the productivity of an economy. So that's the thing I'll start with. In economic history, Dirdar McCloskey, you're from tomorrow, is an economic historian and well aware I'm sure that there's a considerable amount of discussion of real wage rates over long periods of time. It's one of the few measurements we can make on a pretty broad basis that you can take all the way back to Roman times. So there's even some controversy about what a free Roman actually labor earned. What I want to start with is an example. This example comes from Bob Allen, an economic historian, formerly at Oxford, now in all places at Abu Dhabi NYU. This is an example of how to calculate a real wage, and it's a real wage calculated in 1704 to compare London and Canton. The wage rate, of course, was calculated recently, but the data come from 1704 from the East India Company. In the first row, first column, you'll see things that people need to buy. And then in the second column, you'll see the ratio of the price of that thing in pieces of silver, which was the common way to think about measuring something as a currency at that time. At the very bottom, you'll see the ratio of the wage rate in units of silver in England to Canton. Canton is more or less what we call Hong Kong today, but it's also some area near it. In the third column, you'll see the share of the budget the best we know, a typical Englishman's spent or English family spent on each of these items. And in the last column, you'll see what a Chinese family spent on it. You can observe it pretty clearly, a simple point of economics, which is that when something is relatively more expensive, people have a smaller part of that in their budget. And this makes it more complicated to measure a real wage rate when you want to compare two different places. Because the budget shares what people consume and I'll come back to this is not the same in both places. You also see then that I can calculate the ratio of the cost of living in England to the cost of living in Canton. And you'll see that that ratio is very high. It's much more expensive to live in England than it is in Canton, even though the people in England make more money. And if you divide the two, you come up with a real wage rate. And the real wage rate is either, well, it's on the bottom row. And it'll be different, as you can see, according to whether you use Chinese budget shares or English budget shares, which means that a Chinese person facing English prices in China would be much worse off than they would be if they faced the prices they did face. You can also see that as for as a practical purpose, wage rates are about the same in both places. If I take an average of those two numbers, I'll get them round one. Wage rates are very close to each other in Hong Kong and London in 1704. This is an important point, and it's important because it turns out that throughout all of history, real wages have been the same in almost all places. The Roman worker didn't make any more or less than the London guy in real terms. And it's no accident that Malthus and Ricardo in the 19th century wrote treatises in which they predicted they wrote out a model in which they anticipated that real wages would everywhere and always be the same. It was one of the many reasons we call our subject the dismal science. But let me move on a little bit. It turns out that that prediction was wrong. In fact, just almost at the moment that Ricardo and Malthus wrote their treatises on why real wages would remain stable forever, it turned out to be wrong. Perfect example of economists trying to predict. This is an example I like to give to show you what the beginning of that period. This is the United States. And it's two measures of average hourly earnings in cents per hour. This is an eye opener for people that don't realize how little people made back in the 19th century. $0.25 an hour was a lot. Of course, inflation prices were much higher, too. It's two different indices, one that was compiled by a teacher of mine and another that was compiled by Paul Douglas. Both of them were at the University of Chicago at one time or another. And this is an attempt to measure prices. Douglas had come to the conclusion there was no real wage growth in the United States between 1880 and 1914. And Rees came to the opposite conclusion. And the way he came to the opposite conclusion was by constructing a different price index. So you'll see that what happens is that if you actually compare the real wage indices, Rees has a real wage index, the blue index, which is increasing, and Douglas has when it isn't. They're indexed, by the way, so that they're both the same at the very end. An interesting point about this is, how did Rees come up with the altered price index? Douglas relied on information that was primarily available on the Chicago Board of Trade and other places. Rees actually used what was, at the time, the internet of the day. Some of you in this room are old enough to remember, especially in Minnesota, that before Christmas, two very important books came into your house. They were called the Montgomery Awards Catalog and the Sears and Robot Catalog. Those two, I sat by the fire in North Dakota, and that was my entire life, for the entire run-up to Christmas, was dreaming about things I might get out of those two books. So Rees used those books to construct price indices for manufactured goods that could actually be considered comparable over a period of time. Just to get the bottom up to this point about real wages in different parts of the world, this is something I like to use. It's an average real wage rate. It's a wage rate which is a subsistence wage, according to Allen, and I've put it on here for a bunch of different countries. So you can see which ones had taken off and which ones had not. 1914 is commonly used by economic historians as a good ending date, because it's the beginning of the First World War, and things are not quite so normal afterward. You can see that most of the countries in the list here have real wages that are only slightly above subsistence or a little bit above one. You can see Japan, Canton, Beijing, Delhi, they're almost all the same, and then you get down to what were the developed countries? London at seven, Oxford at six, Amsterdam at five, and Chicago at six. By 1914, things had changed. The Malthusian prediction had failed and real wages had grown, so that now they were five or six times subsistence. There are two lessons here. One, things can change, and two, a lot of times they don't. The rest of these countries were left behind. And I'm going to show you in the next half hour that for all practical purposes, many of them still are, not all of them, but many of them still are left behind when you try to measure things with real wage rates. I like to think of the real wage rate as a measure of the standard of living. The idea is that we try to figure out what it would cost to live in a certain way and then compare what we observe today to that amount, and if it's more, then wages have gone up, and if it's not it, then they've gone down. But there's a second way to think about it, and that is to think about it in terms of an important measurement problem, which I want to bring up to you. This is one of the few things that has changed in economics since the time I first started. When you think about comparing living standards between India and the US, one of the most obvious differences is that many of the things that Indians consume are not tradable. So a haircut in India can be really good, but it'll cost much less than it does in the US. An apartment might be less good, but even so, it'll cost less than a comfortable apartment in the US. So there are tradable and non-tradable goods. This makes it very difficult to compare, because it means that almost uniformly, the budget, the basket of goods that you consume won't be the same in the two places. We basically, the way we try to solve this problem is by creating something which is an index, a quasi-tradeable, and I'm going to come back to the quasi-tradeable I use, but the basic idea is that the quasi-tradeable is the price of a good, which has in it something which is a substantial fraction of it which is not tradable, and a substantial fraction of it which is is tradable. And I'll use that to try to figure out what the price level is in one place to compare to another. There's one last point, technical point I wanted to go out of the way, and that is that real wages can also be used as a measure of the productivity of labor. This does require a couple of things. First, it requires that we assume that workers are paid in accordance with their productivity. Secondly, it requires that we collect data on workers that have the same productivity in all these places, and therefore, would be interchangeable if we move them. And then when we compare the two, we can use that to measure productivity differences. So I'm gonna use this in two ways. The idea is that the real wage is a measure of how well-off people are, and it's also a measure of how productive a country is. Both of those things are approximately true. Hopefully this one. Okay, this is the product I'm gonna study. This is gonna be our quasi-tradable. Is there anybody in here who has never consumed a Big Mac? But the people lie. I believe people lie. This is what is commonly called a full meal in the literature. We're gonna study the price of Big Macs, and we're gonna use that as our measure of a price index that we can study over time and that we can study across countries. Now I'm here to tell you that Big Macs are the way to go, partly for a very old reason that's connected to North Dakota of all things. I still remember as a child when I moved from California to North Dakota. Some people often ask me, why did we move? I always say we did it for the weather. My mother, bless her heart, stopped, the first night we stopped was in Yuma, Arizona. I still remember it like it was yesterday. We went into a greasy spoon and we had dinner. We came out and the next morning we were both totally sick. We could barely move. It was just unbelievably bad. My mother announced the fundamental rule that we were gonna follow from then on. We're always going to eat at McDonald's when we're on the road. The key thing about McDonald's for its growth and why it still exists and does so well in most parts of the world is food safety. Most of the people in this room cannot drink the water that comes out of a tap in India or even in most parts of China. But if you go to a McDonald's in those places and I'm here to attest to the fact because I was there, you can drink the water that comes out of, that they serve to you. So food safety is something which will keep this company going for a very long time. It's a very important issue. But I'm gonna use the Big Mac as a measure of a quasi-tradable. Now why do I say that? Well because half of the price of a Big Mac is labor. And that's a non-tradable good. The worker who works in India can't sell his labor in the United States, at least not very easily. The other half of that good is a tradable good. How is it that McDonald's can guarantee food safety everywhere? Well the way they do it is that they basically have central facilities where almost everything is frozen. And where then things are taken care of in terms of both the aesthetics as well as the sanitation and then they're shipped to McDonald's stores around the world. There's typically an area, a distribution center for each of those areas. So that's why a McDonald's Big Mac is more I think the same everywhere you get it. Tom Friedman has an entire chapter about Big Mac's in his books, the olive and the lexus, the olive tree and the lexus and the olive tree. In which he says that he's eaten Big Macs in more countries than anybody alive, which is quite possibly true. He also believes that they are identical in all those places. But of course the reason he wrote this chapter in that book is because he has what he calls the McDonald's theory of conflict. Which is that no two countries that have McDonald's have ever fought wars against each other. That's why we know that the Kashmiri conflict can't bring about a war in India and Pakistan because they both have McDonald's. I'm not sure I ascribe to that theory, but anyway if you read the olive and the lexus and the olive tree you'll see that. This is the first McDonald's, how did I get into this project? I was actually doing some consulting in India and when I was consulting in India the Indians loved to argue about poverty. They never stopped talking about it. Inequality and poverty, love to talk about it. So I just wanted to know, there had been some reforms in this country. What happened to the wages of an ordinary person? Well, I say it depends, it depends on it. Well, so then my answer was, well, is there a McDonald's? Yes. And it turns out that was the McDonald's. There was a cow in front of it. You can't eat cows by the way. The Big Mac that I had there with, those days they made out of lamb, now they're made out of chicken. But they're very good. The lamb one I thought was outstanding, but they can't get that anymore. So I went with my former student, Sergio Bala, a very distinguished Indian journalist and economist now. We went out to a McDonald's and we actually went in there and we saw that, meal combos. You can see that's something at 99 rupees for the chicken Mexican wrap, whatever. Why they have that? I don't know, but there they are. Do I look just like any other place? I bought, here you can see what I bought. I like to show people that I really did go to a McDonald's in India. I'm not making this up. And I got my, it's off this receipt. There's a copy of it. I had a Big Mac. Of course, my Indian friend is a vegetarian. So he had the veg comb, whatever that is. And of course I always like to have a shake. You can see the cash tendered. And this was a receipt from, in fact, it's from 2008, very long time ago. So I got the receipt. And then there was only one thing left to do. I saw, I was, we were sitting there in the restaurant where there were some people that were working behind the counter and some others that were picking up trash from the, so we just asked them, how much do you make here? What's the rate of pay? Well, the Big Mac costs, what is it? 52 rupees. And the workers were making about 18 rupees an hour. In other words, they would have to work almost four hours to buy one single Big Mac. Now that's a very, very simple way to explain that roll wages are lower in India than they are here. They're certainly lower in dollars because 18 rupees translates to about maybe 40 cents. So in dollars, they're lower. But in Big Macs, which is a quasi, it's not exactly what most of those kids would eat. In quasi-tradables, it's a measure of the purchasing power in that country. So I've done on me, well, I wonder if you could do this in other places. So I did like everybody would. I enlisted a co-author. This co-author is from, didn't anybody know where this McDonald's is? It's one of my favorites. It's in Prague, in Czechoslovakia, in the Czech Republic. And I enlisted, we went and visited that one. And there I am with him, my compadre in this effort. We went and visited this place and we got a tour of the entire place. The manager didn't even, he didn't even have any trouble telling me how much the workers make. Partly because the turnover at McDonald's is so high that there's no question what the wage rates are. They can't be a secret. They have to come out. If you call, you'll find out right away because they're almost always trying to hire somebody. So we got together and decided to organize a project studying McWages. The notion is we're gonna compare entry level, basic crew jobs at McDonald's. And why is that useful? Well, first of all, the skills that are used everywhere are identical. We're getting the same, they can't complain about the fact that skills are different, they aren't. The hedonic job qualities are virtually identical. So everything about the job is about the same. Moreover, they're producing an identical product using invariably an identical technology. McDonald's doesn't change the technology going from place to place because they have the McDonald's way of doing business. They're also active in over 140 countries. The activities of these things are governed by what's called their famous operations and training manual. It's a giant thing where you can learn more about it if you go to Hamburger U, which they have just outside of Chicago, where learning how to run on McDonald's is taught in 25 languages. So I thought, well, here we have an example where a giant American corporation, which is producing homogeneous products that everybody gets all mad about, is actually gonna give me the chance to collect some data that will be very useful for lots of things. And that's what we started doing. So let me, oh, incidentally, I should add, if you look at the big handbook, it tells you a lot of things about how to make the food that's in the store. It does not tell you what to pay your workers. The goal there is to pay the workers, your local workers, what the local wage rate is. And the manual is completely silent about how you figure out what that is. They don't want you to violate the law, so I think it would be rare that they would pay anything less than a legal minimum wage, but other than that, it's left up to the owners to do it. Here's the data collection. We now have data for about 64, 65 countries. We started doing this in 2000, so we have some data for 27 countries since that period. We collect the hourly wages of crew member. The sample frame, I don't think we need to talk about that, but we try to find at least two of the major cities in a couple of stores. And I can tell you that the correlation of wage rates is pretty high as we've measured them independently. Now I do usually make an offer at this moment in this talk, and I'll tell you what the offer is, but I'm very reluctant to make the offer here today. My usual offer is that, you know, we always like to check our data. So, if you were to travel to a really esoteric place and you bought a Big Mac and you sent me the receipt, you can have a drink, too, and you sent me the receipt and you interviewed someone in the store and found out what the wage rate was, and you sent that to me so I could validate the wage rate data we collect, I'll pay for your dinner. I used to say anybody who did that anywhere, and then I was over my head there on that. I don't really want thousands of Big Mac receipts from England. That's easy to measure what goes on in England. But maybe if there's something way up in Northern Minnesota, I might consider that. That's an offer. I'll pay for the full meal, including a Big Mac. All you gotta do is send me the receipt and you gotta write on the back where you were, when you did it, and what the wage rate of a crew member was. And the easy way to do that for the students in this room, of course, is just to say, I'm thinking about, maybe I'd like to work here, how much would you pay me? We've also collected in more recent years data from Starbucks. Starbucks has expanded all over. I thought, well, here's another giant American company, let's take advantage of these people. Starbucks is interesting because there is, of course, we're gonna have a, I don't think it's quite as good as the Big Mac. We can also have a quasi-tradables price. What would that be? Well, it's a cup of coffee. So I could also measure how well off you are. Coffee, by the way, has the same characteristic. A lot of the cost of the coffee is local, making it, but the beans are hardly ever local. So some of the raw materials that go into it come from elsewhere. But we did this primarily as a validity test so we could see what was actually going on. We started doing that in 2011. Now let me give you the theory of what I think this is about. It's gonna make it very simple. In each country, for workers of a certain quality, there's a demand curve. And there's some supply, which is probably very unelastic. What that means is that what drives the wage rate in a country from one year to the next or from one place to another is demand. Demand is determined primarily by productivity of workers. So what I'm really thinking about is that the McDonald's wage is just a measure. It's a glimpse, it's like a picture of what's going on inside of a country. It reflects what wages are in the country as a whole. It happens to be a particularly nice example of a wage in the country as a whole because it's for the same exact skills and the same hedonic qualities for jobs in all the different places and over the whole time period that I've been studying it. But to think of it, you should just think of it as like an indicator of what's going on in a much larger place. You could think of it as like in your gas gauge. It's a measure of what's in the tank. It's not perfect, but it should be highly related to it. Okay. We've done quite a bit of reliability studies, but as I say, one of them is we get people to help us by collecting things. There are a couple of limitations they ought to mention. First is the wage rate of market wage. Wage rates are not gonna reflect productivity if they aren't based on the market. So when there's a minimum wage, for example, and they can be very high in some countries, that minimum wage means that the workers in that particular place have to produce enough output to justify themselves, but it doesn't reflect what the wage rate is in the country as a whole, what the wage rate would be if there weren't some manipulation of the wage itself. So you can think about that. The same thing is also true about the fast food price. It could be that there isn't really competition in the fast food market, and so the price might, for some reason or another, be a monopoly price, not a competitive price. Those are both limitations to keep in mind when you do this. Let me show you a simple example of a validity study. Here's, these are wage rates, this is a chart. I don't think it's too complicated. It has, on one axis, the x-axis, wage rates in dollars at McDonald's in different countries, and on the y-axis it has wages at Starbucks in different countries. The way we've done this is we've just fixed it so that the wage rate of the Starbucks wage rate relative to the Big Mac, the wage rate of McDonald's is one. So basically, if wage rates were the same ratio, Starbucks to McDonald's in all countries, all these points would lie on a 45 degree line. That's the line that I've drawn in there. So there's no complex statistical analysis. Just a 45 degree line. If the points lie on the line, it means that the ratio of the wage rate at Starbucks to the wage rate at McDonald's is the same in that country as it is in the U.S. And you can see, even though some of these wage rates are very high, they fall right on the 45 degree line. So this does suggest that what we're measuring is an indicator of what the wage rate is in the country as a whole. This is an example now. What I've done is I've taken Big Macs per hour. So let's think about that for a second. I get the price of a Big Mac. That's in dollars per Big Mac. And then I divide that into the wage rate. So I end up with the amount of Big Macs that you get per an hour of work. So I'm gonna call it the BMPH further on. You can also do it with coffee. And then I've given you that same plot. You can see there's some exceptions. Coffee and the real wage measured in coffee and the real wage measured in Big Macs is not identical in some places, they're off. But you can see they're very highly correlated. Here's another one. This is an attempt to measure, to compare the wage rates at McDonald's to wage rates in manufacturing across different countries. And you can see, except for way up here, does my finger show up? Yeah, you can see that. See my finger, I can touch this with a finger. Except for these are countries that have very high minimum wages. So those are examples where the wage rate, the minimum wage is relatively high compared to the manufacturing wage because the manufacturing wage is not affected by the minimum wage, statutory wage, but the wage rate in the McDonald's will be. There's another measure for people who are really get esoteric. There's something that I get involved with called the International Comparison Project that's run by the UN. And it tries to create purchasing power indices for different countries. And this is a comparison of the wage rates at McDonald's using our purchasing power peri measure and the purchasing power parity measure that the ICP uses. You can see they're pretty close. Okay. This is a pretty good example of the purchasing power parity measure of the wage rate versus the wage rate I've measured in Big Macs per hour. So I'm just trying to assess whether or not this is a closely related measure of purchasing power in the two country, in all these different countries and it seems like it really is. So on the whole, that's a change in the purchasing power parity. You can see it's highly correlated with the change in the Big Mac things. So those are all kind of results that should give you some notion. I like to, whenever someone tells me about some fact I like to know where it came from, how do you know it's any good? And that's the attempt that I'm trying to show you here is how we did it, a little bit of an attempt to show you what it's really like. Now let me show you a new graph. This is something I just did this summer. It is brand new. It's new enough so that when we constructed the PowerPoint and we were using it a little bit ago, it wasn't actually working. But apparently it is working. I can see it's working down here on this other computer now. So the first thing I'm gonna do is I'm gonna compare real wages for you across the United States. These are gonna be real wages measured at Big Macs per hour. We're gonna find out how Minnesota looks compared to Texas using a measure Big Macs per hour. Another thing I'm gonna do is I'm gonna show you real wage differences across countries. Third thing I'm gonna do is I'm gonna show you real wage changes over time. And I'm gonna do that in two periods before the financial crisis and afterward. And I'm gonna do that for countries outside the U.S. because I don't have these data for inside the U.S. except for a relatively small number of parts of the country. And then I'm gonna show you what happens to real wages on the path of development. A little bit about the old, that quote I traded about I mentioned before and then I'm gonna show you an application which is to migration. And then I think we'll be done and I think I'll be just about on time. So let's take a look. That's Big Macs per hour this summer. These are county medians. This is a result of surveying 5,000 stores and then constructing a county median and making this map. You can see the legend down here. I can hardly read it myself. If you look at the bottom, the dark blue is two to 2.4 Big Macs per hour. The low wage ones are one to 1.8 Big Macs per hour. Now the amazing thing here, so the blue ones all have Big Macs per hour more than two. By the way, if there's nothing, if it's all white that means we don't have any data. The ones that are light and gray, those are where people make less than two Big Macs per hour and typically the 25% less than two Big Macs per hour. So you can see the story. There are very high real wage rates along the west coast across the northern part of the US into the eastern side and then everything else is lower and there was some exception being Florida. So there actually are two countries here. We are not integrated. People could move, low wage workers could move from these poorer places to richer places and the least measured in Big Macs per hour would be better off. We've done this with, there are now regional purchasing power parity indices that our US Department of Commerce puts out. We've done the same thing with those. They show the same thing. So this demonstrate its real terms. There are as much as 30% differences in the real wage rates people are receiving in different parts of the country. I put it into a map that has the state aggregated. It's a little bit easier to see it that way and you can see exactly what I said before. The real wage areas of the country that are high are up here. They come down through here. They're a little bit down here. But the middle of the country and the deep south is still, has a relatively low real wage and these are for people who have the same skills and are doing the same thing. So some people have, I noticed that in the poverty report recently some people have mentioned one solution for some of these people is to move. Come to Minnesota or I'm sure they'd be welcome. So that's sort of the current level in the US. I'm gonna show you some details now for 2007. This is an interesting period because I picked it on purpose because it's, and I've aggregated these data. I picked it on purpose because 2007 is really before the financial crisis got rolling. And the way I've done this in the first column is the name of a country or a region. The trouble is that some of the countries are very small and I concluded that it might make more sense to aggregate them together. So I put Latin America together at the bottom, for example. The countries everybody wants to see are the US, Canada's pretty large. They're your neighbors to the north here. Russia, South Africa, interesting country. But the two big ones that everybody cares about are China and India. And then Japan and then the rest of Asia and then I put Western Europe together because their countries are relatively small too. So you can see right away, these are now measured in dollars per hour. Now what is measured with exchange rates as of that date? What is an exchange rate? It's something that converts whatever's being measured in another country into dollars at a convertible exchange rate. It's basically, the exchange rate should be, it isn't always, but it should be determined so as to equilibrate prices in different countries. In other words, the exchange rate should really equilibrate so that a price of an iPhone is the same everywhere. Or so the price of, well, as the economist says, the price of a, not quite right, but the price of a Big Mac, at least would be the same where the amount of tradable goods is the same in each component. So the notion is that the exchange rate is supposed to give us equation of wages in a dollar measure, in tradable goods measure. So that really tells us what the wage rate is measured in tradable goods. It's not a measure of welfare because in many of these countries, people don't consume as much tradable goods as we do. So you can see the US is not the top. The Europeans have the highest wage rates and of course, a lot of that is a result of high minimum wages. The US is pretty high. In 2007 it was $7.33. You can see Canada at $6.80. South Africa $1.69. The two I think you should look at are China, which in 2007 had someone in McDonald's making about 81 cents an hour. And someone in India who had a wage rate, as I more or less mentioned earlier, of 46 cents per hour. Now I'm gonna show you what happened to that over time, but you shouldn't be too optimistic. You're not gonna get from 81 cents an hour to $7.30 an hour, even at 10% growth rates all that fast. Then there's the, in the third column, I put the ratio of the wage in the country that you're looking at to the wage rate in the US. So you can see how, what fraction they make of the US. You can see in this period, most countries were below the US. Then the Big Mac price, you can see Big Macs are cheap in poor countries. In China it was only $1.42. India $1.29. That's a little more. Now I paid 52 rupees, that's probably pretty accurate. It's about what I showed you in that original slide that I gave you. And then Big Macs per hour over on the right. So back in 2007 in the US, it was about 2.4. Our Canadian brothers, 2.2. Russians were at a little over one. South Africans at 0.81. China at 0.6. India at 0.38. Japan is high, was the highest at that time. And of course the highest of all are the Western Europeans. So that's a cross section picture. You should probably keep it in your mind as we move forward. This is a picture of the difference. It's how many Big Macs you get per day. I did this on purpose. So we have a nice American person putting their hand over their heart at McDonald's. And we have a nice Chinese guy putting his hand over his heart in McDonald's. He's came actually off the McDonald's websites in those places. So you can see the gap in actual Big Mac things. This is really hard to see. You know what, I'm going to skip over it. This is an attempt, a very crude attempt to give you an index of how the distribution of wage rates looks across the world. It's population weighted. So what you see here is a population weight. And you can see that it's somewhere between 0.7 and this is $5 here. So this is about 250 here. So these are all below $2 an hour. And you can see that's where the vast majority of the population is. It's right in here. And then you can see there are some countries that come along here. A lot of them, some of these are Eastern Europe. And then there's the US because we're a big country. And then there are some Europeans who are in that group. And then there are some that drag out way into the far end. So this is a kind of a picture of the way wages look in the world. The vast majority of people are in low wage countries. Now this is what happened to growth. This is the most optimistic I can get. If you're going to get one take home thing, this is a really optimistic one to take home. What I've done here is not, this is complicated. So the wage ratio is the wage rate in the US in 2007 compared to 2000. So this number right here tells you that in the US wage rates at McDonald's grew by 13%. In Canada over the period 2000, 2007. So that's a period that kind of from the first recession, first Bush recession into the 2007 period. Canada grew by 1.50%. Now take a look at Russia, four and a half times. The Russian wage rate grew by four and a half times between 2000 and 2007. Look at China, it doubled. Look at India, went up by 60%. Japan down by 5%. This is the sort of beginning of what we're gonna see in a second. Some countries are progressing and converging. The Russian one by the way is one to take home because this of course corresponds precisely with Putin's rise. It's no accident that Russians love him. The man has presided over the greatest growth in Russian history. Now much of that is a result of the fact that it was a gigantic depression in the late 1990s. So some of that's a recovery from that. But it's still the case that even in Russia, real wages are. Now that's the wage rate measured in dollars. This is that same ratio measured to the US. So this tells us that Canada grew 34% faster than the US. Russia 411% faster, China 71% faster, India 40% faster, and Japan fell compared to the US. This is what happened to the price ratios. And here's what happens. This is what happened to the growth in the Big Mac. This is the Big Mac per hour ratio. So you can see what happened to the US. Actually it fell by 7% during this period. The wage of an ordinary person working in McDonald's went down, saying it was true in Canada, but it was not true in Russia, or in China, or in India. These were periods of rapid real growth in all three of those countries. Japan grew by a little bit. So this is the good news. When I first started collecting these data, and we branched out in 2007 or 2008, I thought this is gonna be phenomenal. I mean, the world may be coming together. We may have, this accounts for two and a half billion people right here. This, if we could actually get those countries to grow at that rate for a long period, we would have phenomenal convergence in the world. Well, it didn't happen. Starting in 2007, things just collapsed. I think the thing you probably wanna look at the most is to go over to the, this is what happened to the wage ratio from 2007. This is what happened to the wage in the US. It went up by 6% between 2007 and 11. But this is probably what you wanna look at, which is what happened to the Big Macs per hour while they continued to fall in the US. They actually fell in Canada also. Russia had an increase. South Africa had a decrease. China had a 24% increase. Now you saw in the earlier period, China's road wage was growing really fast. It started to grow this over four years, so it's still growing pretty fast. But it's down now, it's not growing at 10% a year. It's growing at more like 5% a year during this period. And the rest of these countries have all had declines. Real wages have declined post-financial crisis. I've only taken this to 2011, but actually there's no recovery in the period since, although it hasn't continued. So the good news was we had a phenomenal growth period in poor countries in 2000 to 2007. And that came to a pretty abrupt halt with the exception of China, which is the one that still seems to be growing, and the exception of Russia. You see the growth rate in Russia, backwards it's not 400%, so it's not growing the way it had before. Nevertheless, Putin has not had to face anything like a problem of economic stagnation the way people have thought that he had to. So for him, this has been good news, and it shouldn't be surprising that he's a very popular man in Russia. Let me take a few minutes to talk about McWage's along the development path. Blassa and Samuelson, so this is a little bit of economics. Basically the point here is if you look at this equation, this is the price of my quasi-tradable, and I'm imagining it, it's a function of the wage rate in the local country, and the price of goods, and you notice there's no subscript on that one, so it's the price of goods that are tradable. What that implies is if you look down here, is that the wage rate measured as something in trade, quasi-tradable goods, in other words, measured as a way of thinking about welfare benefits, is actually a concave function, this is for economists, of the wage rate measured in dollars. I wanna show you, this is something I didn't know much about when I was younger, but it's a common attempt to explain why it is that you can have the price of some things that are the same all over the world, and the price of some things that are so different. So the idea is that some of them are tradable and some are not, and what drives that is the fact that in the tradable section of the economy, some countries are much more efficient than others. That's more or less the explanation everybody, I think, would give for the fact that real wages are not the same in all these countries. So China just hasn't had real wage growth, it's also had growth in productivity per capita. So I'm gonna show you that there's something to this. There is that concave function, though it is actually true that what happens is as the wage rate grows over the development path, the Big Mac price grows, but it doesn't grow at as fast a rate, which basically means that real wages converge, but when you get higher up the chain, they converge at a slower pace, because you're replacing the non-tradable good with something which costs much more money. Anyway, that's the thing, this is a more modern version where I've actually made the size of the ball equal to the size of the country. You can see again, there are some exceptions by the way, but you can see again that this kind of notion that there's a development path, hopefully a development path, holds up in these data too. Now this is a way of showing you that growth has been negatively related to levels. So here's the US, here are the poor countries. The growth rate has been much higher in the poor countries than in the rich countries. That's really good news over the period 2000 to 2004, but remember most of that happened between 2000 and 2007, because I don't want to exaggerate the potential for future change. Let me just take one more minute and then I'll close this up. Well that's an example, I should show you this. So this is GDP per hour worked on this axis, and on this axis is a purchasing power measure of the wage rate, and you can see they're pretty close together. So what this does is it allows you to actually get a measure of GDP per capita by just going to one restaurant in some country. Much cheaper way to collect data, most people expect. In any event, I think it does demonstrate that any measure you have, this is total factor productivity measured here, and this is a measure from our McDonald's number of total factor productivity, and you can see they fall right along a nice line like this. So it looks as if a country that gets richer is gonna have workers that get paid more, and the big problem is to move into the higher level. Let me just take a moment to talk about welfare and migration. This is a topic, it's a very hot topic right now politically. Now I'm gonna talk about as an economist, not politically. Many people, there is a group that thinks that the far and the way the easiest way to get rid of poor people in the world is to move them to the US. Or another way to put it is, if we could move people from the very poor countries to the US, all the evidence we have is that they would make more money. And one of the proofs of that is this is what I've just shown you. If we can certainly find an Indian guy working in a McDonald's who we can move from India to here, and his wage rate will go from 50 cents an hour to seven or eight dollars an hour. So migration that's from low wage to high wage countries increases the world's output, and it increases incomes. So one kind of way to think about migration instead of thinking about it as a political problem is, maybe this is a poverty alleviation effort. Maybe this is a way to think about alleviating poverty that does require some cost, and it's certainly a cost that has to be borne in part by the person who's moving. But as long as the groups are not really large compared to the base group, this is a way to actually bring incomes together between different groups around the world. Now one of the problems with trying to estimate the welfare gains from migration is that if you look at actual migrants, you don't really get what I'm showing you when I measure these wage rates. Actual migrants are selected. The people who show up here are not, I've met two students in fact who are my hosts who have come to the U.S., and for all I know they'll stay in the U.S., and there are students at Gustavus Adolphus. Those are not typical people from those countries. So these people have been selected. Finding out what they do here is not a good way to find out how well they would do compared to somewhere else. So this is an example where we can actually take migration and ask ourselves if we were to compare these people, this is not a measure of the actual net effect, but it's certainly an estimate of it, take those people, think about what they're making in one country, what they're making in another. I have ways to measure both of those, and then we can add them up if we have data on migration itself. There's several papers that construct migration measures like this. So this allows us to get at measures of the potential benefits of migration that are not related to selection, and let me show you some analysis of it. Now, first of all, I'll show you some tables too. On this axis is the difference between the purchasing power measured wage in the country that people move to. These are all migration pairs across all countries, by the way. So this is actually measuring the best we can, depends on what period we use it. All migration in the whole world, it's net migration in the whole world. So it takes every country pair, figures out what the wage rate is in the home country and in the new resident country, and figures out how big that difference is, and I put that down on this axis. You can see these are quite large numbers. They range up to $10 an hour, and down as low, maybe, as one or two. The vast majority are in the middle. So these are the wage gains that people potentially would make by moving from the, that have made, basically. It's a measure of what they did make by actually making their migration decision. You can see it's a lot. One of the interesting features, by the way, of this chart is you notice it's not, it isn't true that there are more migrants capturing the higher gains. So it isn't, there is no connection here. There's no correlation between the size of the gain and the number of migrants that you see, which is on this other axis. So in a way, this is, migration is happening, but it's certainly not explained by this. But on the other hand, it does show large gains. Here's a fairly good example. You can see in the typical, in 2006 and 2007, the average person power wage gain for a migrant was about five bucks an hour. And this is in a later period. We have 550 country pairs down here and 96 up here. This is a later period when more country pairs. But generally speaking, there's no sign of a decline in this. And the average wage gain is about $5 an hour. Well, you can do that. This is a total gauge. These are in the millions, of course. But you can figure out what that is for 2000 hours a year. That's $10,000. For someone in a poor country, that's a phenomenal amount of money. So this is kind of an application of what you can do with things like this once you have those data. You can try to construct measures of wage rates that you can use to measure migration. And we've done that again in another data set. So what I wanna let end with is the following. My goal here was to try to show you a research project, how you construct one. I realize it's a fair amount of work. How you can try to validate a little bit. What it shows, sometimes quite surprising to people. For example, the differences in wages across the US and probably the growth in wages in some countries, but not in others. And then I wanted to show you an application. The application, there are many other applications you can do, but I wanted to show you what you in principle can do with it. And let me leave you with one last thing and you're welcome to call out if you know. I'm gonna show you some really pretty pictures of restaurants all in different countries. All of which I can say I have been to. Where's that? Yeah, got it. Where's that? I think that's China. There's only one I remember actually. Where's that? I'm not sure. I don't know where that is either, I've forgotten. Oh, we all know where that is. And this one I remember, that's Korea. Thank you very much. Thank you, Professor Ashlenfelter. And we will again invite our panelists up to the stage and we will be gathering your questions throughout the arena. So please turn those in and we'll reconvene in a few moments.