The Impact of the COVID-19 Pandemic on the Global Value Chain of the Manufacturing Industry

Abstract

This paper adopts the GDYN model to estimate the dynamic impact of the COVID-19 pandemic on global manufacturing industry and the value chain. Our simulation finds that (1) In the short run, the low-tech manufacturing industries will suffer greater shocks, with a decline of output growth in 2021 by 6.0%. The growth rate of the high-tech manufacturing industry showed an increasing trend of 3.7% in 2021. (2) In the post-epidemic period, the total manufacturing output will return to the baseline level, from which the growth rate of low-tech manufacturing will rebound, demonstrating a V-shaped development trajectory. (3) From the perspective of Global Value Chain (GVC), the participation in GVCs of manufacturers in countries along the Belt and Road, the European Union and the United States will weaken, while China’s manufacturing industry has witnessed an obvious improvement in export competitiveness. The import added value of China has decreased, which shows that its ability to meet domestic demand has been improving. This indicates that the COVID-19 pandemic is providing a crucial opportunity for China to upgrade its manufacturing value chain, which contributes to the accelerated construction of a new dual-cycle development pattern.

1. Introduction

The COVID-19 epidemic continues to rage, putting humanity through a public health and economic crisis with far-reaching implications. Preliminary evidence shows that the COVID-19 crisis is considerably more profound than the 2008 Global Financial Crisis [1,2]. According to official data released by the countries, the GDP of the United States, Japan, Germany and the United Kingdom contracted by 3.5%, 4.4%, 4.9%, and 9.9% in 2020 [3,4,5,6]. COVID-19 is already having far-ranging economic consequences, and the end is not yet in sight. Although recent vaccine approvals have raised hopes of a turnaround in the pandemic later this year, renewed waves and new variants of the virus pose concerns for the outlook [7]. The cross-border flow of products and components is still greatly restricted worldwide, signifying a pessimistic outlook on the development trend of the global manufacturing value chain. Comprehensively measuring and predicting the impact of the COVID-19 epidemic on the global manufacturing industry and its Value Chain is an important prerequisite for accurately dealing with the impact of the epidemic and building a new dual-cycle development pattern in the context of globalization.

The Global Value Chain (GVC) theory originated from the value chain theory proposed and developed by international commercial researchers in the 1980s [8,9]. The global value chain trade accounting method can be traced back to the discussion of quantitative vertical specialization in Hummels et al. [10]. They defined a country’s import intermediate input for production and export as vertical specialization. After that, scholars continued to relax the strict assumption of Hummels et al. along this logic and gradually constructed a general accounting formula for value-added trade) [10,11,12]. In particular, Koopman et al. and Wang et al. divided gross exports into four categories: domestic added value eventually absorbed abroad, domestic added value converted back to home after export, foreign added value and pure double counting items, and incorporated relevant indicators such as gross trade, value-added trade, and trade-added value into a unified accounting framework [13,14,15]. On this basis Wang et al. the global value chain indicators are further divided into forward participation and backward participation. From the perspective of forward and backward decomposition, the added value and final products are divided into pure domestic use, traditional trade, simple GVC and complex GVC, and the forward and backward participation are calculated respectively, GVC activities and non GVC activities are distinguished, and the forward and backward participation of GVC activities are calculated [16]. Over the past three decades, the division of labor in the GVC has led to unprecedented rapid growth in the world economy and international trade. At present, the virus continues to mutate, and the development of the new coronavirus epidemic is highly uncertain, causing great concerns about the GVC and the process of economic globalization. Verikios links epidemiological and economic models to capture the transmission from regional populations to regional economies, finding that COVID-19 is likely to be of longer duration and more severe in its economic effects than previous pandemics [17]. Islam and Muyeed use Meta-Analysis approach, exhibiting that this crisis could cost 2.7 trillion U.S. dollars, which is about 3.06% of the global GDP [18]. Guan et al. study the impact of COVID-19 on Global Supply Chain using CGE model based on GTAP framework, finding that regardless of the strategy, the complexity of global supply chains will magnify losses beyond the direct effects of COVID-19 [19]. Li and Chen points out that some countries represented by the United States and Japan will further upgrade the plan to return to domestic manufacturing, with more government subsidies to the regression of manufacturing, and the impact on manufacturing supply chains cannot be ignored [20]. Rajak S., Mathiyazhagan K. et al. summarized the requirements of current supply chain stakeholders and used the CSF method to identify 16 key factors. The study found that during the COVID-19 pandemic, social distance, emergency response to the logistics system, and emergency backup facilities are the three most important factors to stabilize the supply chain [21].

Some studies focused on the impact of COVID-19 on specific countries and industries. Walmsley et al. use CGE model to estimate the macroeconomic impacts of mandatory business closures in the U.S. in order to control the spread of the COVID-19, finding that for the three-month scenario, there will be a 20.3% decline in GDP and a decline of 22.4% in the employment of the U.S [22]. Duan et al., evaluate the economic impacts of COVID-19 outbreak on both national and industrial levels by employing quarterly CGE model, revealing that the epidemic may lower China’s economic growth in 2021 by 1.2–2.7% [23]. Zhao and Yang use GDYN model and find that the real GDP of China will increase by 0.9%. To be specific, the added value of textile and clothing, papermaking and printing products, chemical and pharmaceutical products and metal products will rise by 1.2%, 1.8%, 2.2%, and 2.0%. The added value of high-end manufacturing, such as machinery and transportation equipment industries will rise 0.8% and 1.4% respectively [24]. Shen and Xu suggest that the epidemic has led to a large upstream supply shortage in China. Moreover, the energy industry and the electronic information industry are facing high downstream demand [25]. Zhu et al. find that textile clothing, electrical, metal, metal smelting and other industries may have a higher risk of a shift in internal-to-external industry in China [26]. Under the influence of this global public health crisis, many questions arise: What challenges are the global manufacturing industries facing? How will the manufacturing GVC structure change? What role will China play in the GVC in the future?

To date, there have been few quantitative studies on the impact of the epidemic on the global manufacturing value chain. In addition, most of the GTAP model methods used to measure the economic impact of COVID-19 are relatively static or adopt the Chinese CGE model for the 2017 domestic input-output table, which cannot accurately describe the impact of the epidemic on the global manufacturing value chain. Based on the OECD international input-output table, this study expands the GTAP database, adopts the modified GDYN model to measure the dynamic impact of the epidemic on the global high-tech, medium-tech and low-tech manufacturing industries, and analyzes the evolution of the global manufacturing value chain structure from the perspective of value added.

2. Materials and Methods

2.1. Data Sources and Processing

This study adjusts the GTAP10 dynamic database and constructs a standard international input-output table based on GTAP data. For industries, the 65 industries in the GTAP database match 34 industries in the OECD database, and ultimately, 23 manufacturing related categories are retained in this study. The database is summarized according to the national and regional division, and the manufacturing industry in China, the United States, Japan, South Korea, ASEAN nations and the region are analyzed (

Table 1. Country (or region) Division.

Country Division	GTAP Primitive Country Division
Country	Australia (AUS), China (CHN), Japan (JPN), South Korea (KOR), India (IND), United State of America (USA), Germany (DEU), Brazil (RUS), Russia (RUS), South Africa (ZAF), United Kingdom (UK)
Region	ASEAN, Latin America (except Brazil, LAM), European Union (except France and Germany, EU), Belt and Road Initiative (TBTR), other countries in the world (ROW)

Note: Detailed national name in the region; see Schedule 1.

). In terms of trade flow, the flow of goods in the GTAP10 database are divided to distinguish commodity flow and distribution, and the import share in the database is adjusted based on the 2014 OECD International Input Output Table.

According to the European Union’s statistical office, the manufacturing sector is divided into high technology, medium to high technology, medium to low technology and low technology manufacturing. In this paper, according to the existing classification of industries, the medium and high technology manufacturing industry and the medium and low technology manufacturing industry are merged into the medium technology manufacturing industry, so the manufacturing industry is divided into high technology, medium technology and low technology three categories as in

Table 2. Industry division.

Industry Division
Low tech	Textile and clothing; Leather products industry; Wood processing; Papermaking and printing, other manufacturing industries
Medium tech	Petroleum and coal processing; Rubber and plastic manufacturing; Ferrous metal smelting and processing; Non-ferrous metal smelting and processing; Metal products industry; Other fuel processing industries;
High tech	Chemical fiber manufacturing industry; Medical manufacturing industry; Automobile manufacturing; Transportation equipment manufacturing; Computer, communications and other electronic equipment manufacturing (referred to as electronic equipment manufacturing); Electrical equipment; Instrument manufacturing; General and special equipment manufacturing (mechanical equipment manufacturing for short)

.

To ensure that the total input equals the total output in the global and national levels, we adjusted the tax. Equation, in which all income taxes are removed from intermediate input and final demand. In addition, income is divided into income for each component to ensure that the total investment is equal in the three levels of the industry, the country, and the total output. $VIMS\_N_{i,r,s}$ represents the total amount of goods imported by country s from country r, $VIG{M}_{i,s}$ is the total amount of final consumer goods imported by the government of country s, $VIP{M}_{i,s}$ is the total amount of final consumer goods imported by households in country s, and $VIF{M}_{i.r.s}$ is the total amount of intermediate products imported from r by manufacturers in country s. $shr\_F_{t,r,s,“Government”}$, $shr\_F_{t,r,s,“\mathrm{private}”}$ and $shr\_I_{i.r.j.s}$ are the share of a country’s government’s source of final consumer products, the share of household consumption’s source of final products, and the flow of imported intermediate products among different industries in the country calculated using the 2014 OCED Inter-Country Input-Output Table. (Equation (1)).

$$\begin{array}{ll}VIMS\_N_{i,r,s}& =VIG{M}_{i,s}\times shr\_F_{t,r,s,“Government”}\\ & \stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}+VIP{M}_{i,s}\times shr\_F_{t,r,s,“\mathrm{private}”}\\ & \stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}\stackrel{}{{\displaystyle }}+{\displaystyle \sum _{jj}VIF{M}_{i.r.s}}\times shr\_I_{i.r.jj.s}\end{array}$$

(1)

The data of consumption, unemployment and investment used in the simulation are all from the annual data of major economies in the wind database. and HP filtering is used to remove the cycle term for the consumption, unemployment, and investment data, and strive to obtain the true impact of the COVID-19 decrease trend, and then find the rate of decline caused by the COVID-19.

The equivalent tariff data is calculated based on Minor and customs clearance time data on the World Bank’s doing business website [27].

2.2. GYDN Model Improvements

2.2.1. GYDN Model

This study is based on the application of a Computable General Equilibrium (CGE) model, which has been applied successfully to examine economic impacts of health threats [28]. Global Trade Model (GTAP) is one of the most widely used CGE models. This model is a global static equilibrium model developed by Professor Hertel Thomas from Purdue University and has the extended capability for examining employment and supply chain impacts linked to economic activity and policies [29].

In particular, we use the Global Dynamic General Equilibrium Model (GDYN) is a dynamic extended version of the GTAP model, developed by Ianchovichina and Walmsley [30]. The GDYN model has two outstanding advantages: first, it introduces a dynamic mechanism of investment and capital accumulation, allowing investment to be allocated on a global scale according to the difference in the rate of return of various countries, thus having an impact on the total capital of each country. It also allows fierce competition. Second, asset income is distributed globally according to the relationship of asset ownership, which accurately matches the relationship between capital ownership and the corresponding income; that is, assets in a country are not entirely owned by residents of that country, making the model analysis more accurate.

Based on the database’s dimension setting, the framework of the original GYDN model is modified, and the GYDN model is connected to the GVC decomposition model. In addition, the front and rear indicators of the GVC model are expanded, and the source of the added value is defined. The specific improvement is as follows.

As shown in Figure 1, first, the GYDN model framework is improved. The dimensions of the partial coefficient are extended, and variables and related formulas are added based on the original model. In addition, the elastic coefficients in the model are adjusted to portray the elasticity difference between the intermediate components and the final product. New variables are defined, the CES demand function is used to define new variables to describe the source and distribution of imports. Specifically, it includes the consumption variables of enterprises ($qfm{s}_{i,r,j,s}$ and $pfm{s}_{i,r,j,s}$), households ($qpm{s}_{i,r,s}$ and $ppm{s}_{i,r,s}$) and governments ($qgm{s}_{i,r,s}$ and $pgm{s}_{i,r,s}$), as well as the relationship between import and export trade volume. See Equations (2)–(5).

$$qfm{s}_{i,r,j,s}=\frac{qf{m}_{i,j,s}}{am{s}_{i,r,s}}\times {\left[\frac{1}{am{s}_{i,r,s}}\times \frac{pfm{s}_{i,r,j,s}}{pf{m}_i}\right]}^{-ESUBN{f}_{j,s,i}}$$

(2)

$$qpm{s}_{i,r,s}=\frac{qp{m}_{i,s}}{am{s}_{i,r,s}}\times {\left[\frac{1}{am{s}_{i,r,s}}\times \frac{ppm{s}_{i,r,s}}{pp{m}_{i,s}}\right]}^{-ESUBM{h}_{s,i}}$$

(3)

$$qgm{s}_{i,r,s}=\frac{qg{m}_{i,s}}{am{s}_{i,r,s}}\times {\left[\frac{1}{am{s}_{i,r,s}}\times \frac{pgm{s}_{i,r,s}}{pg{m}_{i,s}}\right]}^{-ESUBM{h}_{s,i}}$$

(4)

$$qx{s}_{i,r,s}=qi{m}_{i,r,s}$$

(5)

The second improvement is redefine some of the equations and use the weighted average method to link the original variable with the new variable. Among them, the prices of enterprises ($pf{m}_{i,j,s}$), households ($pp{m}_{i,s}$) and the government ($pg{m}_{i,s}$), as well as the prices ($pi{m}_{i,s}$) and quantity ($qi{m}_{i,r,s}$) of imports are redefined, see Equations (6)–(10) for details;

$$pf{m}_{i,j,s}={\displaystyle \sum _r\left[\frac{VIF{A}_{i,r,j,s}}{{\displaystyle \sum _{r}VIF{A}_{i,rr,j,s}}}\times \left(\frac{pfm{s}_{i,r,j,s}}{am{s}_{i,r,s}}\right)\right]}$$

(6)

$${\displaystyle \sum _{r}VIP{A}_{i,r,s}\times pp{m}_{i,s}=}{\displaystyle \sum _{r}VIP{A}_{i,r,s}\times \frac{ppm{s}_{i,r,s}}{am{s}_{i,r,s}}}$$

(7)

$${\displaystyle \sum _{r}VIG{A}_{i,r,s}\times pg{m}_{i,s}=}{\displaystyle \sum _{r}VIG{A}_{i,r,s}\times \frac{pgm{s}_{i,r,s}}{am{s}_{i,r,s}}}$$

(8)

$$pi{m}_{i,s}={\displaystyle \sum _r\left(\frac{VIM{S}_{i,r,s}}{{\displaystyle \sum _{k}VIM{S}_{i,k,s}}}\times \frac{pm{s}_{i,k,s}}{am{s}_{i,k,s}}\right)}$$

(9)

$$qi{m}_{i,r,s}=\left[{\displaystyle \sum _j\left(\frac{VIF{M}_{i,r,j,s}}{VI{M}_{i,r,s}}\times qfm{s}_{i,r,j,s}\right)}\right]+\left(\frac{VIP{M}_{i,r,j,s}}{VI{M}_{i,r,s}}\times qpm{s}_{i,r,s}\right)+\left(\frac{VIG{M}_{i,r,j,s}}{VI{M}_{i,r,s}}\times qgm{s}_{i,r,s}\right)$$

(10)

The third improvement the link between the GYDN model and the GVC model. The front and rear decomposition formulas are expanded on the basis of Koopman et al. (2014) and Wang Zhi et al. (2017b), value-added trade is transferred to value-added income (GVC income), the source of the value added is defined, and the front and rear summed results are calculated in Equations (11) and (12).

$$\begin{array}{ll}{\left(V{a}^s\right)}^{\prime }& ={\widehat{V}}_{}^s\times L_{}^{ss}\times Y_{}^{ss}+{\widehat{V}}_{}^{ss}\times L_{}^{ss}\times {\displaystyle \sum _{r\ne s}^G{Y}_{}^{sr}}\hspace{0.33em}\\ & +{\widehat{V}}_{}^{ss}\times L_{}^{ss}\times {\displaystyle \sum _{r\ne s}^G{A}_{}^{sr}}\times L_{}^{rr}\times Y_{}^{rr}+{\widehat{V}}^s\times L^{ss}\times {\displaystyle \sum _{r\ne s}^G{A}^{sr}}\times {\displaystyle \sum _u^G{B}^{ru}}\times Y^{us}\\ & +{\widehat{V}}^s\times L^{ss}\times \left({\displaystyle \sum _{r\ne s}^G{\displaystyle \sum _{t\ne s}^G{A}^{st}}{\displaystyle \sum _u^G{B}^{tu}}}{Y}^{ur}-{\displaystyle \sum _{r\ne s}^G{A}^{sr}}\times L^{rr}\times Y^{rr}\right)\end{array}$$

(11)

$$\begin{array}{ll}{Y}^{\prime }& =V_{}^s\times L_{}^{ss}\times {\widehat{Y}}_{}^{ss}+V_{}^{ss}\times L_{}^{ss}\times {\displaystyle \sum _{r\ne s}^G{\widehat{Y}}_{}^{sr}}\\ & +V_{}^{ss}\times L_{}^{ss}\times {\displaystyle \sum _{r\ne s}^G{A}_{}^{sr}}\times L_{}^{rr}\times {\widehat{Y}}_{}^{rr}+V^s\times L^{ss}\times {\displaystyle \sum _{r\ne s}^G{A}^{sr}}\times {\displaystyle \sum _u^G{B}^{ru}}\times {\widehat{Y}}^{us}\\ & +V^s\times L^{ss}\times \left({\displaystyle \sum _{r\ne s}^G{\displaystyle \sum _{t\ne s}^G{A}^{st}}{\displaystyle \sum _u^G{B}^{tu}}}{\widehat{Y}}^{ur}-{\displaystyle \sum _{r\ne s}^G{A}^{sr}}\times L^{rr}\times {\widehat{Y}}^{rr}\right)\end{array}$$

(12)

L is the local Leontief inverse matrix, B is the global Leontief inverse matrix, A is the direct consumption coefficient matrix, V is the value added coefficient matrix, and Y is the final demand matrix. Taking Equation (11) as an example, ${\widehat{V}}_{}^s\times L_{}^{ss}\times Y_{}^{ss}$ represents the domestic production and consumption of a country, ${\widehat{V}}_{}^{ss}\times L_{}^{ss}\times {\displaystyle \sum _{r\ne s}^G{Y}_{}^{sr}}$ indicates the value-added flow of a country’s traditional trade, ${\widehat{V}}_{}^{ss}\times L_{}^{ss}\times {\displaystyle \sum _{r\ne s}^G{A}_{}^{sr}}\times L_{}^{rr}\times Y_{}^{rr}$ indicates the flow direction of added value of a country’s simple GVC and ${\widehat{V}}^s\times L^{ss}\times {\displaystyle \sum _{r\ne s}^G{A}^{sr}}\times {\displaystyle \sum _u^G{B}^{ru}}\times Y^{us}+{\widehat{V}}^s\times L^{ss}\times \left({\displaystyle \sum _{r\ne s}^G{\displaystyle \sum _{t\ne s}^G{A}^{st}}{\displaystyle \sum _u^G{B}^{tu}}}{Y}^{ur}-{\displaystyle \sum _{r\ne s}^G{A}^{sr}}\times L^{rr}\times Y^{rr}\right)$ represents flow direction of added value of a country’s complex GVC.

In Equation (12), $V_{}^s\times L_{}^{ss}\times {\widehat{Y}}_{}^{ss}$ represents the part of domestic production and domestic consumption of a country, $V_{}^{ss}\times L_{}^{ss}\times {\displaystyle \sum _{r\ne s}^G{\widehat{Y}}_{}^{sr}}$ represents the source of value added in a country’s traditional trade, $V_{}^{ss}\times L_{}^{ss}\times {\displaystyle \sum _{r\ne s}^G{A}_{}^{sr}}\times L_{}^{rr}\times {\widehat{Y}}_{}^{rr}$ represents the source of value added in a country’s simple GVC, and ${\widehat{V}}^s\times L^{ss}\times {\displaystyle \sum _{r\ne s}^G{A}^{sr}}\times {\displaystyle \sum _u^G{B}^{ru}}\times Y^{us}+{\widehat{V}}^s\times L^{ss}\times \left({\displaystyle \sum _{r\ne s}^G{\displaystyle \sum _{t\ne s}^G{A}^{st}}{\displaystyle \sum _u^G{B}^{tu}}}{Y}^{ur}-{\displaystyle \sum _{r\ne s}^G{A}^{sr}}\times L^{rr}\times Y^{rr}\right)$ represents the source of value added in a country’s complex GVC.

The third improvement is the benchmark scenario for building a dynamic model. Referring to the recursive dynamic method of Chappuis [31] and using the forecast data of international authoritative institutions such as CEPII, IMF and the World Bank, the benchmark scenario for the endogenous GDP growth rate is obtained (2015–2035), ensuring that there is no difference between the endogenous GDP growth rate and the previous exogenous GDP growth rate.

2.2.2. Method of Social Network Analysis

Based on Meng Bo’s method [32], the demand centers and supply centers of the global manufacturing trade are identified using the increased value of imports and exports. From the supply perspective, a country is a value-added supply center if the increased value imports in most countries in the region come from that particular country. In terms of demand, a country is a regional value-added demand center if the value-added exports of most countries in the region flow to that particular country.

2.2.3. Scenario Setting

The decrease of GDP in 2020 reflected the decreases in personal consumption expenditures, exports, private inventory investment, nonresidential fixed investment, and government expenditures [33]. By early May in 2020, the total volume of online job vacancies had fallen by over 50% in five OECD countries (Australia, Canada, New Zealand, the United Kingdom, and the United States) with respect to the beginning of the year, with even larger declines in some sectors [2]. As the limits to personnel flow and travel restrictions directly lead to a decrease in consumption demand and labor supply, the labor-intensive industry is impacted from both the supply and demand sides. Dunn et al. (2020) evaluate the economic effects of the COVID-19 pandemic on consumer spending using daily card transaction data and estimate an aggregate effect of −27.8% on consumer spending after mitigation measures have had time to take hold [34].

From the perspective of international trade, governments have implemented restrictions, closed ports, and strengthened inspections and quarantines, and the epidemic is mainly influencing import demand. In addition, trade facilitation affects the goods trade and manufacturing value chain structure. When the financial section of the chain is impacted, the economic downtown is expected to increase, the trend in domestic international investment slows down, and effective capital stock and capital prices both decline due to the impacts on the international supply chain.

Based on the research and analysis above, this study selects employment, consumption, trade facilitation and capital stock as proxy variables for shock.

Employment Rate

We remove the trend in the unemployment rate of countries with HP filtering for nearly 10 years and retain the cycle costs. The unemployment data are converted into labor data for each year of employment data. On this basis, the employment losses of various countries in the context of the epidemic are calculated to obtain the 2020 employment rate by country. It is expected that the epidemic will continue to affect the employment rate in 2021, which maintained a level approximately 2/3 of the previous average in 2020. Projections are that the epidemic will be controlled in 2022, and the labor supply will gradually be restored to the steady state of the economy.

Consumption

With HP filtering, the trend item is removed; the 2020 consumption growth rates for countries are used to reduce their 2019 consumption growth rate by selecting the most recently published consumption expenditure growth rate of the various countries. Based on this, the decline in consumption is obtained under the influence of COVID-19. In the GTAP model, the relationship between personal private consumption demand, the price of different consumer products and personal private consumption expenditures is as follows in Equation (13):

$$\frac{q{p}_{i,r}}{po{p}_r}=\left({\displaystyle \sum _{k}p{p}_{k,r}{}^{E{P}_{i,k,r}}}\right)\times {\left(\frac{y{p}_r}{po{p}_r}\right)}^{E{Y}_{i,r}}+f\_q{p}_r$$

(13)

In the above formula, $f\_q{p}_r$ is the mobile variable for the consumption of region r. The changes in the preference for private consumption demand caused by impact $f\_q{p}_r$ and the preference for private consumption demand by household are obtained. This study expects to find a long-term impact from the epidemic through final consumption. It is expected that the final consumption at the family level in 2021 will remain at 2/3 of the level in 2020. On this basis, the rebound mechanism is set, assuming the model will return to a steady level in 2025.

Trade Facilitation

According to Minor P and Tsigas M’s GTAP database-based association of different types of imports and exports to the imports and exports of different industries and considering the effects of the epidemic over Time multiplied, equivalent tariffs are determined in various industries [27]. According to the import and export trade volume as a weight, the equivalent tariff results are matched with merged countries and industries. Equivalent tariffs in different industries in two countries are determined by adding import and export equivalent tariffs [35]. This study assumes that the flight restrictions following the epidemic have restricted transportation capacity, countries have increased efforts to review immigration products, enhance inspection and quarantine standards, etc., causing import customs transfer time to increase to 14 days, while export customs transfer time does not change.

Investment Level

The annual investment data comes from IMF. To accurately show the investment changes, we remove the trend items with the method of HP filter. The investment data adjusted by HP filter are matched with the country and region categories in this paper. Investment is assumed to rebound to a steady level after 2022.

2.3. Global Value Chain Participation Index

Global value chain participation indexes the connotation of global value chain participation is the proportion of a country or region’s participation in the global value chain in its added value. The greater the value of the index, the stronger the participation of the country/region in the global value chain; the smaller the value of the index, the weaker the participation of the country/region in the global value chain. Equation (14) VAS_GVC represents the global value chain participation index based on forward decomposition. Equation (15) VAS_GVC_S represents the activity participation index of simple value chain based on forward decomposition. Equation (16) VAS_ GVC_C represents the complex participation index of complex value chain based on forward decomposition. Equation (17) FGYS_ GVC represents the global value chain participation index based on backward decomposition. Equation (18) FGYS_ GVC_ S represents the activity participation index of simple value chain based on backward decomposition. Equation (19) FGYS_ GVC_ C represents the complex participation index of complex value chain based on backward decomposition. The global value chain forward participation index aims to answer the question “what proportion of production factors are used by countries/sectors for global value chain production”; The global value chain backward participation index aims to answer “what proportion of national/sectoral production inputs for final products are based on global value chain production activities”. Compared with the traditional global value chain participation index, the GVC participation index of Wang et al. (2017) perfectly deconstructs the GVC participation at the national and departmental levels, effectively makes up for the shortcomings of the traditional global value chain participation index system, and clarifies the source and destination of added value at different levels from the perspective of global value chain.

$$VAs\_GVC=\frac{VA\_GVC}{V{a}^{\prime }}=\frac{VA\_GVC\_S}{V{a}^{\prime }}+\frac{VA\_GVC\_C}{V{a}^{\prime }}$$

(14)

$$VAs\_GVC\_S=\frac{VA\_GVC\_S}{V{a}^{\prime }}$$

(15)

$$VAs\_GVC\_C=\frac{VA\_GVC\_C}{V{a}^{\prime }}$$

(16)

$$FGYs\_GVC=\frac{FGY\_GVC}{Y^{\prime }}=\frac{FGY\_GVC\_S}{Y^{\prime }}+\frac{FGY\_GVC\_C}{Y^{\prime }}$$

(17)

$$FGYs\_GVC\_S=\frac{FGY\_GVC\_S}{Y^{\prime }}$$

(18)

$$FGYs\_GVC\_C=\frac{FGY\_GVC\_C}{Y^{\prime }}$$

(19)

3. Results and Analysis

Based on the general law of the development of infectious diseases, this study assumes that the development of the epidemic will undergo three stages: Peak period, effective control period, and recession period. In addition, the changes in all periods will affect factors such as employment, consumption, trade facilitation, and capital stocks. (1) In the peak period of the COVID-19 epidemic (2020–2021), businesses face declines in short-term consumption demand, employment and investment and challenges in trade. (2) Once the epidemic is under effective control, the previously suppressed consumer demand, employment and investment will gradually rebound after the epidemic is effectively controlled (2022–2025). (3) In the third stage, various types of elements will gradually return to their normal level, and supply and demand present a stable situation after the epidemic recedes (2026–2035).

In order to further analyze specific industries, this study will carry out analysis on a few representative industries from the three categories. The detailed simulation results and analysis are as follows.

3.1. The Impact on the Output of Global Manufacturing

The simulation results show that by 2035, the prices of production factors, manufacturing output, and total imports and exports of most countries and regions will generally show a downward trend. Due to the large decline in labor wages, the decrease in primary factor prices in countries such as France, the United Kingdom, and South Africa will be significantly affected. The wages of unskilled labor in the above three countries will fall by 21.19%, 16.61%, and 14.96%, and the wages of skilled labor fall by 20.5%, 17.81%, and 15.39%. The wages of skilled labor, wages of unskilled labor, and capital prices in traditional manufacturing countries such as China, Japan, South Korea, and Germany will all declined, resulting in a relatively large negative impact on the prices of primary factors. From the perspective of import and export, India, South Africa, France, the United Kingdom, and Brazil have the largest decline in exports, with 0.55%, 0.50%, 0.44%, 0.42%, and 0.40% respectively. China’s manufacturing imports will fall the most, reaching 0.42%, but the decline in China’s manufacturing exports was limited, only 0.24%. However, the decline in exports in the United States, India, Japan, the United Kingdom, France, Europe, and the United States and other countries/regions is greater than the decline in imports. This indicates that after the epidemic, China’s import dependence on the international market will decrease. Affected by factors such as factor prices, demand, and customs clearance time, the manufacturing output of South Africa, India, and France has dropped significantly, by 0.48%, 0.43%, and 0.39%, respectively.

Due to the length of the value chain and the degree of globalization varies in different industries, the impact of epidemic shows great differences (Figure 2). Global low-end manufacturing industry was largely negatively impacted in the short term compared to the baseline scenario. Its growth rate of output will decrease by 4.2% in 2021 and will gradually return to normal levels after the epidemic is effectively controlled. On the whole, it exhibits a V-type development path, with an output of 1% in 2030. On the other hand, the high-end manufacturing industry will grow rapidly in the short term, with an increasing of output growth rate by 2.15% in 2021, which in turn will drive global manufacturing growth. The overall development trend of the manufacturing industry will gradually return to the benchmark level after the epidemic recedes. We choose a few representative industries for further analysis as in Figure 3.

The medium and low-tech manufacturing industries suffer large impact, especially for textiles and clothing, forest products, papermaking, and printing, with the outputs dropped by 8.4%, 5.8%, and 5.4% in 2021. The main reason for the decline is that most of these industries are labor-intensive. As the epidemic spreads, unemployment keeps rising, and the production cost of those labor-intensive industries greatly increases, resulting in a decline in output. On the other hand, these industries provide a large number of final products (to the consumer). They are affected by declines in consumption and residents’ income, and the output of traditional manufacturing will further decline. In addition, international demand is reduced due to the challenges in transport logistics during the epidemic and the reduced facility of trade. Most manufacturing will gradually return to normal levels after the epidemic recedes, which will further promote industrial output growth for textiles and clothing, forest products, and petrochemical products, which will experience a pulling effect from the rebound of consumption rebound and return of capital flows.

The high-tech manufacturing industries experience relatively small effects from the epidemic in the short term, and some industrial output has increased significantly. The industrial output of mechanical equipment, electronic products, transportation equipment, and electrical appliances grows faster in 2021, with growth rates of 5.5%, 4.5%, 4.0%, and 3.5% compared with the benchmark scenario. The main driver of this growth is that these industries belong to technically intensive industries, with high levels of production automation and digital technology leading to relatively small impacts from unemployment. For example, China’s steel industry has built a “black light factory” that requires little labor; thus, it can easily ensure virus prevention and control during the epidemic while maintaining efficient production operation. The second driver is the importance of investment capital in the above industries. Because the epidemic leads to a large decline in capital, the production cost of these capital-intensive industries is significantly reduced, but demand remains for the output. Third, these industries are far down the chain from final consumption, and so the demand side impact is smaller during the short term. Finally, high-tech manufacturing is situated in the middle of the chain from the perspective of investment; for example, it receives investment in electronic information manufacturing from automotive and parts industries, mechanical equipment, automobiles, and the component industry. As a consequence, the epidemic has a small influence in the short term.

3.2. Impact on the GVC of the Global Manufacturing Industry

After the epidemic, how will the GVC structure of global manufacturing industry change in the long run? Will the changes be different for specific industries? As the world’s largest manufacturing center, what role will China play in the GVC trade? This study identifies the trade relationships between countries according to the source and destination of the value added trade and analyzes the trade between major countries/regions in the world after the epidemic subsides by constructing a social network structure diagram.

First, the former and latter term are decomposed from the perspective of added value trade, and the source and orientation of added value are identified to build trade relationships among countries. The output of an industry can be divided into four parts: pure domestic production, traditional trade, the simple GVC trade, and the complex GVC trade. “Pure domestic production” refers to the added value of a final product produced to be absorbed by the domestic market; “traditional trade” refers to the added value of a final product for trade; “simple GVC” refers to an intermediate input product that is used for local production or for export, which is a single product cross-border export or import; “complex GVC” is two or more cross-border exports or imports of intermediate products in trade. The flow of GVC can be divided considering forward and backward decomposition; forward decomposition adopts the perspective of added value production, see Equation (5) for details, while backward decomposition takes the perspective of added value demand, see Equation (6) for details.

Second, a social network structure diagram is built to show the trade share between trading partners and analyze the changes in the value chain structure. In the social network structure diagram, the size of each node indicates the forward and backward added value of a country, the thickness of the line between countries/regions reflects the share of the forward and backward added value between trading partners in the forward and backward added value of each country, and the arrow represents the direction of flow. What should be noted is that the presence of a relationship between two countries/regions in the social network structure diagram is determined by two conditions. Take forward participation as an example: first, if country A’s share of the value added of country B’s imports is larger, there is an association between country A and country B; second, if country A’s share of the value added of country B’s imports is greater than 25%, country A and country B have an association. Backward participation considers the share of the value added of country A’s exports to country B.

3.2.1. Impact of Novel Coronavirus Pneumonia on the Participation of Global Value Chains at the National Level

As can be seen from

Table 3. Impact of novel coronavirus pneumonia on the participation of global value chains at the national level (Cumulative impact of 2020–2035 relative baseline scenario).

	Forward					Backwards
	VA_D (%)	VA_RT (%)	VA_GVC (%)	VA_GVC_S (%)	VA_GVC_C (%)	FGY_D (%)	FDY_RT (%)	FGY_GVC (%)	FGY_GVC_S (%)	FGY_GVC_C (%)
AUS	−0.01	−0.07	−0.01	0.05	−0.20	−0.01	−0.07	−0.11	−0.05	−0.26
NZL	0.00	−0.05	−0.17	−0.11	−0.34	0.00	−0.05	−0.11	−0.05	−0.21
CHN	−0.03	−0.07	−0.18	−0.07	−0.36	−0.03	−0.07	−0.32	−0.24	−0.47
CAN	0.02	−0.18	−0.08	0.02	−0.29	0.02	−0.18	−0.04	0.05	−0.23
JPN	−0.06	−0.02	−0.19	−0.11	−0.34	−0.06	−0.02	−0.14	−0.08	−0.28
KOR	−0.12	0.04	−0.13	−0.01	−0.31	−0.12	0.04	−0.14	−0.10	−0.22
IND	0.08	−0.35	−0.30	−0.24	−0.44	0.08	−0.35	−0.10	0.02	−0.45
USA	0.05	−0.21	−0.22	−0.14	−0.37	0.05	−0.21	−0.13	−0.04	−0.31
DEU	−0.06	−0.06	−0.12	−0.02	−0.32	−0.06	−0.06	−0.10	−0.04	−0.19
GBR	−0.16	−0.17	−0.25	−0.11	−0.46	−0.16	−0.17	−0.18	−0.08	−0.25
FRA	−0.18	−0.19	−0.18	−0.07	−0.38	−0.18	−0.19	−0.17	−0.08	−0.23
BRA	0.03	−0.16	−0.30	−0.24	−0.47	0.03	−0.16	−0.13	−0.06	−0.33
RUS	0.04	−0.15	−0.14	−0.01	−0.33	0.04	−0.15	−0.07	0.00	−0.28
ZAF	−0.09	−0.04	−0.47	−0.36	−0.67	−0.09	−0.04	−0.20	−0.14	−0.34
ROEP	0.02	−0.12	−0.21	−0.13	−0.40	0.02	−0.12	−0.16	−0.06	−0.33
TBTR	0.03	−0.09	−0.20	−0.11	−0.38	0.03	−0.09	−0.19	−0.10	−0.37
LAM	0.10	−0.26	−0.30	−0.23	−0.46	0.10	−0.26	−0.15	−0.06	−0.37
AssAN	0.02	−0.21	−0.22	−0.15	−0.39	0.02	−0.21	−0.28	−0.14	−0.47
AFR	0.04	−0.17	1.00	−0.04	−0.36	0.04	−0.17	1.00	−0.10	−0.35
ROW	−0.29	0.04	1.00	0.06	−0.29	−0.29	0.04	1.00	−0.24	−0.31

Data source: The GDYN simulation results. Note: the EU region refers to other EU countries except Britain, Germany and France.

, by 2035, COVID-19 had a strong impact on the forward and backward manufacturing industries of the world’s major economies. From a forward-looking perspective, in terms of traditional trade, India, the United States, France, and Canada suffered the most serious impact, with a decrease of 0.35%, 0.21%, 0.19%, and 0.18%, respectively, compared with the benchmark scenario. In terms of GVC, South Africa, Latin America, India, and the United States experienced the largest decline, with a decrease of 0.47%, 0.30%, 0.30%, and 0.22% respectively. In terms of complex GVC, South Africa, Brazil, the United Kingdom, India, and other countries were most affected, with a decrease of 0.67%, 0.47%, 0.46%, and 0.44% respectively. In contrast, the forward participation of China’s GVC has also decreased, but the decline is far lower than the global average, which makes the production capacity of GVC products in China’s manufacturing industry relatively sufficient. From a backward perspective, in terms of GVC, China, ASEAN and South Africa decreased the most, by 0.32%, 0.28%, and 0.19% respectively. Among them, China’s complex GVC decreased by 0.24%, and ASEAN and South Africa’s complex GVC decreased by 0.14%. Different from the forward participation, the countries with a large decline in backward GVC participation are those that were good at processing trade before the epidemic, which shows that COVID-19 may provide an opportunity for the industrial upgrading of the above countries to reduce the dependence of manufacturing intermediates on foreign countries, especially China and ASEAN, and their manufacturing value chain has shown an upward trend.

3.2.2. Structural Changes in the GVC of Traditional Manufacturing

As seen in Figure 4, the simulation results reveal that after the pandemic, the final product trade among the world’s major economies will be relatively stable compared to the base scenario in 2035, and there will be no major change in the value chain structure. From the perspective of export added value, except for the Asia-Pacific region, China’s finished products mainly flow to the United States and countries along the “Belt and Road”; from the perspective of import added value, the United States and the countries along the “Belt and Road” represent the essential sources of China’s final product imports. In 2035, China represents the core of the finished product supply and demand system of manufacturing in the Asia-Pacific region, and there is a close trade relationship among EU countries.

3.2.3. Structural Changes in the Manufacturing GVC from the Angle of Simple GVC

As seen in Figure 5, after the pandemic, the global simple GVC will see a regional development trend. The demand for simple GVC worldwide falls into two main sectors: one is the supply and demand network of the European Union and the countries along the “Belt and Road”, and the other is the “Asia Pacific-North America” GVC trade network with China as the core. There is closer cooperation within the sectors. For instance, the simple GVC trade between China and the major economies in the Asia-Pacific region will be further strengthened, as it will no longer rely on the markets of countries along the “Belt and Road”, and exports and imports of intermediate products to the United States will decrease.

Examining export added value, compared with the baseline scenario, China’s supply scale and status in the Asia-Pacific manufacturing value chain will have improved by 2035, and it will see heightened trade with the United States. It will rely less on the GVC markets of the countries along the “Belt and Road”, and India will be included in its supply network. These changes indicate that the international competitiveness of simple GVC products is increasing. The added value of exports of countries along the “Belt and Road” will show a downward trend; they are closer to the European Union and will exchange less in trade with India. From the perspective of import added value, there will be less import added value for China from the United States, ASEAN countries, Latin America, and countries along the “Belt and Road”, which indicates that China’s domestic products will be beginning to replace imported intermediate products and that China’s ability to meet domestic demand will be strengthened.

3.2.4. Structural Changes in Complex GVC of Manufacturing Industries

In the post-pandemic period, there will be an obvious trend of centralization in the complex GVC of manufacturing industries. The pandemic has offered an opportunity for the transformation and upgrading of China’s manufacturing industry. China will become the largest supply center and demand market in the world, and its core position in the GVC will improve. The main reason behind this change is the deep integration of the industrial chain among China, the United States, the European Union, Japan, and South Korea [36]. In the medium and long term, the decrease in the export supply of the United States, the European Union, Japan, South Korea, and the countries along the Belt and Road has provided an opportunity to fill the global industry gap and transform and upgrade the industrial chain for China.

In general, the global manufacturing value chain will still mainly be represented as two supply and demand networks: one is the complex GVC supply and demand network of the European Union and the countries along the “Belt and Road”, while the other is the complex GVC supply and demand network of Asia Pacific-North America, with China as the core. In terms of export added value, China will step forward to become the core of the global manufacturing complex GVC supply network based on its complete industrial chain and manufacturing cost advantages. The major global economies will all be taken into the supply network of China’s manufacturing complex GVC. From the perspective of import value added, China’s import demand will show a significant decrease, especially for imports from ASEAN countries and the United States. This shows that China’s manufacturing capacity for high-tech products has strengthened, the export competitiveness and ability to meet domestic demand have improved, and China will be striding forward to the high end of the value chain. Although the pandemic will have some implications for the United States, its complex GVC activities and main trading partners change little. The main reasons for this are that, first, the complex GVC production procedures in the United States are principally implemented in China, second, its GVC activities are mainly concentrated at the two ends of the value chain (R&D and consumer demand). Hence, the impact of the pandemic on the status and activities of US manufacturing in the complex GVC is not significant. What is worth noting is that China’s import demand for complex GVC products from the United States will remain relatively large.

4. Discussion

4.1. Modification of the GDYN Model and the Data Base

The application of the traditional CGE model is mostly concentrated in the environment and energy fields. Fujimori et al. used CGE model to study the impact of Japan’s environmental policies on the Japanese economy, and Kapitza et al. quantified the impact of biophysical and socio-economic intermediaries on Vietnam [37,38]. Li et al. use a global adaptive multiregional input–output model and scenarios of lockdown and fiscal counter measures, showing that compared with a no-pandemic baseline scenario, global emissions from economic sectors will decrease by 3.9 to 5.6% in 5 years due to COVID-19 [39]. Till now, there are few studies on the impact of COVID-19 under CGE model. Some scholars have begun to use the GTAP model to study the economic impact of COVID -19 and GVC related problems which are referable for this study [38,39,40,41,42]. However, there is still room for improvement in existing researches. First, the GTAP database has issues when setting the flow direction of products. Since the flow direction of imported intermediates is not explained in GTAP data, previous studies have generally assumed that the flow direction of imported intermediates is represented by the proportion of domestic intermediates in one flow direction, which is obviously inconsistent with actual product flow direction. As a result, this strong assumption leads to two serious concerns: one is that segmentation after the GTAP database is constructed will be carried out in accordance with the input and output table after the merger. After the synthesis, if the input and output portions of the industry’s total output and the total are not equal, if the inverse matrix and local Lyon initial inverse matrix show significant deviation when calculating the global Lyon, the GVC decomposition may lead to incorrect results. Second, there is a forced assumption in the information on the product flow, which makes it difficult to truly reflect the status of intermediate goods trade among countries globally, and the GTAP database cannot depict the dependence of upstream and downstream industries among countries. Any simulation results based on these assumptions will be far from the real results.

To solve the above problems, this study links the GDYN model and WWYZ model to estimate the impact of COVID-19 on the GVC of manufacturing industry. And the world input-output table structure is embedded into the GTAP 10 dynamic database. In the GDYN model and GTAP database, the import source country and domestic flow direction of products are divided. This approach can not only more accurately reflect the status of intermediate goods trade among countries worldwide but also more clearly and carefully depict the dependence of upstream and downstream industries among countries, and the simulation results are more accurate. At the same time, it ensures that the database results will balance global inputs and outputs when summed; that is, the total input and total output of any country and industry are equal, the global and local Leontief inverse matrices obtained are both real and valid, and the GVC decomposition results based on this approach are valid and credible.

4.2. The Impact of Epidemics on Specific Manufacturing Industries

Due to the different industrial chain positions, production characteristics and factor densities of different industries, the structure of the GVC of various industries and the position of countries in the industrial value chain vary. This paper finds that in the post-epidemic period, the importance of China’s manufacturing industry in different types of trade activities will increase—in particular, the export value added of China’s simple GVC products and complex GVC products will increase significantly—and most of the world’s major economies will be included in the supply network of China’s manufacturing complex GVC. The results show that industries with more cross-border intermediate products and higher technical complexity will be “made in China” in the long run. To verify this conclusion, this paper selected the high-tech manufacturing industry and its internal industry, the electronics industry, to conduct a further analysis of the complex GVC product trade.

As seen in Figure 6, Figure 7 and Figure 8 on the supply side (export added value), the complex GVC structure of the high-tech manufacturing industry and electronics industry is basically the same as that of the manufacturing industry. However, it is worth noting that from the perspective of imported added value, China will become the core of the imported added value supply of the electronics industry, indicating that there are certain differences in the structural changes in the complex GVC in the electronics industry and those in the manufacturing industry and high-tech manufacturing industry. This finding shows that with the gradual rise of technical barriers, China’s influence in the GVC will gradually increase, and China’s core position will become more prominent. The overall trend of the impact of the epidemic on the manufacturing industry is consistent, but there will be a certain degree of industry heterogeneity. In general, COVID-19 does provide an opportunity for China’s manufacturing industry, especially high-end manufacturing, to move up the GVC.

4.3. Limitations

As with most studies, the design of the current study is subject to limitations. (1) Therefore, this study may have allocated insufficient time in its design to the development stage of the epidemic. (2) The study focused on the GVC and trade relationships, but it does not incorporate the complex and changeable international economic situation into the simulation scenario. For example, it does not reflect the trade cooperation framework between China, the “Belt and Road” countries and RCEP; at the same time, how China participates in the GVC and its competitive strength need to be further strengthened. (3) The third limitation is that nonmarket factors such as Sino-U.S. trade frictions and Sino-European trade frictions are not fully considered in the simulation mechanism. Therefore, in the context of anti-globalization, China’s participation in GVCs and its competitiveness need to be further studied. In future studies, we will further improve the model structure, the design of simulation scenarios, and the overall international situation as much as possible in the model to make the analysis more accurate.

5. Conclusions

This paper chooses employment, consumption, trade facilitation, and capital as simulation shock variables and adopts the GDYN model to simulate and analyze the dynamic impact of the COVID-19 pandemic on the output and GVC of the manufacturing industry. The main conclusions are as follows:

(1) The extent of the impact is projected to vary significantly across industries. In the short term, the low-tech manufacturing sector will be greatly affected by the epidemic, while the output of some high-tech manufacturing sectors will see a growth trend. Compared with the baseline scenario, the global output of low-tech manufacturing for textile and clothing, forest products, paper and printing, will decline by 8.4%, 5.8%, and 5.4%, respectively, in 2021, due to factors such as reduced labor supply and weak consumer demand. High-tech industries such as mechanical equipment, electronic and optical products, transportation equipment, and electrical equipment will grow by 5.5%, 4.5%, 4.0%, and 3.5%, respectively, in 2021, due to their relatively stable output, which stems from their distance from final consumption and their high digitization.

(2) After COVID-19 subsides, the overall trend of manufacturing output will return to the baseline level. Post-COVID-19 recovery will improve the output of the affected medium- and low-tech manufacturing industries for textiles and clothing, Petroleum and coal processing, and Rubber and plastic manufacturing, which will be 2.3%, 5.2%, and 1.2% higher than the baseline scenario in 2030, showing a V-shaped development path. The output of most high-tech manufacturing industries in 2030 will be lower than that in the baseline scenario. For example, machinery and transportation equipment will decline by 3.4% and 3.0% by 2030, but this trend will slow.

(3) From the perspective of GVCs, the influence of China’s manufacturing industry in GVCs will increase once the COVID-19 pandemic fades, while the participation in GVCs of manufacturers in countries along the Belt and Road, the European Union and the United States will weaken. In 2035, China will become the country with the fastest growth in value-added exports in the GVC, especially in the export of complex GVC products, and Imports will decrease significantly, indicating that the epidemic represents an important opportunity for China to accelerate its move toward the higher end of the value chain. China will continue to rely heavily on complex GVC products from the United States, indicating that the United States will continue to have a certain influence on China’s high-end manufacturing industry in the long term.