Cross-country comparisons of wage rates: the Big Mac index

<p><em>Romesh Vaitilingam interviews Orley Ashenfelter for Vox<br />
<br />
January 2009<br />
<br />
Transcription of an VoxEU audio interview [http://www.voxeu.org/index.php?q=node/3487</em></p>
<p><strong>Romesh Vaitilingam</strong>: Welcome to Vox Talks, a series of audio interviews with leading economists from around the world. My name is Romesh Vaitilingam, and today's interview is with Orley Ashenfelter, Professor of Economics at Princeton University.</p>
<p>Orley and I met at the American Economic Association's annual meetings in San Francisco in January 2009, where we talked about his work comparing wage rates around the world using data collected from McDonald's, the hamburger chain. I began by asking Orley how he first became interested in this issue.</p>
<p><strong>Orley Ashenfelter</strong>: I became really interested in this project when I was visiting India in a project involved with McKinsey, the consulting group, the old Global Institute that was doing a kind of country study in India.</p>
<p>The way it came to me as being of interest is that there was this continuous interest in development in India. Most people would argue a lot about the data - whether poverty had increased or not as a result of a more open economy and so on. This was back in the late 1990s, so it was 10 years ago.</p>
<p>And if you're interested in measures as an economist, it seems sort of silly to argue about something that you can&rsquo;t measure. So one day, I realized that in New Delhi there was a place that was very much like the place where workers worked and people bought things in the US, and that was the McDonald's.</p>
<p>So I went with an Indian friend who's a journalist now, a kind of well-known right wing pundit in New Delhi and my only student who was a development economist. We went to eat. I still kept the receipt in fact, always being worried like people used to claim that Cyril Burt faked all of his data.</p>
<p>I actually ate at McDonald's. I ate at McDonald's in New Delhi, and I had a big Mac, which was called a Maharaja Mac. Because the Indians in general don't eat beef, this was actually made from lamb. They offer the vegetarian version of that in many places, but this was really terrific I found, the lamb burger. And I had a shake.</p>
<p>I suddenly realized that this was similar to what would have happened in the US in the 1950s, in that here was a technology. The store looks exactly the same as any other one. The manager took us on a little tour. The back, the kitchen area, was exactly the same. They freeze things, so they have to have their own internal power generation system, because you can't rely on the power in India.</p>
<p>I suddenly realized that this was the same idea as McDonald's had back in the 50's. Not just that the food was something good, but it was safe. I could go in there at eat that, even though I had been warned by everyone not to drink the water, to drink only bottled water and so on, but that's not true in at McDonald's. You can go into McDonald's, and you can have a Big Mac, a shake, a coffee, a soda and whatever, and you'll be fine. That's what McDonald's is selling me, to some extent.</p>
<p>So at that time, McDonald's was expanding pretty rapidly in India. It wasn't very big, but it was there. And I got the idea of following up on The Economist and not just collecting data on what a Big Mac costs, but collecting data on what a crew member at a McDonald's made. At that time, I bought a Maharaja Mac for about 40-45 rupees, which is maybe I think about $1 or $1.25. The worker was making about - if I remember correctly - maybe 10 rupees an hour, which is probably 25 cents.</p>
<p>I realized that you could do cross-country comparisons. So, we got together with McKinsey, and we did a pilot study in 2000, as I recall, I may be off by a year, just to see whether or not we could collect information on this using their intelligence survey people.</p>
<p>Obviously, McDonald's, which is headquartered in Illinois, Hamburger University teaches in 20 languages. They have a many-hundred page handbook that they use to standardize products all over the world. McDonald's clearly does know all that I'm finding out. But they won't share that information, and I don't blame them.</p>
<p>McDonald's restaurants pay market rates, I'm sure. You don't pay below-market; they wouldn't get any workers. But it always bothers Americans and people in developed countries to find out what people in those countries are making. So, I'm sure they don't want to share it. So we started collecting data.</p>
<p><strong>Romesh</strong>: So you have to go in and ask them. You look at the menu for the prices, and then you ask them.</p>
<p><strong>Orley</strong>: And then you have to find someone to ask. In fact in our case, we just asked one of the workers. I mean the workers have no reason not to tell you. So in fact, I offer now anybody who has a McDonald's meal, send me the bill, and ask the person who served you how much they make. Write it on the back and I'll pay for it.</p>
<p>I've got some from Cyprus and Denmark. People have definitely taken me up on this. We started then collecting. After we realized that we could do it, we collected data from more countries, about 30 or 35 countries,</p>
<p>We did collect data on India and China, but we were especially interested in Eastern European countries. Because in the early 2000s, the real issue then was, how fast were they going to rejoin the western European countries? So we did. We collected data then in 2001, and then we went back in 2007 and expanded the project quite considerably.</p>
<p>We also hired&hellip;I went to a firm that does this. It's a business in Germany, a German firm. They must speak at least 10 languages, maybe 20, and they specialize in obtaining information of this type for other companies. I of course am not a company; I'm just doing it for myself.</p>
<p>So we strategized about a better technique for collecting data. I wanted to collect data at multiple levels in India, not just in the capital city but also in other cities. And by this time of course, that's possible to do in India and also in China. They are both growing quite rapidly. Also, we were interested in whether or not wage rates would be different in larger and smaller cities, and what would happen over time.</p>
<p>So we collected data in '07, and we'll go back again in '09 to see what happened. We would have done it anyway, but the '09 will take on a special interest because of the fact of what's happening in the economies of all these countries in the last year. I think every two years is pretty fine for us.</p>
<p>Anyway, it allows us to do two things. We can ask the question &quot;what does a worker make?&quot; and value that in dollars using the current exchange rate. If you're a businessman, you're asking, &quot;If I'm going to put a plant somewhere and I have to pay someone to do some work for me, I want to know what it's going to cost me in dollars, because I'm going to sell the product in dollars in some other place&rdquo;.</p>
<p>The other thing you can do is you can think about how much a person makes and value what they make in Western standards. So you could ask, for example, how many Big Macs they can make per hour. The results are pretty interesting. Basically, probably not a bad way to think about it is, if a country can provide wage rates for the people at the bottom of something like 5-7 dollars, 5-8 dollars, that means that you're going to be able to provide 2-3 Big Macs per hour per person. That's sort of the developed level that we see in the data.</p>
<p>In other words, if you think of Western Europe, US, Canada, and Japan, they're pretty close together. So, in fact, when there was the question about opening up the borders in Europe, before the East Europeans had joined, it was quite apparent, looking at the Big Mac wage index, that wages were not very different across European countries. So when people worried about this gigantic migration that might potentially take place between Italy and England or wherever, I didn't think there would be very much because the wage rates just weren't that different.</p>
<p><strong>Romesh</strong>: So that was quite a surprising finding from your data?</p>
<p><strong>Orley</strong>: Yeah. I think, at the time we did it, it was a little bit surprising, but I think that's one way you can use it. Now, Eastern Europe is a little different. Eastern Europe, the wage rates are lower, and so either there will have to be some blockages on migration or you will see considerable amounts of that.</p>
<p>So one interesting thing was there's kind of a group of countries that are at the top. Then there's a group of countries that are toward the bottom: India and China make those up. These are countries where wage rates are below 50 cents an hour. And that means, back then, that the Indian worker would have to work four hours to get a Big Mac. Of course, they wouldn't buy that. I mean, it'd be silly to expend so much effort to buy a Big Mac when you could buy a perfectly good paneer or something that would cost you much less.</p>
<p>So that's one lesson learned. And then there are some countries in the middle - Singapore, Hong Kong, pretty good examples of those - where wage rates are much higher. Like, for example, the wage rate for someone at a McDonald's in Hong Kong is much higher than the wage rate for someone at a McDonald's in China.</p>
<p>People will often ask me and say, &quot;Well, how can that be? Aren't they the same country?&quot; And of course, the answer is there's no free migration from China to Hong Kong. And Hong Kong can more or less make that wage rate what they want it to be by how much migration they bring in.</p>
<p>I notice now that Hong Kong, which was supposed to be the great example on the old Milton Friedman &quot;Free to Choose&quot; television show about free markets and so on: no one has a minimum wage. I'm sure Friedman especially, who lived in this city, San Francisco, would probably turn over in his grave.</p>
<p>But I can understand the potential problem of income inequality, because the per-capita income in Hong Kong is more or less at developed-country levels.</p>
<p>So there are a few countries like that, Singapore and so forth, in the middle. And we saw a few very unusual things. Our first data collection in Russia was immediately after their crisis. So their wage rates were abnormally low. I mean, it's quite a remarkable fact, I think well-known, that the Russian collapse back in the early part of the 2000s - that actually did lead to phenomenal real wage declines in Russia itself. Now, that's come way back. So they're still a poor country, but nothing like the level that they were operating at when we first started to collect data.</p>
<p>So that's one lesson, which is there's kind of a distribution, the Western European countries, US, Canada, and Japan being in that kind of OECD group. And they really do get that minimum level of wage rates that I think defines, in some ways, a developed country.</p>
<p>The second thing we learned was - the growth in China and India, that's not a fake. In China, a person can more or less buy twice as many...now, they couldn't buy very many to start, but they can buy pretty much twice as many Big Macs as they could in the early 2000s. So they are growing at a phenomenal clip. India is not growing as rapidly, but fast.</p>
<p>If we saw growth rates in the Western European countries, or the US, for wage rates that were like what you're seeing in China and India, it would just be incredible. Our income-distribution problems would go away almost immediately. I mean, to see wage rates growing at 8, 9, 10 percent a year, real wage rates, that you can measure them either in dollars...</p>
<p>I realize, in the last year or two, the Big Mac index has another huge attraction, which is it's completely impervious to exchange-rate changes. You don't have to worry about that. So when, for example, the Czech crown got out of control, a Big Mac would be enormously expensive, by American standards, in Prague. But of course, it's not so expensive anymore, because the exchange rates have collapsed relative to the dollar. But that would not affect that comparison of the wage rate to the Big Mac.</p>
<p>So, in a way, even though there are disadvantages to using that as your purchasing-power-parity measure, there is an advantage in that it keeps you completely immune from exchange-rate fluctuations that don't seem to be&hellip;like the pound. I mean, when the pound was at two dollars, it's quite a different story from when it was $1.50. But those don't affect the ratio. So there's a huge amount of growth.</p>
<p>I think what we're interested in, that there are several different aspects to the economics of this. One is it does let you get a handle on some very basic appraisals of public policy, and especially in places where it's so hard to get any real information that seems comparable.</p>
<p>I think this is one of the problems&hellip;as you wander around the meetings here in San Francisco, this is one of the problems in the comparisons between the current period and the Great Depression.</p>
<p>Most people don't know this, but the US government didn't have official unemployment figures until the late 1940s. So the unemployment figures that are talked about in the Great Depression are constructed, mainly by Stanley Lebergott, who constructed them by trying to estimate employment and then extrapolating what he thought happened to labour-force participation. So they're not real numbers, in the sense that we would call them today.</p>
<p>So I feel a little bit like that. I'm imagining here, in developing countries whether people always wonder, &quot;Can you trust those Chinese data? Can you trust those Indian data?&quot; These are data that they're not as broad-scaled as you like, because they don't cover all parts of the income distribution, but they do cover that one part. So we'll be able to see now what happens over time within countries like China and India.</p>
<p>In a sense, it's an incredibly optimistic deal because...well, we'll see what happens in the next two years, but this is probably the convergence. To have half the world's population grow at such a fast rate, even at the bottom, there's probably not been a period, certainly in my lifetime, when that's happened before.</p>
<p><strong>Romesh</strong>: Are there some issues of comparability, though? I mean, you talk about the standard McDonald's being pretty much the same. But are there issues about, say, the status of a McDonald's job in different countries?</p>
<p><strong>Orley</strong>: Well, that's true.</p>
<p><strong>Romesh</strong>: I mean, in the United States and Europe, they might think that it's a fairly low-status job, but perhaps it's a higher status in China and India. And are there differences in productivity, or is productivity of a typical McDonald's worker pretty much the same?</p>
<p><strong>Orley</strong>: Well, you have raised two questions, both of them good ones. I think it's true that a McDonald's job is... because so many of the jobs are at a relatively low level with low pay in a developing country, a McDonald's job probably doesn't have the same hamburger-flipping aspect to it; although certainly the workers that I've observed, typically they don't look much different from the workers you'd see in the US: very young. But it is probably true that that's the case.</p>
<p>To the extent that it is the case, it does mean that it could be the case that the workers at a McDonald's in a developing country are really better, relatively speaking, than the workers in a McDonald's in the US. What that would tend to suggest is that whatever gaps I measure are actually understated, that the gap is even larger.</p>
<p>I actually don't think that's a very big issue because I think that the McDonald's wages is kind of &quot;the wage.&quot; In other words, there is a wage out there... I don't mean by that it's to the penny, but just like there is the price of a cup of coffee or the price of a haircut in a given place, there is a kind of going rate for that kind of work.</p>
<p>The biggest problem I think is not with the developing countries, because in the developing countries there really is a market for that kind of work and the government doesn't involve itself in it very much. McDonald's, which has a lot of control of the stores that they're engaged in, the one thing that they don't control is the wage rates at which they hire people.</p>
<p>The place where we get the trouble, I think, is in Europe where there are either giant taxes on wages that are used to fund health insurance benefits and other things, or where there are very, very high minimum wages. That can happen in some countries. If you have a minimum wage that's very, very high then clearly that's not a market rate that you're looking at. Even though it does tell you something about how well-off someone is who works in a McDonald's there, it's not really telling you about what we want to know about, which is the core level of work. So I think the biggest problem actually is in developed countries.</p>
<p>And we've made a little bit of effort to try and make adjustments for income taxes and social security benefits and things like that in some of the European countries. Once you do that you're starting to move away from the pure data aspect of what we've done so I'm less comfortable with it. But that's where I think the biggest problem is.</p>
<p>Now the other question you have: productivity. I occasionally do hear people say that you might get your burgers quickly, or something. For example, I've heard that said about Barcelona. Someone went into a McDonald's and observed it and thought it was a little slower. It could be a little bit different. I observed them too and they didn't seem that different to me, but it could be a little different.</p>
<p>One aspect of this which we haven't paid much attention to but which I think is important is the extent to which there is competition in the McDonald's market. We don't think about that too much, you wouldn't in a country like England or the US, because we don't imagine that there would be any sense in which there would be less competition. In other words, here we have Roy Rogers and Wendy's and Burger King and whatever. They're all kind of in competition.</p>
<p>They're not identical but they're... Subway, I went to a Subway yesterday: that's a kind of a healthy version of McDonald's. Actually, the BLT is pretty good. But there are some countries where there doesn't seem to be very much competition. I noticed even in Barcelona there don't seem to be any Wendy's and Burger Kings and I'm quite unclear why that is.</p>
<p>I do know that in some countries you'll observe that the price-to-wage gap is enormously higher than others. Not very many. A country like Colombia, for example, in Latin America, seems to have incredibly high prices for a McDonald's compared to the wage rates. That suggests a kind of a price mark-up. Maybe there's been some restriction on competition.</p>
<p>What bothers me about that is not so much that it would tell us anything about the wage rate. It wouldn't affect the way we measure the wage rate. But it would affect using the Big Mac price as a way to deflate the wage rate, because the Big Mac price is really way, way too high. In Colombia, for example, I've always thought that wage rates measured in Big Macs just don't look high enough. One interpretation of that is that Big Macs are too expensive. I think that's also an issue, but it probably more is on the product market side than it is on the labour market side.</p>
<p>In productivity there could be some differences but I don't think there's very many. The way they set the thing up at McDonald's is so routinized. You can understand the reason for it. They don't talk about this. They're not going to talk about food safety. That's not what people talk about, they talk about how good the food tastes. But the reality is that food safety is absolutely paramount.</p>
<p>The one thing that can bring down a company in a development country is like Coca-Cola when they got trouble in India. You just can't allow that. That's the brand, and that brand has got to stick by that. So I think on the whole you get pretty standardized operating mechanisms. Whether that's a good or a bad job is probably about the same as it would be is it a good or a bad job in the US.</p>
<p>I think what we want to do with it is to let us say something about total factory productivity and its growth. It's an amazing fact...if you take a measure of total factory productivity like one, for example, from Hall and Jones...actually, I can predict their measure of total factory productivity with an R-squared of 0.8 just using the wage rate for a person at McDonald's. It's amazing. I don't have any GNP accounts and I can still tell you pretty much how well off the country is in terms of total factory productivity.</p>
<p>One thing it would allow us to do is to take models where people want an index of total factory productivity, try to extract from that. We could probably use this as a way to benchmark that, see how well it works. That then raises questions about trade.</p>
<p>One of the things that seems to show up in what we're looking at is: some models suggest there is almost a dichotomy. The countries that have high capital-labour ratios are the ones that I pretty much mentioned at the beginning - Japan, Western Europe, US, Canada. They're the ones that have those relatively high wage rates for McDonald's workers. Then you move to countries with lower capital-labour ratios and you see these lower wage rates.</p>
<p>Some of it looks almost as if there's a discontinuity. It may very well be that, and I think there's some other evidence for this too, that in the developed countries they're really in one cone of diversification. They're trading products that are things that all of them can more or less make. Then there's another cone of diversification that the poor countries are in, where they're making a completely different set of products. Until they can move to those higher capital-labour ratios that may be one of the secrets to how wage rates will have to grow.</p>
<p>But I'm put off by the idea that it's totally dichotomous just from the fact that the Chinese have grown at such a rapid rate in the last 10 years. It'll be interesting to see how much it slows down during the next year or two. We'll see.</p>
<p><strong>Romesh</strong>: Orley Ashenfelter, thank you very much.</p>
<p><strong>Orley</strong>: My pleasure.</p>