资源描述
Pandemic, Mobile Payment, and Household Consumption: Micro-Evidence from China Taixing Liu , Beixiao Pan, and Zhichao Yin School of Finance, Capital University of Economics and Business, Beijing, China ABSTRACT The novel coronavirus disease (COVID-19) outbreak has signifi- cantly affected many lives, as indicated by widespread lock- downs and restrictions. This study investigates the impact of COVID-19 on Chinese household consumption. It employs the China Household Finance Survey (CHFS) data and finds that there was a significant decline in household consumption dur- ing the outbreak period. Further heterogeneity analysis shows that the pandemic suppresses consumption in urban house- holds; rural households are, however, less affected. Moreover, mobile payment promotes urban household consumption dur- ing the pandemic, while rural households remain unaffected. KEYWORDS COVID-19 pandemic; household consumption; mobile payment; urban-rural gap JEL CLASSIFICATION D12; E21; G51. 1. Introduction The coronavirus (COVID-19) outbreak in December 2019 attracted consider- able media attention (Haroon and Rizvi 2020) and generated global panic (Ali, Alam, and Rizvi 2020), thereby causing much of the economy to halt (Narayan 2020). Moreover, the outbreak has significantly affected many lives, as indi- cated by the widespread lockdowns and restrictions to prevent further infec- tions. This study employs Chinese survey data to investigate the impact of the COVID-19 pandemic on household consumption. As the initial outbreak took off in China and coincided with the Chinese Lunar New Year (which is a traditional festival characterized by high consumption levels), 1 the resultant rapid spread disrupted the consumption plans of many families. Moreover, it has deeply affected consumption behavior in the short- or even long-term. Thus, the existing literature on the impact of the pandemic on consumption can be divided into short- and long-term impact research. Regarding short- term impact studies, irrespective of the focus on the overall changes in con- sumption (see Barro and Ursa 2008; Barro, Ursa, and Weng 2020; Chen, Qian, and Wen 2020; Correia, Luck, and Verner 2020), different category changes in consumption (see Jung et al. 2016), and dynamic changes in consumption (see Baker et al. 2020), most studies employ macro- or micro- level data. However, they either exist the small sample issue or lack sample CONTACT Zhichao Yin School of Finance, Capital University of Economics and Business, Beijing, China EMERGING MARKETS FINANCE AND TRADE 2020, VOL. 56, NO. 10, 23782389 doi/10.1080/1540496X.2020.1788539 2020 Taylor few studies further discuss heterogeneity. Regarding long-term impact studies, scholars posit that isolation creates new consumption demands and patterns. Thus, the more the threat of the con- tagion prolongs, the further adaptive responses become ingrained and resis- tant to reversal (Cohen 2020). However, such studies do not discuss the consumption patterns that change during the pandemic. Thus, unlike Jung et al. (2016), this study employs nationally representative household sampling survey data of China to address the shortcomings of having a small sample size in macro-level data and resolve the serious problem of sample representation in micro-level data. The study finds that COVID-19 suppresses Chinese household consumption significantly. Secondly, from the perspective of the urban-rural dual economy structure, 2 there is considerable heterogeneity in consumption changes between Chinese urban and rural households. Specifically, urban households suffer more declines; rural house- holds are, however, less affected by the pandemic. Thirdly, since the long-term isolation results in significant changes in consumption patterns (Cohen 2020), as reflected in the transition from offline to online consumption, we further investigate the impact of mobile payment on household consumption during the pandemic period. 3 The results show that mobile payment promotes urban household consumption. Hence, this study bridges the research gap in the literature by answering the following questions. How has household consumption responded to the pan- demic? Is there heterogeneity among families? Can mobile payment change the impact of the pandemic on household consumption? By addressing these questions, we can obtain a more intuitive understanding of consumption adjustments during the pandemic and contribute to the literature on the evaluation of COVID-19. Moreover, we can provide empirical evidence for the formulation of consumption policies during (and even after) the pandemic. We arrange the data of confirmed, recovered, and death COVID-19 cases of Chinese prefecture-level cities and match the data with the nationally repre- sentative sampling survey data to evaluate the impact of COVID-19 on house- hold consumption during the Spring Festival. Our results show that household consumption is significantly affected by the pandemic; consumption continues to decline as the effect increases in severity. Further, we explore the heterogeneity between urban and rural families regarding the differences in consumption patterns and human capital. Considering consumption patterns, China is a typical urban-rural dual econ- omy. Moreover, there are significant differences in consumption patterns between urban and rural families. Considering commodity consumption, rural families mostly own the right to use village collective lands, which enables them to engage in agricultural production to provide surplus EMERGING MARKETS FINANCE AND TRADE 2379consumer goods for sale and satisfy their own consumption. Meanwhile, urban families always act as the demand side in the consumer goods market. Regarding human capital, given the low income in agricultural production, many rural labor transfers from agricultural production to non-agricultural sectors, such as the construction and service industries, comprise a large group of migrant workers. Generally, the human capital of migrant workers is relatively low. Thus, most workers engage in low-skilled jobs in the manufac- turing, construction, and service industries. The pandemic has significantly affected many low-skilled jobs, which may lead to a sharp rise in the unem- ployment risk among migrant workers. Relatively urban families have higher human capital. Thus, they face lower unemployment risk and income uncer- tainty. However, our results show that the negative impact of the pandemic on household consumption is more obvious in urban families, while the self- sufficient rural families are less affected. The differences in consumption patterns dominate the impact of the pandemic on consumption. Moreover, given that consumption patterns changed significantly during the pandemic, we evaluate the function of new payment tools. Notably, mobile payment tools can induce the transition from offline to online consumption, thus overcoming space-time limitations, reducing unnecessary personnel mobility, and meeting the needs of consumers and businesses during the pandemic. Mobile payment plays a significant role in promoting consumption during the pandemic; however, it is only evident in urban families. This study contributes to the literature in three aspects. First, it is first to employ nationally representative sampling survey data to evaluate the impact of COVID-19 on household consumption. Insight from the findings addresses the overall situation in China regarding how household consumption responds to the pandemic on average. Second, this study employs the urban- rural dual economy structure as the division criteria to examine heterogeneity among urban and rural households. Nearly 600 million rural residents in China are not significantly affected by the pandemic. Thus, from the perspec- tive of urban-rural economic activity differences, 4 consumption continues to decline as the dependence on the consumer goods market (urban households) increases. The study findings can supplement the literature on the heteroge- neous impact of the pandemic on household consumption. Third, long-term isolation creates new consumption demands and patterns (Cohen 2020). Offline consumption may be significantly restricted, while online consump- tion may be less affected, given mobile payment. Thus, we further investigate whether mobile payment can alleviate household consumption during the pandemic. We find that mobile payment promotes urban household con- sumption during the pandemic, while rural households are unaffected. The findings can encourage the government to put more resources into developing the mobile payment market. 2380 T. LIU ET ALThe rest of the paper is structured as follows. Section 2 provides an in-depth discussion of the data and describes the empirical specifications employed. Section 3 analyzes the estimation results. Section 4 concludes. 2. Data and Empirical Methodology The data employed in the study can be grouped into two. The first group presents the confirmed, recovered, and death cases released by the health commission of various prefecture-level cities in China. The second group presents the data from the sampling survey on the impact of the COVID-19 pandemic on production and life. The survey was conducted by the Survey and Research Center for China Household Finance in the first quarter of 2020. The sample is partly (mostly) derived from a randomly selected individual sample (the 2019 CHFS individual sample, which is adjusted according to the sam- pling weight). 5 The samples are nationally representative, with which we match the data from the various prefecture-level cities to derived a sample size of 2,767. The online survey comprised two batches, and the questionnaire was launched from February 12 to March 22, 2020. The ques- tionnaire includes subjects such as income and consumption, industrial and commercial operation, and financial market investment. We use the ordinary least square (OLS) method to analyze the impact of the pandemic on household consumption. The econometric model was set as follows: Consumption hc 0 0 Confirmed c 0 X ihc 0 Z hc c ihc (1) From the pandemic survey questionnaire, Consumption refers to the changes in household consumption during the 2020 Spring Festival, as compared with 2019. Confirmed is the core explanatory variable that measures the pandemic intensity. X ihc is the control variable at the individual interviewee level, Z hc is the control variable at the household interviewee level, c is the city-level fixed effect, and ihc is the error term. The two batches of questionnaires on the choice of consumption changes differ in the following ways. In the first batch, the consumption options include a sharp decrease, a slight decrease, basically unchanged, a slight increase, and a sharp increase. The second batch includes a decrease of more than 50%, decreases by 30% to 50% and 10% to 30%, and a decrease of less than 10%, as well as basically unchanged. It also includes an increase of less than 10%, increases by 10% to 30% and 30% to 50%, and an increase of more than 50%. We categorize the decrease of more than 30%, decrease of less than 30%, increase of less than 30%, and increase of more than 30% under sharp decrease, slight decrease, slight increase, and sharp increase, respectively, to integrate the two batches of consumption data. 6 We then set 1 to sharp decrease, 2 to slight decrease, in that order, until 5 is set to sharp increase. EMERGING MARKETS FINANCE AND TRADE 2381Regarding the pandemic intensity variable, we first employ the number of newly confirmed cases during the Spring Festival to measure the pandemic intensity. Moreover, since the Spring Festival period coincided with the initial outbreak of COVID-19 in China, we also employ the number of cumulative and existing confirmed cases at the end of the Spring Festival to measure the pandemic intensity. We add 1 to the three indicators and take the logarithm to measure the pandemic intensity. Furthermore, the COVID-19 pandemic situation is exceptional in Wuhan. For instance, Wuhan observed the first confirmed cases more than one and a half months earlier than other cities. Thus, we exclude the Wuhan household samples in our estimation to avoid outliers that can interfere with the results. We also control for several interviewee characteristics, following Campbell and Cocoo (2007) and Li and Chen (2014). There is strong evidence from the literature that households are heterogeneous in several dimensions. The age variable captures the consumption difference in the life cycle of different periods. Moreover, following Li and Chen (2014), we include the gender, marital status, educational background, and occupation of respondents to avoid missing variable problems since they might impact household consumption. Several household characteristics that may affect household consumption are also included in our model. Moreover, it is intuitive and necessary to consider the family size variable in the consumption model (Campbell and Cocoo 2007) because the larger the population, the higher the consumption. We also control for the children and elderly ratios, following Li and Chen (2014), to capture the demography-induced consumption difference. Moreover, according to Dynan (2012) and Kukk (2016), household income and asset variables largely reflect the familys economic situation and deter- mines its consumption ability. Thus, we include them in our model. Moreover, to capture the impact of inherent differences or heterogeneity (such as cultural environment, regional consumption habits, and savings preferences) at the regional level on household consumption, we control for the city-level fixed effect following Chen, Qian, and Wen (2020). Table 1 presents the descriptive statistical results of variables. 3. Results and Discussion Table 2 presents the estimation results of the OLS method. We find that pandemic intensity has a significantly negative impact on household con- sumption. From column 1, the pandemic coefficient is 0.095, which indicates that for every 10% increase in newly confirmed cases, household consumption decreases by nearly 0.01. Since the consumption variables are sequencing indicators, it is difficult to conduct a quantitative analysis. However, when viewed from the macro-level (such as city, province, and even national level), 2382 T. LIU ET ALthe impact is significant. The explanation of the result can be approached from two perspectives. Areas with dire pandemic situations will not only adopt stricter countermeasures but also reduce the freedom of movement, given the high risk of infection. Columns 2 and 3 employ the number of cumulative and existing confirmed cases to measure the pandemic intensity, and the same results were obtained. Table 1. Descriptive Statistics. Obs. Mean Std. Dev. Min Max Consumption 2575 2.201 1.085 1 5 Confirmed_Added 313 68.543 221.727 1 2410 Confirmed_Cumul 3
展开阅读全文