Income Disparity, Consumption Patterns, and Trends of International Consumption Center City Construction, Based on a Test of China’s Consumer Market

Abstract

In order to promote an upgrade of China’s emerging consumer market and build an international consumer center city, and after analyzing the heterogeneity of the household consumption patterns in 14 cities in the east, middle, and west of China according to the fifth income group, the QUAIDS model was used. This revealed that (1) China’s consumer cities still have a high tendency towards basic consumption, the proportion of development-oriented consumption is obviously low, and the consumption of enjoyment has not yet developed a large-scale market. (2) The characteristics of China’s “dual-consumption” market are obvious: the trend of the structure of household consumption upgrading in eastern cities was obvious, and structural factors have a significant impact on the upgrade of consumption; however, urban consumption in the middle and western regions is still in the stage of basic consumption, and the trend of structural upgrading is weak. (3) Significant differences in consumption were found among urban household groups in different regions: the income gap between the middle and western groups had a significant impact on the heterogeneity and imbalance in the consumption market; as a result, the heterogeneity of the upgraded structure of household consumption in the central and western regions was prominent. However, the income gap among the eastern cities had less influence on the imbalance in the consumption market, and the trend towards and upgrade in the consumption patterns showed good consistency.

1. Introduction

The rapid development of China’s economy has promoted the continuous upgrade of the central cities in the aspects of their influence on international communication, their scientific and technological innovation ability, their guaranteed level of service, and so on. Studies have shown that upgrading the consumption patterns can not only directly promote the development of a high-quality urban economy, but can also indirectly promote the development of a high-quality urban economy by promoting upgrades to the industrial structure [1]. This shows that the cultivation and construction of the international consumption center city has had a far-reaching influence on the expansion of the market and the enhancement of consumption, while playing a greater role via spillover effects and the hub function [2]. Therefore, compared with other cities in China, the enthusiasm of the newly rising central cities to build international consumption centers stems from the internationalization of their own cities and the expansion of the urban consumption scale. However, the increasing income gap among urban residents has resulted in a weak foundation and the prominent contradictions in the upgrades of the structure of consumption, and thus the construction of international consumption center cities is faced with the challenge of finding a long-term balance between increasing the residents’ income and upgrading of consumption. Therefore, this study examined the relationships among income differences, the consumption patterns and the development of international consumption center cities, and probed the influence of income effects and regional effects on the trend over time of the consumption patterns of urban residents. The results are of great significance for realizing the transformation and upgrade of China’s emerging central cities.

Based on data on the consumption patterns of urban residents in 15 central cities in China, this study focused on the following issues: (1) the mechanisms and effects of income heterogeneity and spatial heterogeneity on the upgrading of the consumption patterns of urban residents, and (2) the heterogeneous evolutionary path of the construction of consumption center cities, which we explored in terms of the two dimensions of inter- and intra-central cities. Compared with previous research, the main contributions of this study are that it aimed to integrate macro- and micro-consumption theory, using the quadratic almost ideal demand system (QUAIDS) to analyze the heterogeneity and evolutionary trend of the consumption patterns of urban residents in different income groups, and then find a reasonable explanation for China’s trend of constructing international consumption center cities.

The rest of this article is arranged as follows. The second part contains the literature review, the third part presents the consumption characteristics and consumption trends of urban households in a consumption center; the fourth part describes the model’s design and the empirical analysis, the fifth part presents the predictions of the consumption trends and prospects of urban residents, and the sixth part includes the conclusion and the suggestions.

2. Review

As a method of researching the structure of consumption, Stone (1954) derived the linear expenditure system model (LES) on the basis of Klein’s utility function (1947), and carried out further analyses of the level of consumption and income changes on the impact of decisions regarding consumption [3]. However, because the LES model does not account for the impact of savings, it is far from the reality of consumption. On the basis of the LES model, Liuch (1974) proposed the extended linear expenditure system model (ELES), which is a better model for analyzing the effects of household expenditure on consumption and consumption elasticity, as well as the characteristics of household consumption patterns, and for quantitative analyses of their changing trends [4]. Deaton and Muellbauer (1980) proposed an approximate ideal demand system model (AIDS) based on the linearization of the Stone price index [5]. The AIDS model has been widely used in academic circles since it was put forward. However, Buse (1994) considered that the AIDS model was not consistent and lacked enough approximation, so the Green-Alston method was used to expand and modify the model; to some extent, the problem of variable consistency in AIDS model was solved [6]. Pashardes (1993) also proposed the use of 3SLS to solve the endogenous problems in AIDS models [7]. Although the AIDS model was modified by Muellbauer et al., it still does not adequately address the fact that the Ernst–Engel curve, which is based on the linear hypothesis, faces the general problem of non-linearity in reality [5]. Therefore, Bank et al. (1997) put forward a QUAIDS model including household consumption and thoroughly analyzed the direct consumption effect and the indirect consumption effect, under conditions of market consistency [8]. This explained the impact of changes in household expenditure on the structure of consumption more reasonably. Guanghua Wan (1998), a Chinese scholar, made suggestions how to improve various linearization methods for AIDS estimation by means of the Monte Carto experiment [9].

In recent years, Chinese scholars have carried out in-depth research on the characteristics and changing rules of Chinese residents’ consumption patterns by using models such as ELES and AIDS, and have achieved fruitful results. On the macro level, Huang Jun and Li Jikai (2018) used the ELES model to analyze consumption from 2013 to 2016 and found that the proportion of high-level consumption expenditure increased significantly [10]. Mingyang Zhang and Qi Zhang (2015), and Yong Wang (2018) concluded that the characteristics of rural households, such as the location of the household, the age of the head of the household, and the income bracket, significantly affected their food consumption behavior [11,12]. Huifang Zhang and Yaling Zhu (2017) studied the impact factors of and differences in the consumption patterns of urban and rural residents in China from the perspective of the income structure of urban and rural residents. At the micro level [13], Xindong Zhao and Yong Wang (2013) studied the impact of rising food prices on urban households’ consumption behavior and welfare based on data from the China’s 2007 household income survey [14]. Yingxi Zhang (2014) studied the demand elasticity of urban residents in China from seven aspects of service consumption, such as family service, education, and medical care [15]. Huilian Yuan et al. (2016), using CHIP1995–2002 data, found that the burden of medical care, childcare and housing for China’s urban residents has continued to grow [16]. Tang et al. (2018) used data from 2013 to find that the rising share of housing consumption significantly crowded out other households and reduced aggregate demand [17]. Comparatively, in addition to analyses of differences in the data, macro-level studies are mainly concerned with the evolution of the consumption patterns, while micro-level studies are more concerned about the differences in the consumption patterns.

As a comprehensive method of researching the consumption patterns, from direct effects to indirect effects, and both linear and non-linear, the system model of analyzing the consumption patterns reflects the consistency of the relationships between household expenditure and consumption, and the real consumption market. This empirical study of the consumption patterns went from the macro-level to the micro-level, and carried out quantitative and qualitative research on the consumption patterns and trends across time of households, including urban and rural income groups. The Belt and Road initiative has great importance for cities and regional development, and urban consumption has become an important driver of regional economic growth. On the basis of the existing research results, this study will try to integrate the theories of macro-level and micro-level consumption with the new center cities in China, which have the potential to be international consumption centers. Using the QUAIDS model, this study analyzed the heterogeneity and trend across time of the consumption patterns of urban households in different income groups in order to reflect the heterogeneity of household consumption upgrades in the emerging consumer cities of China. It also provided a reasonable explanation for China’s trend of constructing international consumption center cities, probed into the problems facing the construction of China’s international consumption center cities, and put forward some corresponding countermeasures and suggestions.

3. Characteristics and Trends of Urban Household Consumption in International Consumer Centers

In the process of urban internationalization, the globalization of consumption is becoming more and more obvious, and the consumption of residents has gradually evolved from a closed system to an open system. Webster (2000) first proposed that the goal of urban internationalization is to improve the living standards of urban residents from the perspective of welfare, and proposed that urban capacity is embodied in the ability to produce and sell high-quality products [18]. The improvement in residents’ consumption capacity is considered to be the central core driving force of urban development. On this basis, the concept of the international consumption center city was proposed. Jing Wang (2019) argued that international consumption centers are the product of economic globalization and the deepening of the division of labor among cities, with strong global consumption resource allocation and capacity for leading consumption through innovation [19]. Cities that can effectively satisfy domestic and foreign residents’ consumption of daily life will gather talents through consumption, thus promoting the development of a productive service industry with human capital and knowledge capital as the core input, and stimulate the vitality of innovation. Therefore, an international consumption center city, as an important result of the city’s international development in the field of consumption, will upgrade the space, structure, and quality of consumption, along with continuous improvements in the city’s international level. In September 2018, the Chinese government, in its opinions on improving the mechanism for promoting consumption to further stimulate the consumption potential of the population, proposed the construction of a number of cities as international consumption centers, further indicating that the international consumption center cities have become a new urban development strategy.

Therefore, the capacity and scale of consumption have become important components of a region for building an international consumption center city, and the continuous evolution and upgrading of the consumption patterns will explain the meaning of a consumption city from the perspective of quality. Therefore, consumption is not only the ultimate goal and power of production, but is also a direct reflection of people’s needs for a better life. The existing research results show that the trend of changes in households’ consumption patterns brought about by urban internationalization has great similarity at the global scale. Here, we draw on the Economist Intelligence Unit’s list of the world’s most expensive city cities for living for 2021, compiled from the Worldwide Cost of Living 2021 analysis of household consumption in 133 major cities carried out by the world’s largest price aggregator, Numbeo. As shown in

Table 1. The composition of consumption of urban residents in the world’s major consumption centers in 2021.

Rank	City	Food	Clothing	Living	Equipment Housing	Transportation	Education and Entertainment	Medical Services	Other
1	Singapore	8.5	10.93	15.3	6.22	17.35	10.81	10.28	20.64
2	Paris	14.92	5.5	24.76	5.7	15.09	12.66	11.6	10.45
3	Zurich	12.34	5.7	20.64	5.45	15.73	10.87	11.78	17.5
4	Hong Kong	15.61	14.37	14.98	6.02	7.37	8.77	8.80	24.13
5	Oslo	15.71	5.90	22	5.35	16.09	11.99	9.36	14.4
6	Geneva	14.20	6.0	20.5	5.02	15.21	10.35	11.5	16.5
7	Seoul	13.4	6.17	25.55	5.65	16.37	10.46	8.90	14.5
8	Copenhagen	14.51	5.13	21.36	5.20	16.70	12.27	11.1	17.27
9	TelAviv	12.12	5.27	22.83	5.11	15.73	11.08	10.87	14
10	Sydney	12.70	5.32	23.48	5.63	14.57	14.05	9.25	14.06
	Chinese city	27.72	6.92	23.95	6.24	13.3	11.39	7.84	2.64

.

The composition of household consumption in the world’s top 10 cities in 2021 shows that the major areas of household consumption in the world’s consumer centers are housing, transport, communications, education, entertainment and other goods. These four items of expenditure account for 70% of the total household expenditure, while the consumption of services also accounts for more than 50% of total household consumption expenditure, indicating that urban households in international consumption centers are dominated by developmental and enjoyment-based consumption, in addition to housing, food, clothing, and household equipment. The share of basic consumer goods in household consumption is only 30%, and this trend continues to decrease significantly. By contrast, China’s urban household consumption is still dominated by food and housing, which account for 50% of total consumption, and is more than twice the average for urban food consumption in developed countries. Moreover, there are significant differences in the share of consumption of health services, education, recreation, and other commodities compared with cities in developed countries, reflecting the fact that urban household consumption in China is still at a stage of development dominated by the consumption of basic commodities. Development-oriented consumption and enjoyment-oriented consumption, which are mainly service goods, make up a low proportion of household consumption, which shows the huge gap between the urban development of China and the international consumption center cities.

4. Model and Results

4.1. Model Design

As a method for researching consumption demand, Stone (1954) derived the LES model on the basis of Klein’s utility function [3]. Since then, both the super-log model and the Rotterdam model have been used to test the homogeneity and symmetry limits of demand theory, but because of the strict application of the linear Ernst–Engel curve, it is difficult to solve the problems of non-linear estimation and non-parametric linear constraints in the estimation of the model. In response to this, Deaton and Muellbauer (1980) proposed the AIDS model, which could manage the problem of households’ consumption patterns without invoking a strictly linear Ernst–Engel curve [6]. The linear restriction of the fixed parameters simplifies the test of the restriction of homogeneity and symmetry, and thus the AIDS model has been widely used. The model assumes that market consumers are rational and that market demand is the sum of all consumer demand. Under the condition of a constant price, the expenditure function of consumption is expressed as the minimum necessary expenditure under a certain utility level. According to the Hikes demand function, let u be the utility, C be the consumption function, a(p) be the commodity price index and b(p) be the sum of C-D function prices. We thus have

lnC(p,u) = lna(p) + ub(p)

$$\ln a(p_t)=a_0+{\displaystyle \sum _{j=1}^n{\alpha }_j\ln p_{jt}+\frac{1}{2}}{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{r}_{ij}\ln (p_{it})\ln (p_{jt})}}$$

(1)

$$b(p_t)={\displaystyle \prod _{i=1}^n{p}_{it}{}^{{\beta }_i}}$$

We can obtain the consumers’ spending function by

$$w_i={\alpha }_i+{\displaystyle \sum _{j=1}^n}{r}_{ij}\ln p_j+{\beta }_i\ln \frac{m}{a(p)}$$

(2)

where $w_i$ indicates the share of expenditure on class I goods, $p_j$ indicates the price of class j goods, m indicates total consumer spending, $r_{ij}$ indicates the impact of the change in the commodity price of category j on the share of expenditure on commodities of category i when the actual expenditure is constant, and ${\beta }_i$ indicates a change in the share of expenditure on category i commodities resulting from a change in actual expenditure when the prices are constant. When ${\beta }_i$ > 0, goods are considered luxuries, and when ${\beta }_i$ < 0, goods are considered necessities.

In reality, however, the Ernst–Engel curve is not always linear. Banks et al. (1997) proposed an approximate ideal consumption model (QUAIDS) including the quadratic terms of household consumption that was based on the AIDS model [10]:

$$w_{it}=a_i+{\displaystyle \sum _{i=1}^n{r}_{ij}}ln{p}_{jt}+b_i\left[\frac{m}{a_i(p_t)}\right]+\frac{l_i}{b_i(p_t)}{\left\{ln\left[\frac{m}{a_i(p_t)}\right]\right\}}^2+u_{it}$$

(3)

where $\lambda (p_t)={\displaystyle \sum _{\mathrm{i}=1}^{\mathrm{n}}{\lambda }_{i}ln{p}_{it}}$, $\lambda \left(p\right)$ is the homogeneous function of price p. The model’s additive condition is ${\displaystyle \sum _{i=1}^n{\alpha }_i=1;}{\displaystyle \sum _{i=1}^n{\beta }_i=0;{\displaystyle \sum _{i=1}^n{r}_i=0;{\displaystyle \sum _{i=1}^n{\lambda }_i=0}}}$; the homogeneous condition is ${\displaystyle \sum {\lambda }_{ij}=0}$; the symmetry condition is $r_{ij}=r_{ji}$ (i ≠ j).

The QUAIDS model introduced the quadratic term $\lambda$ of household consumption. By explaining the influence of the change in real expenditure on the share of different kinds of commodity expenditure, the model not only analyzes ${\beta }_i$, but also analyzes ${\lambda }_i$. Because it can reflect the attributes of the non-linear change in the marginal expenditure share of goods with an increase in income level, and the scenario of expenditure on consumer goods accords with the real market situation, the model can explain the structural change in household consumption more reasonably.

In order to calculate the elasticity of the quantized model, we use the differential equations for lnm, lnp_j, respectively:

$$u_i=\frac{\partial w_i}{\partial lnm}=b_i+\frac{2l_i}{b(p)}\left\{ln\left[\frac{m}{a_i(p_t)}\right]\right\}$$

(4)

$$u_{ij}=\frac{\partial w_i}{\partial ln{p}_j}=r_{ij}-u_i({\alpha }_j+{\displaystyle \sum _k{r}_{ij}ln{p}_k)-}\frac{{\lambda }_i{\beta }_j}{b(p)}{\left\{ln\left[\frac{m}{a_i(p_t)}\right]\right\}}^2$$

(5)

Moreover, in Banks et al. (1997) [10], the elasticity of consumer spending and the non-compensated cross-price elasticity were obtained as follows:

$$u_i{=1+\mathrm{w}}^{-1}\left[{\beta }_i+\frac{2{\lambda }_i}{b(p)}ln\left(\frac{x}{a(p)}\right)\right]$$

(6)

$$u_{ij}=w_i{}^{-1}\left\{{\gamma }_{ij}-({\beta }_i+\frac{2{\lambda }_i}{b(p)})\left[ln\frac{x}{a(p)}\right]({\alpha }_j+{\displaystyle \sum _k^n{\gamma }_{jk}ln(p_k)})-\frac{{\lambda }_i{\beta }_j}{b(p)}{\left[ln\frac{x}{a(p)}\right]}^2\right\}-{\delta }_{ij}$$

(7)

When i = j, ${\delta }_{ij}$ = 1; when i $\ne$ j; ${\delta }_{ij}$ = 0.

4.2. Variables and Measures

To select the variables for our study, we relied on the availability of data and the credibility of indicators, and adopted the variables used by domestic and foreign scholars as the basis of our selection, and made adaptive adjustments according to the actual situation in China.

(1)

Consumption patterns of urban households

China’s new consumption center cities have a high level of internationalization. However, China’s main goal is the cultivation and construction of international consumption center cities. In particular, municipalities and provincial capitals can have more urban development factors, so these are considered in the construction of an international consumption center city. In this regard, the study used the composite panel data of China’s new consumption center cities for 2002–2021. In terms of the consumption indicators, according to the structure of household expenditure most commonly used in Chinese academic classification, the consumption of household expenditure by urban residents was divided into eight categories: food, clothing, housing, transportation and communication, household equipment, medical care, culture, and education and entertainment. The consumers’ spending data and the price index data were sourced from the Chinese Statistical Annals for 2002–2021.

(2)

Statistical description

The consumption data of urban households in 14 of China’s new consumption center cities are collated in

Table 2. Results on the composition of household consumption expenditure.

Variables	Sample Size	Mean	S.D.	max	min
Food	1904	8.68	7.82	9.51	7.60
Clothing	1904	7.36	6.53	8.02	6.03
Living	1904	7.92	7.94	9.61	6.17
Equipment housing	1904	7.07	6.42	7.95	5.55
Transportation	1904	7.76	7.27	8.82	6.3
Education and Entertainment	1904	7.79	7.20	8.84	6.46
Medical services,	1904	7.23	6.59	8.50	6.03
Other	1904	6.46	5.91	7.46	4.86

.

The results in Table 2 show that the household consumption expenditure of these 14 cities is mostly on food, housing, transportation and communication, and education and entertainment. The differences in the structure of consumption are significant, especially for housing, household equipment, transportation and communications, education, healthcare, and other areas, showing the widening gaps among regions.

4.3. Results

(1)

Comparison of the consumption patterns

We used the QUDIAS model to evaluate the household consumption of 14 cities, and estimated the values of α, β, γ, and λ in Stata19.0. The results obtained by Equations (6) and (7) are shown in

Table 3. Non-compensation elasticity of various consumption expenditure items of urban residents.

	Food	Clothing	Living	Equipment Housing	Transportation	Education and Entertainment	Medical Services	Other
Food	−0.0982
	(1.19)
Clothing	0.6667 ***	1.8099 ***
	(2.7)	(6.49)
Living	−0.2991 **	−1.6654 ***	0.7155 ***
	(2.15)	(4.8)	(2.6)
Equipment housing	−0.0509	−0.4450 **	0.1907 *	0.0557 **
	(0.97)	(2.19)	(1.92)	(2.18)
Transportation	0.0575 *	−0.2753	0.0412	0.0376	0.0810 **
	(1.83)	(0.97)	(0.35)	(1.23)	(1.69)
Education and Entertainment	−0.0709 ***	−0.0434	0.0531	0.0120	−0.0454 ***	0.0371 ***
	(2.57)	(0.31)	(0.85)	(0.64)	(3.0)	(3.09)
Medical services	−0.473 **	−2.6018 ***	1.1763 ***	0.2656 *	0.1911	0.0612	1.751 ***
	(2.43)	(6.08)	(5.17)	(1.9)	(1.0)	(0.62)	(3.43)
Other	0.0715 *	0.5542 ***	−0.2123 ***	−0.066 **	−0.088*	−0.0037	−0.3702 ***	0.1142 ***
	(1.89)	(4.07)	(2.73)	(2.2)	(1.72)	(0.18)	(3.58)	(2.68)

Note: The z-value of the coefficient’s test statistic are shown in parentheses, and ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

and

Table 4. Comparison of income elasticity and the share of marginal expenditure between Chinese and foreign urban residents.

	China Residents			International Residents
	Income Elasticity	Expenditure (%)	Marginal Expenditure (%)	Income Elasticity	Expenditure (%)	Marginal Expenditure (%)
Food	1.1314 **	32.47	36.74	0.1109 **	9.22	0.55
Clothing	0.8882 ***	8.69	7.72	0.34	3.78	(0.17)
Living	1.3730 ***	15.07	20.55	0.51	25.23	13.27
Equipment housing	0.1109 **	6.51	7.75	0.25	5.75	7.30
Transportation	0.733	12.87	9.43	1.2051	16.90	19.06
Education and Entertainment	0.55	13.27	7.3	1.0370	8.45	0.50
Medical services	0.6013	7.65	4.6	0.2583	4.17	6.13
Other	0.1447	3.47	2.11	0.7793	26.5	30.68

Note: The z-value of the coefficient’s test statistic are shown in parentheses, and ***, **, indicate significance at the 1%, 5% levels, respectively.

:

From the results of self-elasticity of consumption items, we can see that clothing (1.81), housing (0.72), and medical services (1.75) have high self-elasticity, which indicates that residents have stronger sensitivity regarding these categories of consumption, that is, price fluctuations can trigger significant changes in the sale of goods. The elasticity of food (0.098), household equipment (0.06), transportation and communication (0.08), education and entertainment (0.04), and other (0.11) categories was lower, indicating that the consumption sensitivity of these five categories of commodities was lower. In comparison, clothing, housing, and medical services are luxury goods within household consumption in new consumption center cities, whereas food, education and entertainment, transportation and communications, household equipment, and other commodities have the nature of necessities. The income elasticity data reflect the degree of the impact on household consumption from the perspective of income, with the income elasticity values of food, clothing, housing, transportation and communication, education, and medical services ranging from 0.5 to 1, showing strong sensitivity, whereas expenditure on household equipment and other categories had low sensitivity. The share of marginal spending is used to indicate the degree to which growth in consumption affects the consumption of different goods under constant prices. Growth in each unit of spending tended to increase the proportion of a particular consumer category. The results showed that the marginal expenditure share of food and housing was large, which indicates that with the continuous growth in urban household incomes, the consumption of food and housing will increase rapidly. Growth in the consumption of transportation, education and entertainment, household equipment, clothing, healthcare, and other goods will be relatively slow.

The abovementioned analysis of the overall consumption of China’s emerging urban households ignores the heterogeneity of consumption of households with different levels of income, which may lead to biased analysis results and not reflect the actual situation. Therefore, we need to study the characteristics of the consumption patterns of different income groups further. To this end, using the income classification method commonly used in Chinese statistics, the total annual expenditure of urban households was divided into the low-income group (Y_l), the low–middle-income group (Y_lm), the middle-income group (Y_m), the middle–high-income group (Y_mh), and the high-income group (Y_h). The QUAIDS estimates are given in

Table 5. The non-compensation price elasticity of urban residents in each income group.

	Non-Compensating Price Elasticity					Income Elasticity
	Yl	Ylm	Ym	Ymh	Yh	Yl	Ylm	Ym	Ymh	Yh
Food	−0.2548 *** (4.96)	−0.2853 *** (5.24)	−0.1927 *** (3.73)	−0.2406 *** (4.48)	−0.2804 *** (4.8)	1.3664 *** (7.85)	1.3642 *** (8.25)	1.1944 *** (9.05)	1.0287 *** (6.51)	1.3861 *** (8.56)
Clothing	0.1192 (1.69)	0.3567 *** (6.02)	−0.1030 (1.6)	−0.396 *** (7.31)	0.3128 (4.80)	0.1013 (1.41)	0.5031 *** (3.04)	0.701 (1.42)	0.7642 *** (3.94)	1.0678 *** (2.81)
Living	0.6708 *** (5.99)	0.557 ** (2.42)	−0.4175 *** (3.39)	0.3377 *** (2.91)	0.1670 (1.25)	0.2150 *** (3.28)	0.660 *** (2.67)	0.7485 * (1.86)	0.6932 * (1.83)	0.2273 (1.09)
Equipment housing	0.1985 (0.29)	−0.1064 (1.83)	−0.3455 *** (8.14)	0.1961 ** (2.48)	−0.2240 ** (2.28)	0.0615 ** (2.12)	0.0772 (1.51)	0.5486 *** (5.13)	0.2480 (1.70)	0.4852 *** (9.4)
Transportation	−0.1508 * (1.52)	−0.0611 * (1.7)	0.3470 *** (3.38)	−0.323 *** (2.85)	−0.1908 (1.47)	0.1204 (0.93)	0.0309 (0.63)	0.6687 ** (2.62)	0.6260 * (1.85)	0.2916 (1.32)
Education and Entertainment	−0.1527 ** (2.51)	−0.2402 ** (2.37)	−0.1738 ** (−2.31)	0.0708 (1.57)	0.070 * (1.85)	0.1494 ** (1.98)	0.496 *** (3.07)	1.368 * (1.94)	1.398 (1.31)	1.843 (1.61)
Medical services	−0.1256 (1.09)	0.7884 *** (2.68)	0.8975 *** (6.02)	0.1691 (0.92)	0.1350 (1.41)	0.2765 (1.74)	0.3202 *** (3.22)	0.3878 *** (3.57)	0.5128 ** (2.15)	1.1001 (0.60)
Other	0.1259 *** (−3.8)	0.0142 (−1.41)	0.1685 *** (−4.29)	0.2120 *** (3.47)	0.234 (0.39)	0.0842 ** (2.34)	0.0230 *** (3.36)	0.1838 *** (3.98)	0.2120 *** (3.47)	0.2938 * (1.90)

Note: The z-value of the coefficient’s test statistic are shown in parentheses, and ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

.

For the low-income group, food and living have the greatest elasticity. Changes in their prices will make the corresponding demand for commodities increase significantly. In terms of consumer spending, the impact of consumer spending on food, housing and health services on low-income groups is significant, meaning that the increased income of low-income groups will mainly be spent on food, housing, and health services. This reflects the current underconsumption of housing and health services by low-income households, as well as the resulting low standard of living and the consumption patterns.

For the low–middle-income group, clothing, housing, and medical services are more resilient, and the corresponding decline in commodity prices can increase the consumption of low–middle-income households, leasing to their relatively rapid growth. In contrast, spending on food, housing, clothing, education and entertainment, and health services was more flexible than the spending by low-income groups, whereas the demand for education and entertainment by low–middle-income households shows a growing trend.

For the middle-income group, the self-elasticity of housing, household equipment, transportation, communication, and medical services was greater, and the corresponding commodity price had a more obvious impact on middle-income households. In terms of consumption expenditure, the elasticity of expenditure on food, clothing, housing, household equipment, transportation, communications, and education and entertainment was significant, reflecting the strong willingness of urban middle-income households to consume various commodities. In particular, these households have begun to focus on spending on growth-oriented consumption.

For the middle–high-income group, the elasticity of clothing, housing and transportation is greater than that of food, clothing, transportation, and education and entertainment. Compared with the middle-income group, the middle–high-income group has lower elasticity in basic consumption and development consumption, but higher elasticity in the consumption of other commodities. This shows that the consumption of middle–high-income households is gradually transitioning from development-oriented consumption to enjoyment-oriented consumption.

For the high income group, the low self-elasticity of the eight commodities indicates that the relationship between price fluctuations and consumption is weak for the high-income households. It is worth noting that high-income households have the highest propensity to spend on food, clothing, household equipment, medical services, and other goods out of all the income groups. This indicates that the high-income households have strong potential in the development-type consumption and enjoyment-type consumption markets.

A comprehensive comparison of the consumption patterns of households in different income groups clearly showed the “smile curve” of China’s urban household consumption (Figure 1). The when the income of households is lower, the flexibility of spending on basic goods is greater. With the growth of household income, the elasticity of expenditure on basic commodities decreases, and the elasticity of expenditure on developmental services increases. After entering the stage of high income, the elasticity of expenditure on basic commodities increases again; at this point, the elasticity of spending on developmental services goods will continue to grow. Therefore, it also indicates the gradual decline in the consumption of basic commodities and the gradual increase in consumption of developmental commodities in the process of the income growth of urban households in China.

(2)

Comparison of the consumption patterns among different regions

The particular structure of China’s “dual economy”, the distinct regional development gap and the resulting heterogeneity in the evolution of urban households’ consumption patterns are other major issues. Here, we divide the 14 cities into the eastern, central and western regions. A total of 7 of the 14 cities were located in the eastern region, 3 were in the central region and 4 were in the western region. Given the QUAIDS model’s requirements in terms of sample size, the eastern region showed a significant divergence from the central and western regions, while the central and western regions had a smaller gap between them (Lu and Sun, 2018) [20]. Therefore, the study combined the central and western cities into a single study group. The self-elasticity of consumption (

Table 6. Self-elasticity of household consumption in different regions.

Eastern China						Mid-Western China
	Yl	Ylm	Ym	Ymh	Yh	Yl	Ylm	Ym	Ymh	Yh
Food	0.1290	−0.3664 ***	−0.1175	0.6328 ***	−0.5304 ***	−0.1931 ***	−0.1257 **	−0.2723 ***	−0.3733 *	−0.2052 ***
	(0.15)	(−5.13)	(−1.37)	(5.57)	(−5.55)	(−3.12)	(−2.17)	(−7.97)	(−1.58)	(−3.39)
Clothing	−0.2192 ***	0.0656	0.173 **	0.1344	−0.1172 *	0.4330 ***	−0.5721 ***	−0.4631 ***	−0.51 ***	−0.5381 ***
	(−3.06)	(0.74)	(2.4)	(1.32)	(−1.6)	(4.39)	(−7.53)	(−7.51)	(−6.81)	(−0.6)
Living	0.059	0.3739 **	0.61 ***	−0.3694 *	1.0167 ***	0.8046 ***	0.4467 **	0.2693 ***	0.4486 **	1.1898 ***
	(0.45)	(2.08)	(4.21)	(−1.99)	(7.56)	(6.02)	(2.15)	(3.08)	(2.33)	(9.65)
Equipment housing	−0.46 ***	−0.2801 ***	0.6984	−0.3375 **	−0.2126 ***	0.0917	0.1218 *	−0.1702 *	0.2646 ***	0.2500 ***
	(−6.59)	(−4.72)	(0.595)	(−2.05)	(6.11)	(0.81)	(1.5)	(−1.82)	(2.79)	(3.05)
Transportation	0.563 ***	−0.2573 **	0.138	1.16 ***	−0.3816 **	−0.0932	0.2847 ***	−0.1436 ***	0.0088	−0.1444
	(4.37)	(−2.05)	(0.7)	(6.75)	(−2.45)	(−0.7)	(3.12)	(−2.69)	(0.07)	(−1.13)
Education and Entertainment	−0.2 *	−0.0275	0.43 ***	−0.56 ***	0.068	−0.0219	0.1022 *	0.7753 ***	0.1598 ***	0.1764
	(−1.6)	(−0.25)	(4.36)	(−5.35)	(0.05)	(−0.28)	(1.77)	(15.1)	(2.67)	(2.97)
Medical services	0.207 *	0.5489 ***	−1.2 ***	−0.2	−0.7896 ***	−0.8031***	−0.2099	0.0298	−0.2418 *	−1.060
	(1.95)	(3.87)	(−5.5)	(−1.2)	(−6.8)	(−3.06)	(−1.2)	(0.4)	(−1.32)	(−4.92)
Other	−0.015	−0.0571 *	−0.237	−0.5778 ***	0.9673	−0.0218***	−0.0476	−0.0252	−0.0327	−0.1529
	(−0.4)	(−1.32)	(0.8)	(9.19)	(1.26)	(−3.71)	(−1.14)	(−0.19)	(−0.62)	(−3.1)

Note: The z-value of the coefficient’s test statistic is shown in parentheses, and ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

) and the corresponding elasticity of household income (

Table 7. Households’ income elasticity across the regions.

Eastern China						Mid-Western China
	Yl	Ylm	Ym	Ymh	Yh	Yl	Ylm	Ym	Ymh	Yh
Food	0.9386 ***	1.237 ***	1.2178 ***	1.162 ***	1.4112 ***	1.2396 *	1.5738 ***	1.1083 ***	1.2862 ***	1.1861 ***
	(4.4)	(4.41)	(5.23)	(2.7)	(3.33)	(2.30)	(4.53)	(4.0)	(4.52)	(8.56)
Clothing	0.5816 *	0.661	0.7124	0.69	0.687	0.3623 ***	0.8600 ***	0.4577 ***	0.8887 ***	0.3313 **
	(1.47)	(0.28)	(1.02)	(0.41)	(0.86)	(3.46)	(3.84)	(3.63)	(3.52)	(2.21)
Living	0.2985 **	0.5077	1.7614 **	0.84 *	0.2506 ***	1.858 ***	0.8659	0.6711 *	0.8261	2.1003 ***
	(2.427)	(0.95)	(2.27)	(1.38)	(4.16)	(3.28)	(1.01)	(1.91)	(1.01)	(4.95)
Equipment housing	0.8407 ***	0.326 **	0.406	0.428 ***	0.4384 ***	0.2045*	0.3314 **	0.3641	0.4554 **	0.4023 *
	(3.05)	(2.44)	(0.84)	(2.67)	(4.78)	(1.79)	(2.12)	(1.07)	(1.90)	(1.30)
Transportation	1.1941 **	1.2457	1.21 ***	1.332 ***	1.4895	0.3961**	0.5283 **	0.5336.	0.6700	0.8942 ***
	(2.04)	(0.91)	(2.65)	(3.33)	(0.87)	(0.61)	(2.01)	(1.7)	(0.69)	(2.70)
Education and Entertainment	0.1715	0.4127	0.7006 *	1.312 **	1.1561 ***	0.196 ***	0.2281 **	0.8765 ***	0.5963 **	0.3142 ***
	(1.31)	(0.71)	(1.92)	(2.51)	(3.45)	(3.31)	(2.37)	(7.12)	(2.26)	(2.65)
Medical services	0.6182	1.2391	1.3874 ***	2.04	2.2057 ***	0.1382 *	0.2196	0.3056 *	1.4183	2.3944 **
	(0.59)	(1.88)	(3.48)	(0.1)	(3.22)	(1.54)	(0.17)	(1.49)	(0.32)	(2.41)
Other	0.285 ***	0.357 ***	1.106 ***	1.548 ***	1.1018 *	0.1298 *	0.1480 *	0.4215	0.507 *	0.6733 **
	(5.06)	(2.73)	(2.44)	(4.86)	(1.31)	(1.90)	(1.86)	(0.95)	(1.82)	(2.04)

Note: The z-value of the coefficient’s test statistic is shown in parentheses, and ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

) were estimated for eastern and central–western households.

The results of the analysis shown in Table 4 and Table 5 indicate that in terms of inter-regional heterogeneity, the food price elasticity of high-income households in the eastern region was relatively large, while that of the central–western households is relatively small. The income elasticity of the eastern middle-income group was significantly lower than that of the corresponding central–western group, but the elasticity of food consumption in the eastern region was higher than that of the central–western group after households entered the middle–high-income group. This reflects the gradual evolution of food consumption to a higher level in the east, whereas in the central–western regions, it remained at the level of basic consumer goods. In terms of clothing consumption, households in the central–western regions were generally more resilient, indicating that clothing consumption in the central–western regions is a luxury. In terms of housing, the elasticity of prices and income first increased and then decreased in the eastern region, while the elasticity of the central–western regions first decreased and then increased, reflecting the significant regional divergence in China’s real estate consumption market. High housing prices in the eastern cities made housing consumption a luxury for the low- to middle-income groups, while housing prices in the central–western regions’ cities were relatively low. When income increased to the middle level, it could meet the household demand for housing consumption. As for the higher price elasticity of the high-income group, this was the result of household asset allocation under the influence of expectation. In terms of developmental consumption, the general consumption elasticity of households at all income levels in the eastern region was lower than that of the same group of households in the central–western regions, especially regarding the consumption of transportation, communication, medical services, and education and entertainment. This reflects that urban household consumption in eastern China has a relatively high structure. In terms of income elasticity, in addition to food, urban households in the east pay more attention to transportation, education, entertainment, and other types of consumption, while households in the central–western regions pay more attention to the consumption of clothing and medical services.

In terms of the heterogeneity among different income groups in different regions, the differences in household consumption among the income groups in the eastern regions was small (0.51): low-income households paid attention to the consumption of household equipment and transportation; low–middle-income households paid attention to transportation and medical services; middle-income households were concerned with housing, transportation, communication, and medical services; and middle–high-income households were concerned with education, entertainment, transportation, medical services, and other commodities. High-income households increased their consumption of household equipment compared with the middle–high-income group, indicating that household equipment has been upgraded from basic commodities to luxury goods. The difference was significant (0.75). All income groups were concerned about housing consumption, except the middle–high-income group and the high-income group, which were concerned about medical services. The heterogeneity of self-elasticity and income elasticity in the consumption of other commodities was greater. This showed that the gaps in household income in the central–western regions had a significant impact on the heterogeneity and imbalance of the consumption market, indicating that the diversification of the regional consumption market is relatively low. The income gap in eastern cities had less influence on the imbalance of the consumer market, which reflects the higher diversification of the consumer market’s structure in eastern China.

5. The future Trends of the Consumption Patterns of China’s Urban Households

5.1. The Trends of the Consumption Patterns in Different Income Groups

Table 8. The share of marginal expenditure by urban residents in each income group.

Variables	Yl (%)	Ylm (%)	Ym (%)	Ymh (%)	Yh (%)
Food	51.81	48.46	36.44	28.32	30.63
Clothing	4.27	4.45	6.32	6.55	6.98
Living	14.42	21.76	23.85	17.62	10.59
Equipment housing	5.68	4.94	6.47	5.76	6.57
Transportation	5.78	6.52	8.16	10.03	12.34
Education and Entertainment	4.94	7.17	6.15	9.15	14.29
Medical services	4.46	4.69	6.22	7.82	10.78
Other	2.87	4.75	5.48	6.82	8.06

shows the future trend of households’ consumption patterns in 14 cities of China. In general, 60–70% of the expenditure of the low–middle-groups and the low-income groups is spent on food and household goods, although the propensity to consume basic commodities will decline to about 40% and the propensity to consume developmentally will increase from about 20 to 40% as incomes move towards higher groups; this is still far from the average 50% share of developmental consumption by households in the world’s major international consumption centers.

In terms of the specific composition, the consumption of basic commodities by urban households in the different income groups indicates that the income elasticity of food consumption has begun to decline, up to the high-income level, where it has gradually increased. This may be due to the higher frequency of eating out by high-income households (Zheng et al., 2015) [21], which, in turn, leads to an increase in the elasticity of spending on food consumption. Compared with the continuous decline in food consumption in developed countries, the proportion of food consumption by urban households in China is much higher than the average of 13% in the international consumption center cities. In contrast to the declining share of food consumption by urban residents in developed countries, this trend also reflects the change in the eating habits of urban residents in China and the resulting upgrade of the consumption patterns of the catering industry. In addition, after the income level of China’s urban households reaches the middle-income level, the elasticity of housing expenditure begins to decline, and the higher the household income, the lower the propensity for housing consumption. This reflects the living conditions of households in China’s emerging consumer cities in the process of continuous improvement. The consumption of real estate and related commodities has shown a declining trend. The general theory is that urban internationalization will increase resident’’ income and promote urban housing prices (Shen et al., 2012) [22], and the continuous decline in housing consumption with growth in income is contrary to that of international consumption center cities. Favara and Song (2014) put forward a plausible explanation for this problem, namely the persistence of expectations of future house prices, where income affects house prices and rents, with the expected differences for different household income levels [23]. In the context of the Chinese government’s governance of the real estate market for many years, the continuous decline in housing consumption of the middle–high-income group and the high-income group testifies to the objective reality of the negative influence of market expectations and differences in income on housing prices.

In terms of service consumption, urban households in China tend to spend more on transportation, communication, education, entertainment, and medical services, while urban households in developed countries tend to spend more on transportation, communication and other goods. On the one hand, China’s level of economic development is on the low side. Although China’s per capita GDP exceeded USD 10,000 in 2019, reaching the level of medium-developed countries, the gap between China and developed countries is still significant. Development-oriented consumption, such as transportation, communication, medical treatment, and education, is still the main form of service consumption. On the other hand, the high welfare subsidy policies of the governments of developed countries have made up for private expenditure on education, healthcare, etc. to a considerable extent; as a result, households in developed countries tend to spend less on these developmental services as a whole, but they tend to spend more on the services they enjoy.

From the perspective of differences in income, there is a significant difference in the future trends of the consumption patterns among households with different income levels in China. The low-income households mainly consume food, and the middle–low-income households’ consumption demand for food and housing will increase rapidly. The propensity to consume food in the middle-income group declined rapidly, but their expenditure on housing consumption was the highest among all the groups, indicating that the consumption patterns of urban households enters a stage of rapid adjustment after they reach the middle-income level and become the main drivers of urban housing demand. The middle–high-income group and the high-income group showed a continuous decline in consumer spending on basic commodities, and their demand for developmental service commodities is constantly increasing. Among them, the middle–high-income group showed a clear increase in spending on transportation and communications, whereas the high-income group was willing to invest more in education and entertainment.

5.2. The Trends of the Consumption Patterns of Different Income Groups in Different Regions

Table 9. China’s share of marginal expenditure by urban residents in various regions and income groups.

	Eastern China					Mid-Western China
	Yl	Ylm	Ym	Ymh	Yh	Yl	Ylm	Ym	Ymh	Yh
Food	47.84	44.0	30.18	32.70	32.86	48.20	57.8	38.33	38.69	32.42
Clothing	3.96	4.62	4.64	4.31	5.39	3.21	8.52	6.76	9.73	3.43
Living	3.57	5.87	17.61	9.54	2.54	14.16	12.01	8.84	10.88	15.46
Equipment housing	4.45	2.1	3.18	3.52	3.25	3.20	2.18	2.53	3.33	3.13
Transportation	10.55	11.67	13.55	15.56	18.09	4.12	5.95	6.51	8.30	12.26
Education and Entertainment	2.29	5.57	9.51	10.70	15.76	2.27	2.75	11.04	7.31	10.16
Medical services	5.06	9.53	11.18	12.65	16.73	1.10	1.72	2.28	10.96	20.09
Other	1.06	1.90	3.3	6.23	5.01	0.25	0.34	1.15	1.52	2.50

shows the share of marginal expenditure by the residents in the cities of the eastern and central–western regions in China. In terms of the regional differences, the decreasing trend of the residents’ expenditure on food consumption in eastern cities is more obvious. That of the central–western regions first increased, then decreased, and then increased again. The strong contrast between regions indicates that the structural imbalance in China’s housing consumption market is prominent between regions. The eastern urban households have tended to increase their share of consumption rapidly with an increase in income, especially on transportation, education, entertainment, and medical services. The average consumption tendency of eastern urban households (11.23%) was significantly higher than that of the central and western urban households (7.12%), households in the central–western regions were willing to spend more on housing (12.27%). The advanced trend of the eastern cities’ consumption patterns is obvious, whereas the central and western cities’ consumption is still in the stage of basic consumption.

In terms of the regional differences, the heterogeneity of the consumption tendency of the eastern cities was small (11.37), and the tendency of food and transportation consumption was high in the low-income group The middle-income group had a higher propensity to consume housing, transportation, and medical services, while the middle–high-income group and high-income group had higher consistency in their consumption patterns and trends. It is worth noting that the housing consumption tendency of middle-income households was much higher than that of other income groups, which reflects that housing consumption has become a watershed in the evolution and upgrading of eastern households from basic consumption to developmental consumption. In other words, through resolving the housing problem, households will show rapid growth in developmental consumption and enjoyment consumption. The difference in the consumption trends of each income group in the central and western cities (13.54) was larger than that in the eastern cities, indicating that the income gap is more important to the structural problems consumption by central and western households. The low-income group and the low–middle-income group had the highest propensity to consume food and housing, the middle-income group had the highest propensity to invest in education and entertainment, and the middle–high-income group had the highest propensity to consume housing and medical services. The high-income group showed high consistency with the eastern cities in terms of developmental consumption except for housing, but the difference was still large for other goods.

6. Conclusions and Discussion

6.1. Conclusions

Based on our analysis, the “smile curve” of household consumption patterns in emerging consumption cities indicates that China’s urban household consumption patterns are in the process of a rapid upgrade: on the one hand, the demand for basic commodities, represented by food consumption, continues to decline; on the other hand, developmental consumption, represented by education, medical care, and transportation, has increased significantly. However, compared with the world’s major consumption centers, China’s consumption cities still have a high tendency towards basic consumption, the proportion of development-oriented consumption is obviously low, and enjoyment-oriented consumption has not yet formed a market.

Regarding the heterogeneity in regional consumption, the trend of upgraded household consumption patterns in eastern cities is obvious, while consumption in the central and western cities is still at the stage of basic consumption, and the trend of upgraded consumption patterns is weak. Among them, in the eastern region, the escalating trend of food consumption is obvious, and the habit of eating out is developing. The central–western regions’ food consumption is in the declining stage, and the households’ traditional food consumption patterns are tending to weaken. This reflects the gradual evolution of food consumption into an enjoyment commodity in the eastern part of the country, whereas the central–western regions’ food consumption is still at the basic commodity stage. Under the influence of market expectations and asset allocation, the housing consumption in the eastern region shows an inverted U-shaped curve, while the central–western regions’ consumption shows a U-shaped curve. In terms of the trends of developmental consumption, that of eastern urban households has increased rapidly with the increase in income, and the consumption of transportation, education, entertainment, and medical services is obviously higher than that of central and western urban households. In addition, although the consumption patterns of households in the eastern and central–western regions tend to synchronize, compared with the central–western regions’ consumer market structure, the eastern consumer market structure is more diversified, which indicates the obvious dual characteristics of China’s consumer market. This duality is becoming an important factor affecting the upgrading of inter-regional consumption patterns.

Heterogeneity can be seen in the consumption patterns of different income groups in different regions. The central–western regions’ income gap has a more significant impact on the heterogeneity and imbalance in the consumer market, which has also led to a serious imbalance in the consumption patterns of households in the central and western regions. The income gap in eastern cities has little influence on the imbalance in the consumer market. The main results are as follows: in the central and western cities, the consumption tendencies of each income group are more diversified under the influence of income differences, and the trends of upgraded consumption in each income group are significantly different. For example, the trend of upgrades in the food and housing consumption of the low- and middle-income groups was obvious. The middle- and high-income groups had a higher tendency to upgrade their consumption of housing and medical services, while the high-income group had a higher tendency towards developmental consumption, in accordance with the eastern cities, but for the other goods, the difference from the eastern region was greater. In the eastern region, housing consumption, as the watershed in the shift from basic consumption to developmental consumption, showed high consistency and has become one of the main restraining factors that hinder the evolution of household consumption to higher levels. That is to say, after the problem of housing consumption was resolved, the tendency of households’ consumption towards developmental consumption and enjoyment consumption has increased gradually.

6.2. Discussion

In this study, we found that for China’s emerging cities to build an international consumption center city, they should first clarify their own developmental stages and developmental positioning. The gap between the consumption patterns of the eastern cities, represented by Beijing, and that of the main consumption cities around the world is relatively small, and with the rapid development of the regional economy, a global regional consumption center can be formed in a short period of time. The central–western regions’ differences from the consumption patterns of households in major consumer cities around the world indicate the need to optimize the structure of the urban consumption markets in the longer term; at the same time, this may allow the formation of a differentiated consumer market in a city’s characteristic field, e.g., tourism centers, cultural centers, science and technology centers, and so on.

Secondly, China needs to improve the dual-consumer market structure and remove the constraints that hinder the upgrade of household consumption. Due to the differences in the development of China’s consumer market, a “dual-consumer” market has formed, that is, the gap between the consumer market in the eastern cities and that in the central and western cities. To this end, we should continue to deepen the supply-side reform in the eastern region, improve the content and quality of the development-oriented industry, and promote the development of the enjoyment-oriented industry. This way, we will accelerate the restructuring of the consumer market and promote upgrades of the consumption patterns. In the central–western regions, we should pay attention to growth in household incomes, speed up the transformation and upgrade of the basic consumer industries, and gradually improve the supply structure of the developmental consumption market and the enjoyment market to narrow the gap in the consumption patterns between the eastern and central–western regions. Ultimately, we must promote the evolution of the “dual-consumption” market structure towards a single-consumption market structure.

Finally, we must improve the structural system of social welfare. In addition to the gap in the consumption patterns and income, the gap in the social welfare system is also an important factor that causes the upgrade of household consumption. In the eastern part of the country, middle–high-income households and high-income households are close to the level of international consumption centers in terms of their developmental consumption patterns, but if the government subsidies in developed countries for education and healthcare were excluded, the average level of real household expenditure on developmental consumption would be much higher than that of Chinese urban households. Due to the heterogeneity of urban areas and the income levels of urban households in China, the externalities of the “one size fits all” welfare policy has led to less efficient policies. In this regard, from this study, we believe that the process of economic development in China should continue to improve the structure of the social welfare system, which would help residents upgrade their consumption, effectively improving people’s sense of well-being. For example, the eastern region should focus on the transportation and communication consumption by low- and middle-income groups, and middle-income groups should focus on the consumption of housing and education and entertainment. For middle–high-income and high-income groups, the focus is on medical services. In the central–western regions, priority should be given to housing for low- and middle-income households, education for middle-income households, and transportation, communication, and medical services for middle- and high-income households. This may eliminate the consumption bottleneck of households with different income levels in different regions, releasing the consumption potential of households with different income levels, and promote the upgrade of urban household consumption, which is a requirement for promoting national economic productivity, and it is also necessary to ensure the people’s quality of life.