Heterogeneous Impacts of COVID-19 on trade: Evidence from China’s province-level data

Abstract

This study investigates how the coronavirus disease 2019 (COVID-19) pandemic affected China’s province-level trade. Using data on trade for January to April 2019 and 2020 and the number of confirmed COVID-19 cases or deaths, we empirically examine the effects of COVID-19-related damage on provincial exports and imports separately. We find that, on average, the COVID-19 pandemic did not lead to a reduction in provincial trade. However, we find negatively significant effects on trade in some provinces (e.g. those with a smaller number of deaths during a past coronavirus outbreak) and some industries (e.g. labor-intensive industries).

KEYWORDS:

• COVID-19
• International trade
• China

JEL classifications:

• F15
• F61
• I18

Acknowledgements

We would like to thank an anonymous referee, Kyoji Fukao, Shujiro Urata, Hitoshi Sato, Satoru Kumagai, and seminar participants in the Institute of Developing Economies for their invaluable comments and suggestions. We gratefully acknowledge financial support from the JSPS in the form of various KAKENHI Grants (20K01644 and 20K01666 to Cai and 19H00594 to Hayakawa). Cai also acknowledges support from the Nanzan University Pache Research Subsidy I-A-2 for the 2021 academic year. The data that support the findings of this study are available from IHS Markit. Restrictions apply to the availability of these data, which were used under license for this study. There are no conflicts of interest to declare. All remaining errors are ours.

Disclosure statement

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

Notes

1 The severe acute respiratory syndrome experience prompted China to pass and enact the Emergency Response Law on November 1, 2007. The Law stipulates that when local governments activate the Level 1 (highest among four levels of severity) Public Health Emergency Response to ‘extremely serious incidents,’ they need to delimitate the control area, lock down the epidemic area, take compulsory actions (i.e., constraining or suspending work, production, and business), implement mobility restrictions, establish quarantine stations, and collect information from each community.

2 Antràs, Redding, and Rossi-Hansberg (2020) provided the theoretical framework for this nexus.

3 Friedt and Zhang (2020) also examined the heterogenous effects of COVID-19 on China’s trade but focused on the characteristics of foreign countries, such as their income levels or geographical proximity to China.

4 https://vis.ucloud365.com/ncov/data_en.html

5 https://connect.ihsmarkit.com/gta/home. One of the remarkable features of this database is the inclusion of monthly trade data. The original source is customs in each country. For China, the data comprise all merchandise passing through China’s customs territory. The trade data are also available according to provinces and trade regimes (see footnote 8). The aggregated values (e.g., total exports or total imports) in the Global Trade Atlas are exactly the same as those reported in China’s customs.

6 Friedt and Zhang (2020) examined China’s trade in January and February 2020 separately by splitting their sum according to the January and February shares in 2019. However, we did not adopt this approach because it seems problematic for at least the following two reasons. First, COVID-19 may have changed production and trade patterns, and thereby the monthly shares, dramatically. Thus, the monthly shares might be considerably different between 2019 and 2020. Second, the Chinese New Year fell in February in 2019 and in January in 2020. Thus, other things being equal, the trade values in 2020, relative to those in 2019, were likely to be smaller in January and larger in February. Thus, we decided to aggregate the trade values by two months.

7 Although Friedt and Zhang’s (2020) empirical model is slightly different from ours, they also found a similar magnitude of negative effects on exports, which was around 0.04%.

8 We also estimate our model for ordinary trade and processing trade separately. ‘Ordinary trade’ includes trade with regimes of barter trade, donation by overseas Chinese, goods on consignment, goods on lease, and ordinary trade. ‘Processing trade’ includes trade with regimes of compensation trade, entrepot trade by bonded area, equipment for processing trade, equipment imported into export process zones, outward processing, process and assembling, process with imported materials, and warehousing trade. In both exports and imports during January–April 2019, ordinary trade accounted for around 60% of total trade while processing trade accounted for 37%. We did not find significant differences in the effects of COVID-19 between ordinary and processing trade. The results are available upon request.

9 Fernandes and Tang (2020) examined the effects of SARS on firm-level trade, finding significantly negative impacts.

10 More specifically, we first compute this share for around 100 non-services sectors in the input-output table. Second, by using the converter table between Harmonized System (HS) eight-digit codes and these input-output sectors, we obtain the share for each HS eight-digit code. Last, we take a simple average of the share by HS two-digit codes.

11 Specifically, the World Trade Organization’s HS codes include 340111 and 340130 (hand soap), 340220 (other cleaning products), 382499 (hand sanitizer), 392690, 630790, and 902000 (face masks), and 900490 (protective goggles and visors).

http://www.wcoomd.org/-/media/wco/public/global/pdf/topics/nomenclature/covid_19/hs-classification-reference_2_1-24_4_20_en.pdf?la=en.

12 Several policy reports discuss China’s ‘mask diplomacy.’ Examples include Park et al. (2020), Baldwin and Freeman (2020), Kahn and Prin (2020), Verma (2020), White (2020), and Wong (2020).

Additional information

Funding

This work was supported by JSPS: [Grant Number 19H00594,20K01644,20K01666]; Nanzan University: [Pache Research Subsidy I-A-2].