The relationship between housing prices (HP) and household non-housing consumption (CON) is a significant global economic issue. In China, particularly the underdeveloped central and western regions, there's a phenomenon of low income, low CON rates, and high HPs. The central government has implemented various policies to stimulate consumption, yet CON remains sluggish. This study aims to address the following questions: How do HPs and housing credit constraints (HCC) affect CON in central and western China? What regional differences exist in the response of CON to HPs and HCCs? This study contributes by focusing on underdeveloped regions of China, which has been neglected in previous research. Additionally, it treats HCC as a threshold variable in a varying coefficient panel model, unlike previous studies that considered it as an exogenous variable. This allows for more objective and unbiased identification of thresholds.

Existing literature on the relationship between HPs and CON focuses primarily on developed countries, where rising HPs generally promote consumption based on the Life Cycle Hypothesis and Permanent Income Hypothesis. However, in China, the relationship is less clear, with some studies suggesting a crowding-out effect, while others show a positive impact. The impact depends on various mechanisms: the direct wealth effect (adjustments based on changes in housing wealth) and the collateral constraint effect (increased consumption facilitated by access to mortgage loans). In China, home equity withdrawals are minimal. Furthermore, cultural factors such as the social importance of homeownership and its association with better public education (the "school district effect") significantly influence the wealth effect. The "housing slave effect" describes how the pressure of home purchase leads to reduced consumption in order to save for a home purchase. Studies concerning credit constraints and consumption have generally examined countries with developed credit markets where home equity is easily accessed, and less attention has been paid to China's underdeveloped credit markets where home equity is difficult to obtain. The effect of housing wealth on household consumption through the mortgage transmission channel demonstrates a mutual feedback loop between increasing household debt, housing liquidity premiums, and rising house prices.

This study uses panel data from 18 provinces in central and western China from 2005 to 2020. The data were sourced from China's Statistical Yearbook, China's Real Estate Statistical Yearbook, and local statistical yearbooks. The data were processed using logarithmic transformation and CPI deflation to control for heteroscedasticity and nominal price effects. The core explanatory variable is housing price fluctuation (HPF), calculated as the percentage change in HPs. The explained variable is LnCON (the natural logarithm of per capita non-housing consumption). The threshold variable is HCC, operationalized as the ratio of mortgage loans to GDP. Control variables include per capita urban resident disposable income (URDINC), child dependency ratio (CDR), and elderly dependency ratio (EDR). A panel threshold model was employed, allowing for the identification of threshold values based on data characteristics. The model tests the null hypothesis that the effect of HPF on CON is constant, regardless of HCC. The likelihood ratio (LR) statistic is used to test the significance of the threshold effect. The LLC test and the Fisher-ADF test were used to test for stationarity in the panel data before the regression analysis.

The results indicate a significant single threshold effect of HCC on the relationship between HPF and CON in both central and western China. The threshold values differ between the regions. When HCC is below the threshold, HPF negatively affects CON in both regions, indicating a crowding-out effect. However, when HCC exceeds the threshold, the relationship changes. In the central region, the effect of HPF on CON becomes positive, suggesting a wealth effect. In contrast, in the western region, the effect remains negative, albeit with reduced elasticity, showing that the crowding-out effect persists even with relaxed credit conditions. Most HCC values exceeded the thresholds during the sample period. While URDINC increased during this period in both regions, the rate of increase in CON lagged behind, resulting in low CON rates. This suggests that factors beyond HPF and HCC, such as high savings rates, are also significant factors affecting CON. The high homeownership rate in both regions, despite high housing prices, indicates a significant demand for homeownership, likely due to the linkage between homeownership and access to better public education resources. This "housing slave effect" might have constrained non-housing consumption significantly. The price-to-income ratio (PIR) in the studied regions clearly indicates unaffordable housing prices.

The findings highlight the complex interaction between HPF, HCC, and CON in the context of China's unique housing market and social norms. The threshold effect of HCC shows that relaxing credit constraints can stimulate consumption, particularly in the central region, but the persistence of negative effects in the western region indicates the dominance of the crowding-out effect, and hence that other factors beyond credit conditions are also influencing household consumption. The high homeownership rate and its connection to better public education resources explain the low CON rate, even with relaxed credit conditions. The combination of high housing prices, the school district effect, and high debt levels led households to prioritize housing expenditures over non-housing consumption, despite increased incomes and relaxed credit constraints. This highlights that the traditional economic models which neglect cultural factors or unique institutional settings might not be applicable in studying the dynamic of housing and consumption in China.

This study contributes to the literature by providing empirical evidence of the threshold effects of HCC on the HPF-CON relationship in underdeveloped regions of China. The simultaneous occurrence of low income, low consumption rates, high housing prices, and high homeownership rates suggests that factors beyond credit constraints and HPF are at play, particularly the unique linkage between homeownership and public education access. Future research should explore these intertwined factors to gain a comprehensive understanding of household consumption patterns in China. Policy implications include the need for stable housing prices, improved income levels, and considering the "school district effect" when designing housing and consumption policies.

This study focuses on macro-level data, which might not fully capture the heterogeneity within provinces. The proxy for HCC may not fully capture all dimensions of credit constraints. Future research could benefit from micro-level data and refined measures of HCC. The study primarily focuses on the central and western regions of China, limiting generalizability to other regions with different housing market dynamics.