Land prices in China have been rising very rapidly; the growth has been as high as 16% per annum over eight years. This has been caused by both demand and supply factors. While it’s not realistic for this rate to continue, don’t expect a slump either, says Wharton real estate professor Joseph Gyourko.

Gyourko, along with colleagues at the National University of Singapore and Tsinghua University in China, has been studying the China housing market for some time now. It is not easy because data is difficult to come by, but the team has put together an index of housing land prices in China and is now planning a second phase to the research, Gyourko tells Knowledge at Wharton.

An edited transcript of the conversation follows.

Knowledge at Wharton: You have collaborated with your colleagues at the National University of Singapore and Tsinghua University in China on a really pioneering project, where Wharton works with other universities to create indices of housing land prices in China. Could you tell us what are some of the most significant takeaways?

Joseph Gyourko: These are real price indices for land values in China; not housing, but the land underneath the housing. The takeaways are: At a national level, there has been remarkably strong price growth in land values. We have eight years of data from 2004 through the first half of 2013, and the compound annual growth rate (CAGR) is just over 16% per year. So that’s really a strikingly large number. And that’s the aggregate over all of China.

There is a lot of variation across regions and cities, however. In that sense, China is somewhat like the U.S.: It’s a very big country, and you do not get the same price growth everywhere. At the city level, Beijing has been growing astronomically — 22% CAGR since 2004. Markets like Jian are very different: 4% to 8% in some of the central and western cities of the country. So one size does not fit all in Chinese land markets.

Knowledge at Wharton: As you looked at all that data, what was the biggest surprise?

Gyourko: I’m not so sure I was surprised. I expected high price growth rates. One of the things that’s clear is you can’t continue to have 16% per annum growth, because prices will go to infinity fairly quickly if you have another decade of that. But I don’t know that I was surprised by that. I think what probably surprised me most is how varied the results are across cities.

Knowledge at Wharton: What implications would you say your findings have for international residential developers who may be interested in opportunities in China?

Gyourko: I think you have to be worried about how high the price growth has been in the past and you have to carefully consider how long you think it can continue at its present high trend rate into the future. You have to be cautious. This is not advice that there’s an obvious bubble, that there’s going to be a crash, because I don’t know that for sure. What I think I do know for sure is that I will be very surprised if we have another eight years of 16% CAGR.

“You have to be cautious. This is not advice that there’s an obvious bubble, that there’s going to be a crash, because I don’t know that for sure.”

Knowledge at Wharton: I’d like to come back a little later to the 16% number, which I think works out to almost 200% over the course of your study. Just to set the background, can you help explain for our audience a little bit about how China’s land market differs from the market in other countries? And what challenges did that pose for you as you went about your research?

Gyourko: China’s really very unique in this sense. If you study housing prices across the world, including in the U.S., you never see land values. You don’t see vacant land being sold in the data. So what you see are house prices. The price of my house or your house in a price index is the sum of the land value and the improvements, the building, itself.

China’s different. The Communist Party of China owns all of the land in urban areas. The local governments do. So if you’re a developer and you want to put up a private residential complex, you have to purchase the land from the local government. What you purchase is actually a land lease. You purchase the rights to use the land for up to 70 years for residential development, which is what our index is all about. We only cover land sales that are going to be used for residential development, not commercial development.

So the prices and the indices we’re reporting are technically those for leasehold estates. Now, the way it works in China is the developer makes a one-time, lump-sum payment for the use rights for up to 70 years, and we treat that one-time payment as the sales price. We think that’s a pretty reasonable assumption, largely because the usage rights are for multiple decades and the present value of anything 50 to 70 years from now is pretty close to zero. So we think that’s a good assumption now; in 25 years, that may not be such a good assumption, because you may be near the end of some of these usage rights. But that’s what makes it very different in China. China’s virtually unique as far as I can tell in that respect.

Knowledge at Wharton: And what challenges did that pose as you went about your research?

Gyourko: Well, there were many challenges. One is that data accumulation in China is very difficult. I’m sure the central government has a lot of data that they don’t release and aren’t going to release to academics like me. So what we did was to scrape the websites of the local land-use authorities. There was a national law passed well over a decade ago that required the posting of the winning bidders in land auctions. It also required that the land be transacted by a visible auction. It was an anti-corruption move. You can imagine before this law came into being, there could be all types of side payments between developers and local officials.

So we have a number of people who literally scraped the websites of the 35 local land authorities from whom we collect data. You might imagine a bunch of students, graduate students and undergrads, particularly in Beijing, who are cleaning the data and making sure it makes sense.

Knowledge at Wharton: Let’s come back now to the question of the rate of growth — 16% a year, 200% over the course of eight years. Just to put those numbers in perspective, do we know how those kinds of price increases compare to housing land price increases in other countries? And what are some of the factors driving that increase in China?

Gyourko: We cannot directly compare them to land prices in other countries, simply because we do not see land value measured consistently in other places. A couple of comments, though, on this: One, in an eight-year period, for prices to go up 200% is really quite extraordinary. It’s much higher than the government’s published data on housing prices, which leads me to believe that they are understating the house price rises, because land has got to be a big component of value in China.

“There is an incredible move from rural areas into urban areas in China, so demand [for housing] is very strong.”

The other thing: If you go to my website and download the data, you will see a couple of different periods. In other words, this 16% figure is not a smooth rise. The first really big increase in aggregate land values in China occurred in 2007. In three quarters, from the first to the third quarter of 2007, real land values in aggregate across the 35 cities we’re tracking rose 71%.

Following that dramatic rise, real prices fell 34% until the first quarter of 2009. From 2009 to the end of 2010, prices more than doubled. They went up 108% according to our index in real terms. That period coincides with the great Chinese stimulus. So it is pretty clear there are a number of drivers. Some of this is government-policy related, but some of it is clearly demand-side related. There is an incredible move from rural areas into urban areas in China, so demand is very strong. In some of these areas, although this is a subject of future research, it’s unclear how much new supply is being created. My two co-authors and I are working on a project to measure supply in these markets.

When you look at the data, it is clear that government policy matters in China. It matters a lot to their local land markets.

Knowledge at Wharton: Apart from the stimulus, did you see any other political factors at work?

Gyourko: Not in our indices. The indices are just created from the data. When you look at the nature of how households in China think about owning, many of them pretty clearly view it as a store of value in a way that you and I in the U.S. don’t. I own a home because it provides shelter for me and my family, my kids get to go to the school district in which my house is located and the like. In China, investment restrictions are much more severe than they are in the U.S. The typical household there cannot diversify by buying the S&P Index. A lot of households are buying because they view this as a good store of value, a hedge and the like. So there are different motivations in China than there are for most U.S. households in particular.

Knowledge at Wharton: You also referred earlier to the fact that land prices in the Beijing area increased much faster than other parts of the country? What explains some of these regional variations?

Gyourko: A couple of things: One, we do know that demand growth, that is in migration to Beijing and particularly some of the East Coast cities, is very strong. So, partly, it’s demand. What I also suspect but don’t know — this is future research — is that it is probably harder to build in some of these coastal markets. But we actually don’t have good measures yet about how much new supply is coming up in each of these markets.

My economics gut [tells me that] it’s both supply and demand. We can pin down demand. We can’t yet pin down the contribution of the supply side to that remarkable growth.

Knowledge at Wharton: One really fascinating phenomenon in China’s housing market that I’ve heard people talk about is that developers acquire land from the government as you explained. Then they just go ahead and build without really estimating what the real demand is from consumers. This has led to the creation of these so-called “ghost cities” [that have] a lot of houses with no one living in them. Did you take some of these factors into account, and what do you think of this phenomenon?

Gyourko: The phenomenon definitely exists. You can see it documented. But in the 35 markets we covered, there aren’t ghost cities per se. China has well over 100 cities or urban areas with more than one million people. We’re looking at 35 large ones. The ghost cities phenomenon is not a major issue in them. You don’t see entire developments or swaths of Beijing, for instance, which are vacant. So while that phenomenon clearly exists in some parts of China, it tends to be in outlying areas and not in the major markets we’re working in.

Knowledge at Wharton: So, in other words, it didn’t affect your data and didn’t have an impact on your findings?

Gyourko: No. We don’t really see that phenomenon a lot. We’re looking and trying to measure vacancy rates. There is no published vacancy rate series in China. It’s one of the things we’re trying to do in current research. We’ve looked first at Beijing. There’s not a lot of vacancy in Beijing. In other words, it doesn’t look like there are ghost developments in that city at all. Maybe we’ll see them in some of the other 34 markets that we track. But, thus far, we just don’t see that in the markets we’re tracking.

“China has well over 100 cities or urban areas with more than one million people. The ghost cities phenomenon is not a major issue in them.”

Knowledge at Wharton: Did you find any evidence of a land bubble in China?

Gyourko: Well, I don’t know what a bubble is per se. I don’t believe you can have another eight years of 16% per annum growth. Does that mean it’s a bubble? I don’t know. It means prices are very high. They’re high relative to the incomes of the people buying the housing units that the developers are putting up on these plots of land. You cannot have this rate of growth continuing. I don’t know whether it’s a bubble. It’s back to my advice to developers and investors. You have to be very cautious. Look at the local demand and supply fundamentals. I think there are some markets where they clearly still make very good sense and others where I’d be really worried about the sustainability.

Knowledge at Wharton: As you look to the future, what impact do you think the economic slowdown that we see these days in China could have on land prices?

Gyourko: It’s going to come through both supply and demand. I’d be more worried about supply effects; that is, developers not cutting back sufficiently on development to accommodate the lower trend growth rate. I suspect there’s going to still be rural migration into urban areas in China, because the income differences between a factory job and being a farmer are very high, and you’re going to see that movement. The question is as the trend growth rate in China slows, will you see development slow enough to accommodate it?

I think the biggest risk will be from oversupply in markets. And we don’t have good measures of oversupply yet, although that’s something we all need to be thinking about going forward.

Knowledge at Wharton: You spoke earlier about some future research. What’s the next phase of your study going to be, and what are you looking at next in China?

Gyourko: A couple of things. One, we will continue to update the indexes. Every quarter, we’ll update the national index. The city and regional indexes only come out every six months or every year. That has to do with data availability issues. So that’s one. We’ll continue to update this. Second, the whole research team is trying to get a handle on how much new supply has been created and how much you need given that demand growth is still going to be substantially positive. But the difference between 7% to 8% growth and 12% growth is a lot.

So the next big project is what has to happen to supply to make the market stay sane or be sane in a world where trend demand growth is 50% to 60% of what it used to be.

Knowledge at Wharton: In conclusion, if there were some CEOs of international real estate companies asking for advice about China and the land market there, what would you tell them based on what you have learned so far?

Gyourko: Well, one, just be very careful. The data in China is hard to come by and hard to create. So make sure you’re looking at real data and good data. Be realistic about what you think demand growth is going to be. And, as carefully as you can, try to measure the supply competition. Because one of the things we learned in the U.S. is that supply matters, not just demand.