Unleashing the real estate market: assessing the impact of relaxed housing purchase restriction policy on house prices in China

ABSTRACT

This study examines the effects of the relaxed housing purchase restriction (RHPR) policy on house prices in China, using panel data from 41 cities between January 2021 and June 2022. The study finds that the RHPR policy has a positive effect on the existing housing market, contributing to its stabilization. However, the impact on new house prices is unclear. Mechanism analysis indicates that the policy effectively boosts house demand and fosters positive market expectations. The study also finds that the policy has a greater impact on second-tier and northern cities. Additionally, the research indicates that the policy has stimulated house demand among non-registered populations and in non-restricted purchase zones. These results emphasize the importance of timely adjustment and optimization of real estate policies to maintain a stable and sustainable market.

KEYWORDS:

• Real estate market
• Relaxed housing purchase restriction policy
• House prices
• Multi-period DID

JEL CLASSIFICATION:

• O18
• R21
• R23
• R28

I. Introduction

In China, the real estate market has been a crucial factor in driving national economic growth and improving people’s quality of life. To ensure the sound development of the real estate market and stability in house prices, the housing purchase restrictions (HPR) policy has been implemented, which is recognized as the most effective measure in stabilizing house prices (J. Li and Xu 2016). However, due to the impact of the COVID-19 pandemic in 2020, real estate market regulation shifted its focus from managing rapid price increases to curbing rapid price decreases. As of March 2022, some Chinese cities have adopted the relaxed housing purchase restriction (RHPR) policy to boost the real estate market and stabilize house prices.

Extensive research has examined the HPR policy’s effectiveness in impacting house prices (Du and Zhang 2015; V. J. Li, Wui Wing Cheng, and Se Cheong 2017; J. Li and Xu 2016; Y. Li et al. 2020; Somerville, Wang, and Yang 2020). However, there is a lack of research investigating the influence of the RHPR policy on house prices. Moreover, existing research primarily focuses on its influence on housing price equilibrium and sales outcomes (Deng, Gyourko, and Li 2019; Jia, Wang, and Fan 2018), with limited knowledge regarding its effects on demand factors such as consumer preferences and pricing strategies of real estate developers. Therefore, this paper aims to assess the impacts of the RHPR policy on house prices as well as understand its underlying mechanism. Additionally, this paper also examines the influence of specific policy details, such as restriction zones and populations.

By treating the RHPR policy as a quasi-natural experiment (Lu, Zhang, and Hong 2021; Wu and Li 2018), we evaluate its impact on house prices. Our research utilizes data from January 2021 to June 2022, encompassing 41 large and medium-sized cities across China. As shown in Figure A1, there are two groups: treatment group consisting of 18 cities, and control group consisting of the remaining 23 cities. The housing price index, sourced from the official website of the China National Bureau of Statistics (http://data.stats.gov.cn/), serves as primary independent variable. It comprises two components: the index for existing house prices and the index for new house prices, representing changes in both segments over a specific period.

II. Methodology

Our research methodology employs the following experimental design: By considering the RHPR policy as a quasi-natural experiment and taking into account its gradual implementation across various cities, we employ a multi-period difference-in-difference (DID) model (Callaway and Sant’anna 2021), represented by Equation (1)(1) $\mathit{Pric}{e}_{\mathit{it}}=\mathit{\alpha }+\mathit{\beta Trea}{t}_i\ast \mathit{Pos}{t}_t+\mathit{\gamma }{X}_{\mathit{it}}+{\mu }_i+{\lambda }_t+{\epsilon }_{\mathit{it}}$(1) .

(1) $\mathit{Pric}{e}_{\mathit{it}}=\mathit{\alpha }+\mathit{\beta Trea}{t}_i\ast \mathit{Pos}{t}_t+\mathit{\gamma }{X}_{\mathit{it}}+{\mu }_i+{\lambda }_t+{\epsilon }_{\mathit{it}}$(1)

In this equation, the variable $\mathit{Pric}{e}_{\mathit{it}}$ denotes the monthly housing price index for city i in month t, serving as the dependent variable. $\mathit{Trea}{t}_i$ is a group dummy variable and is equal to 1 if a city i belongs to the treatment group, and 0 otherwise; $\mathit{Pos}{t}_t$ is a time dummy variable with a value of 1 when the month t is after policy implementation, and 0 otherwise. The control variable $X_{\mathit{it}}$ denotes a vector of control variables that encompasses various economic characteristics and living conditions of each city. ${\lambda }_t$ is the time fixed effect and ${\mu }_i$ is the city fixed effect. The random disturbance term is represented by ${\epsilon }_{\mathit{it}}$. The core parameter $\mathit{\beta }$ to be estimated is used to measure the average treatment effect of the policy on house prices.

III. Empirical results

When studying the policy implications on city house prices, it is essential to control variables that can affect house prices. Table A1 presents a set of main variables and control variables which is related to economic characteristics and living conditions in each city. Table 1 displays the baseline regression results based on Equation (1)(1) $\mathit{Pric}{e}_{\mathit{it}}=\mathit{\alpha }+\mathit{\beta Trea}{t}_i\ast \mathit{Pos}{t}_t+\mathit{\gamma }{X}_{\mathit{it}}+{\mu }_i+{\lambda }_t+{\epsilon }_{\mathit{it}}$(1) . The explanatory variable in Columns (1) - (2) is the new housing price index. The results show that the RHPR policy effect on new housing price is insignificant. Columns (3) - (4) present the findings of regressions where the explanatory variable is the existing housing price index. The regression coefficient for the core explanatory variable Treati*Postt is 0.7284, which is significant at the 1% level of significance. These results suggest that, holding all else constant, the average price of existing housing in cities where purchase restrictions have been relaxed increased by 0.73. These findings indicate that the policy may have effect on stabilize the existing housing price.

Table 1. Baseline regression results.

Variables	Price_new		Price_exist
	(1)	(2)	(3)	(4)
Treati*Postt	0.5249 (0.700)	0.3355 (0.6135)	0.7754*** (3.2222)	0.7284*** (3.4031)
Ln_Revenue		0.0010* (1.17136)		0.0005** (2.0165)
Ln_investment		0.0001 (0.5224)		0.0304*** (2.9693)
Ln_PGDP		0.4755** (2.2568)		0.5214*** (3.0322)
Ln_Income		0.0001 (0.0177)		0.0003*** (4.1238)
Ln_Loan		0.7903* (1.7912)		0.6435* (1.9500)
Ln_Pop		0.0044* (1.7015)		0.0027* (1.7597)
Ln_Education		0.1020 (0.8144)		0.0579** (2.3714)
Ln_Hospital		0.2431 (1.3557)		0.1225 (0.6041)
Ln_Vehicle		0.0032* (1.9495)		0.2166 (1.0801)
Ln_Land		0.0001*** (3.8093)		0.0005*** (10.3848)
Ln_CPI		2.8652* (1.8939)		6.0744*** (3.8367)
Interest rates		10.8292 (1.6604)		14.1251*** (5.7765)
Ln_employee		19.3868** (2.1064)		10.9397*** (3.8771)
Constant	102.0488*** （3.4003）	56.2743** （1.0722）	101.5122*** (5.1739)	43.4810** (2.3039)
Individual fixed effects	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes
N	779	779	779	779
R2_a	0.2579	0.4119	0.3249	0.5024

Note: t statistics in parentheses *p < 0.1, **p < 0.05, ***p < 0.01.

We also perform a series of robustness test, including parallel trend test, placebo tests, other robustness test and heterogeneous robustness test. Figure A2 presents the result of parallel trend test, it can be inferred that the parallel trend hypothesis cannot be rejected. Figure A3 presents the results of placebo tests including randomly select cities to serve as the treatment group and randomly chose time points as policy intervention time. Table A2 presents the results of other robustness test, including substituted the dependent variables¹ and substituted the samples. We also conduct heterogeneous robustness test. To assess the extent of bias resulting from multi-period DID (Baker, Larcker, and Wang 2022), we employ the coefficient decomposition method proposed by Goodman-Bacon (2021) and the results are present in Table A3.

Building on the previous analysis, we aim to investigate the mechanism of this policy. We use the sold area of existing house as a proxy variable to analyse the impact of the policy shock on housing demand. Firstly, we use sold area of existing house and the listed area of existing house as a proxy variable to analyse the impact of the policy shocks on housing demand and supply. The empirical results present in column (1)-(2) of Table 2. The results suggest that the policy can drive price of existing house by boosting demand. Secondly, we further analysed the effects of the RHPR policy on people’s expectations, using one-period lagged existing housing price to assess the impact of policy shock. Our empirical result is presented in column (3) of Table 2. This result suggests that the policy can affect house prices by shaping people’s positive expectations. This finding supports our previous empirical results.

Table 2. Regression results of mechanism analysis.

Variables	sold area	listed area	Price_exist_lag
	(1)	(2)	(3)
Treati*Postt	0.1982** (2.3908)	-0.0350 (-0.3341)	0.5416** (2.5090)
Constant	0.0197 (0.0027)	33.3449*** (3.6114)	45.3601** (2.3835)
Control variables	Yes	Yes	Yes
Individual fixed effects	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes
N	779	779	779
R2_a	0.3198	0.2499	0.4853

Note: t statistics in parentheses *p < 0.1, **p < 0.05, *** p < 0.01.

To evaluate the impact of city heterogeneity on policy effectiveness, we examine the conditional effects of the policy by considering both city level (second-tier and third-tier) and regional division (northern and southern cities, divided by the ‘Qinling and Huaihe line’). The regression results are presented in Table A4, specifically in columns (1)-(2) and (3)-(4). The findings demonstrate the significant effect of this policy in second-tier cities and southern cities.

We further analyses the restricted zone and restricted population. Firstly, we used the Baidu Migration Index to analyses the relaxed purchase restriction population. The results of our analysis are shown in column (1) of Table A5. The empirical results suggest that the policy can influence existing house prices by increasing the demand for housing among the unregistered population. Secondly, we analyses the relaxed purchase restriction zone. Based on their location, we classify the cities into three zones, including Class I, Class II and Class III zones.² Our findings are presented in columns (2)-(4) of Table A5. Our empirical results suggest that policy has a significant effect in Class II zones, but not in Class I and Class III zones.

IV. Conclusion

Using panel data of 41 cities in China spanning from January 2021 to June 2022, this study investigates the impact of the relaxed housing purchase restriction (RHPR) policy on house prices and explores the underlying mechanism. This paper presents the following conclusions. Firstly, the RHPR policy is effective in stabilizing existing house prices to a certain extent. Secondly, through mechanism analysis, it is found that relaxing purchase restrictions can effectively stimulate purchasing demand and unleash market vitality. Thirdly, the policy effects are more pronounced in second-tier and northern cities as revealed by heterogeneity analysis. Fourth, through further analysis, both the expansion of purchase zones and the relaxation of restrictions on specific populations contribute to price stabilization.

Based on the aforementioned findings, this paper provides the following recommendations. Firstly, it is recommended to continue the RHPR policy to stimulate demand, especially in second-tier and northern cities where a noticeable pull effect on existing housing consumption. Secondly, it is necessary to implementing the ‘guaranteed delivery’ loan support scheme to stabilize new house prices. Thirdly, it is recommended to gradually remove the constraints imposed by the household registration system as a means to facilitate intercity population mobility and stimulate demand. Lastly, while the RHPR policy can partially stabilize house prices, it does not warrant a complete elimination of the housing purchase restriction policy.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/13504851.2024.2332587

Additional information

Funding

This work was supported by the National Social Science Fund of China [grant numbers No.20XTJ005]; the Central Government Funds for Guiding Local Science and Technology Development [grant numbers No. YDZX20216200001876]; the Industrial Support Program for Higher Education Institutions in Gansu Province [grant numbers No. 2022CYZC-55]; and the Science and Technology Program of Gansu Province in China (“Excellent Doctorate Program”) [grant numbers No.23JRRA1190].

Notes

¹ The existing housing price can be further divided into three categories based on size: small size (<90 m²), medium size (90 m²-144 m²), and large size (>144 m²).

² Class I zones are downtown area, while Class III zones are suburban areas. Class II zones include areas other than those in Class I and Class III.