Asset Allocation and Consumer Behaviors of Chinese Urban Households: Theory and Evidence, a paper co-authored by our school’s Assistant Professor Dong Bingbing, Jiang Tao from the Survey and Research Center for China Household Finance and Economics of Southwestern University of Finance and Economics, and PhD student Zhang Yuan of CUFE’s Tilburg Program, was published in the 2019 11th issue of Journal of Financial Research.
China and many other countries employ fiscal policy (such as tax reductions and subsidies) to stimulate consumption, but the effect of such policy remains controversial. As poor families show a relatively high marginal propensity to consume, while previous studies mostly emphasized the importance of policies tilting towards these families to make them more effective (Galí et al. 2007, Morita 2015). Moreover, Carroll et al. (2014) had further proved that the greater inequality in an economy, the higher consumption growth caused by the positive impact of transitory income shocks. Most of these studies followed the footsteps of Campbell and Mankiw (1989, 1990), defining poor households as “hand-to-mouth” (HtM) which live paycheck to paycheck and have no other wealth to set them apart from those that have wealth, as the former have a higher marginal propensity to consume. But Kaplan and Violante (2014a, 2014b) found that wealthy households may also have a higher marginal propensity to consume, as some of them may own illiquid assets only, which are difficult to liquidate but have very high future returns. As a result, in the case of a transitory income shock, these households will significantly increase consumption, and hence show a quite high marginal propensity to consume. This has been proved across different countries (including the U.S., Canada, Australia and the U.K.) (Kaplan et al. 2014, Hara et al. 2016, Park 2017). Based on the macro model of these micro household consumer behaviors, it is also found that critical changes can take place to the effects of fiscal and monetary policies (Kaplan and Violante 2014, Kaplan et al. 2018). Therefore, it is of vital importance to identifying and accurately quantifying the consumer behaviors and percentage of such households.
The principal contribution of this paper lies in reviewing the screening criteria for wealthy HtM households and determining the percentage thereof. It comes to a more prudent conclusion by fully utilizing the data of the China Household Finance Survey for 2011, 2013, 2015 and 2017, and mainly taking urban households into account. In selecting the screening criteria for the HtM, it establishes the one of liquid assets accounting for one fourth of household annual income by comparing the percentage of households with different liquid assets vs. annual household income ratios with the criterion of whether consumption exceeds income. In choosing the criterion for wealthy households, it adopts the one of whether the net value of a household’s house is positive, other than that of whether its illiquid assets are positive. These considerations have all reflected the characteristics of Chinese households with a relatively high percentage of liquid assets (more prone to saving) and asset allocation dominated by housing properties.
The study finds that the share of the wealthy HtM in China reached up to 36.7%, higher than that (25.55% and 35.89%) calculated by He Yang and Zang Xuheng (2016) using different criteria, and far higher than the 15.3% obtained by Cui and Feng (2017). Such households have the features consistent with relevant theories in terms of year, age of householder, household consumption, assets, income and household type transfer. It is estimated that the elasticity of consumption and income of the poor and wealthy HtM is 4.5% and 5.9% higher than the wealthy non-HtM, respectively. The marginal propensity to consume of the first two groups is 0.069 and 0.09 respectively, both higher than the latter’s 0.045, which conforms with relevant theories and further validates the rationality of the classification method adopted in this paper.
The second contribution of the paper is to identify and quantify the type of households with Chinese characteristics - the poor non-HtM. Since the main illiquid assets of Chinese households are houses which require a relatively high down payment (usu. higher than 30% of the house price), some households save relentlessly before buying houses. The higher the down payment ratio, the higher percentage of such households. In particular, the marginal propensity to consume of these households is very low, or perhaps even negative - as income growth may prompt them to save more so that they can buy houses earlier. We find that such households, with relatively many liquid assets but few illiquid ones, account for 6% of all the samples, and their marginal propensity to consume and elasticity of consumption and income are both lower than the wealthy non-HtM.
The study has very significant policy implications. First, the existence of a large number of wealthy HtM and their high marginal propensity to consume indicate fiscal policy aimed to stimulate consumption shall also take them into consideration. Next, it is essential to pay particular attention to the special type of Chinese households during policy implementation - the poor non-HtM; as measures designed to boost their consumption through income growth may not turn out as originally expected. Last, the existence of heterogeneous households shows that macro fiscal and monetary policies can generate very strong distributional and welfare effects during implementation, so flexibility shall be allowed.
