Is the Chinese skyscraper boom excessive?

ABSTRACT

According to the Council on Tall Buildings and Urban Habitat, more than half of the skyscrapers completed in recent years, i.e., buildings taller than 200 meters, were in China. In fact, China has been undergoing a dramatic skyscraper boom since the early 2000s. Some claim that this boom is excessive. We examine this claim through two sets of analyses. First, we test whether developers or cities engage in height competition to build taller buildings. Theoretical models demonstrate that such competition results in excessive building height. We find some indications of height competition among Chinese cities. We find concrete evidence for height contests within cities. Second, we build a panel regression model to identify cities where growth in skyscrapers exceeds the trajectory predicted by economic fundamentals. Even though such a model cannot definitely establish whether these cities are building skyscrapers excessively, the results are useful in informing policymakers and developers of potentially higher risks in these cities. Our results suggest that Tier-1 cities are likely less risky than lower-tier cities.

Acknowledgments

We thank Sanghoon Lee, Tsur Somerville, Sheridan Titman and audiences at the 2018 International AREUEA Conference, 2018 RUSE in China, Southwestern University of Finance and Economics, Shanghai University of Finance and Economics, for comments and suggestions. Discussions with Sanghoon Lee inspired us to start this project.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. A literature exists that connects skyscraper boom with economic recessions. It started with Lawrence (1999). Further developments include Thornton (2005), Lawrence (2012), and Barr et al. (2015).

2. It is a bit controversial to include Tianjin in Tier-1. We did this for two reasons: (1) Tianjin is among the first three directly administered municipalities together with Beijing and Shanghai; (2) In recent years, the initiative to develop Binhai New Area added new impetus for Tianjin to revitalize its position in the urban system in China.

3. Table A1 shows the list of cities illustrated in this Figure. They are cities that had built at least one skyscraper by 2018.

4. A caveat exists for using the time-series variations in ANOVA tests, namely height gaps tend to be persistent or serially correlated over time. ANOVA test, however, implicitly requires the sample to be i.i.d. Our test using city-level data, reported later, should be more reliable as cross-sectional correlation tends to be weaker.

5. Multivariate Mean Tests and Pair-wise T-tests are also employed to compare these mean height gaps. Similar or stronger results are obtained as these tests do not adjust for multiple comparisons.

6. The addition of 1 ensures that logarithm height is still defined for city-years with no skyscrapers.

7. The estimates of a Tobit fixed effects model are biased due to some technical issues. We report this result anyway as a robustness check.

8. There are some variations in handling of time trend. RE, FE, Tobit-FE all use year dummies. RE-AR1 and FE-AR1 do not include year dummies. RE-Random Trend accounts for a city-specific random trend. This time component is included in making predictions.

9. Due to space limitations, we do not report regression coefficients for other models. These results are available upon request.

10. We thank an anonymous referee for pointing out this caveat and for suggesting several solutions, which we adopt wholeheartedly.

11. We do not report the correlation coefficients or residual rankings for models with alternative lag periods due to space limitations. These results are similar to results here. They are available upon request.

12. Table A3 reports similar analysis for alternative models with different lag periods. Note that the prediction horizons in Table A3 vary with lag order because we always reserve economic fundamentals data in 2013–16 for prediction. We do not discuss Table A3 because it presents similar findings.

Additional information

Funding

This work was supported by the Humanities & Social Sciences Research Fund (NUS) [No. R-297-000-132-646].

Notes on contributors

Qiang Li

Qiang Li is an Associate Professor at Deakin Business School. Prior to joining Deakin University, Li worked at National University of Singapore for 7 years. He also worked at Shanghai University of Finance and Economics for 5 years. His research and teaching fields are urban economics, real estate finance and economics, and urban planning. He has published papers in Journal of Urban Economics, Regional Science and Urban Economics, and Journal of Housing Economics, among others. His current research interests include ethnic diversity and urban housing markets, the economics of skyscrapers, and housing cycles.

Linlin Wang

Linlin Wang is an Investment Manager in China Jinmao Holdings Group Limited. China Jinmao is a leading real estate developer, operating in more than 30 major cities in China. Wang is responsible for managing its land investments in the Southwest region of China. She specializes in investment analysis, urban development analysis, and housing market research. She received a master’s degree in urban planning from National University of Singapore.