VoxEU Column Development Productivity and Innovation

Local capital scarcity and industrial decline caused by China’s real estate booms

In the new century, China’s large economy features many local real estate booms originating in insufficient land supply. Using a panel of 900,000 Chinese manufacturing firm-year observations with matched firm locations, this column quantifies the causal effects of local real estate booms on local firms. It demonstrates how the diversion of local savings into the real estate sector in cities with real estate booms exerts a large toll on other local industries through higher costs of capital, underinvestment, real wage decreases, and industrial decline.

China experienced a large number of real estate booms across many cities in the first decade of the 21th century. A simple cross-sectional comparison in Figure 1 reveals how uneven this phenomenon was. Ranking cities by their initial residential real estate price index in 2003 (blue line), we see that cities with near-identical real estate price levels in 2003 show starkly different price levels seven years later in 2010. At the 10thpercentile, the housing prices increase by 52%, compared to 182% at the 90thpercentile of local index growth. However, as real estate booms can influence economic activity and vice versa, any directional (or causal) analysis requires us to find a deeper (exogenous) determinant of real estate booms not itself influenced by industrial growth.

Figure 1 Real estate booms across Chinese cities, 2003-2010

What causes local real estate booms?

Shiller (2006) highlights that zoning or other supply restrictions contribute to higher local real estate prices. This is certainly true in China, where the local land supply (in combination with the local price elasticity of supply) features a high correlation with local housing price inflation, as shown in Figure 2. Land supply for construction is a local government monopoly in China and is therefore subject to the uncertainties of planning, coordination, and implementation by different government agencies. This can cause significant delays in land development because of policy conflicts across agencies, central government interference, or local protest. In many cities, real estate developers obtain only a fraction of the ‘planned’ land supply in any given year. On the other hand, pre-existing local industrial and economic conditions appear to have no explanatory power for the actual local land supply. 

We therefore argue that China’s land supply process represents an exogenous source of variation for local financial conditions (Hau and Ouyang 2018). Tracing the effects of local land supply variation allows causal statements to be made about the effect of real estate booms on the competitiveness of the local industrial sector.

Figure 2 Local land supply and local price elasticity as the driver of real estate booms

Industrial decline caused by local credit scarcity

China’s cities represent financially closed economies because of pronounced geographic segmentation of China’s credit market. Regulatory constraints prevent cross-city lending by local banks. Private firms in particular depend on local borrowing and access to local savings. Yet, local real estate booms divert local saving into the land-development and construction sector and can create capital scarcity for the rest of the local economy. 

In our study (Hau and Ouyang 2018), we use a panel on 900,000 Chinese manufacturing firm-year observations with matched firm locations to quantify the causal effects of local real estate booms (traced back to land supply variation) on local firms. A 50% relative increase in a city’s real estate price (due to a shortage in local land supply over the period 2002-2007) increases the borrowing costs of firms by almost a full percentage point annually. It also reduces the share of firms with bank credit by 9% on average. 

More interesting still are the real effects on long-term firm growth. Here, a 50% higher real estate price reduces the average corporate net investment share by 7.3 percentage points, which represents a 34% reduction relative to the sample mean of 21.4 percentage points. The long-run relative output decline amounts to a staggering 35.5% of value-added output. Total factor productivity features a relative decline of nearly 12% for the average manufacturing firm over five years.

State-owned enterprises versus private-owned firms

Firm characteristics clearly matter as they affect bank credit access. Firms with large fixed assets and state-owned enterprises enjoy privileged credit access to the ‘big five’ national banks. This greatly attenuates their exposure to the capital scarcity induced by local real estate booms and reduces their relative competitive decline. Panels A and B in Figure 3 illustrate this difference between state-owned-enterprises and private firms, for which the reduction in valued-added output under real estate booms is much more pronounced.

Figure 3 Relative industrial output decline during real estate booms, 2002-2007

Note: SOE: State-owned enterprise

The negative effects of real estate booms also show up in firms’ financial data: firms located in cities with real estate booms show lower returns on assets and higher financial leverage. In addition, a relative 50% increase in real estate prices raises the market exit rate by 3.3 percentage points a year.

How can real estate booms be so harmful?

Factor price externalities across sectors are well documented in international economics. The so-called Harrod-Balassa-Samuelson effect explains why high productivity in traded products inflates local real wages and makes all non-tradeable goods expensive. Thus, hamburger prices at McDonald’s become a good proxy for industrial productivity across countries. There is much less evidence for factor price externalities working through the local interest rates. Chakraborty et al.(2018) find that local real estate booms in the US also increase corporate borrowing costs —albeit at a more modest scale than in China.Similarly, Martín et al.(2018) find that the housing boom in Spain raised credit demand from the real estate sector and crowded out bank credit to other firms.

Yet, the interest rate externality is potentially more pernicious because causality runs from the non-tradeable construction sector to the industrial sector. Unlike the non-tradeable sector, the industrial sector cannot transfer higher factor costs into higher product prices because of strong national and international product competition. Hence, a competitive setting implies underinvestment and industrial decline in locations with real estate booms.

The Harrod-Balassa-Samuelson framework also predicts declining industrial wages if underinvestment reduces the marginal productivity of labour. The Chinese data confirms this prediction: a 50% higher real estate price tends to depress industrial wages by approximately 16%, during the period 2002–2007. This relative decline of the real wage clearly contrasts with a Dutch Disease scenario, in which a large sectoral shock (often in the oil industry) inflates wages throughout the economy.  


The institutional features of China’s real estate market generates large random variation in local land supply, which in turn explains the magnitude of local real estate booms. Such real estate booms are pernicious in segmented credit markets because they can divert a substantial share of local savings into real estate development, while the tradeable sector is subject to higher capital costs and credit rationing. In China’s geographically fragmented banking system, such exogenous capital cost shocks exert an economically significant adverse influence on local industrial competitiveness through lower corporate investment rates and lower productivity growth. 

Creating a more integrated capital market within China should yield large economic benefits, even if China’s participation in the global capital market remains tenuous. There are also lessons for Europe’s capital market integration. Recent research by Meier (2018) shows that Europe’s regulatory integration in 2004–2009 has reduced barriers to cross-country lending and increased external finance and investment for Europe’s publically listed firms. In light of the Chinese evidence, cheaper and more diverse capital sourcing by Europe’s corporate sector should have substantial long-run benefits. Restricting market access of British or Swiss bankers to Europe’s corporate lending market represents a policy harmful to competitiveness and long-run growth.


Chakraborty, I, I Goldstein and A MacKinlay (2018), “Housing price booms and crowding-out effects in bank lending”, Review of Financial Studies 31(7): 2806–2853. 

Hau, H, and D Ouyang (2018), “Capital scarcity and industrial decline: Evidence from 172 real estate booms in China”, SFI Working Paper 18-38. 

Martín, A, E Moral-Benito and T Schmitz (2018), “The financial transmission of housing bubbles: Evidence from Spain”, CEPR Discussion Paper 12999.

Meier, JMA (2018), “Regulatory integration of international capital markets,” SSRN Working Paper.

Shiller, RJ (2006), “Long-term perspectives on the current boom in home prices”, The Economists' Voice 3(4).

840 Reads