key: cord-0931144-csx6rel8
authors: Liu, Yawen; Cui, Qi; Liu, Yu; Zhang, Jinzhu; Zhou, Meifang; Ali, Tariq; Yang, Lingyu; Feng, Kuishuang; Hubacek, Klaus; Li, Xinbei
title: Countermeasures against Economic Crisis from COVID-19 Pandemic in China: An Analysis of Effectiveness and Trade-offs
date: 2021-09-30
journal: Structural change and economic dynamics
DOI: 10.1016/j.strueco.2021.09.017
sha: 505c0baab71faf26bf767a10b51655ea6c9323b5
doc_id: 931144
cord_uid: csx6rel8

The effectiveness of different countermeasures to economic crisis from the public health emergency is still inadequately understood. We establish an illustrative scenario, specifying the shocks of COVID-19 pandemic and countermeasures applying a general equilibrium model to analyze the effectiveness of countermeasures with a particular focus on trade-offs in the impacts of monetary and fiscal policies. We find that both monetary and fiscal countermeasures could effectively mitigate the economic damages to GDP and employment. However, they would also produce adverse side-effects such as an increase in consumer price by 1.05% and 0.57%, respectively, and a decline in exports by 2.61% and 1.05%, respectively. Monetary policies would exacerbate the damages to external demand by supply-side shocks of the pandemic, but they are more suitable for mitigating demand-side shocks. While fiscal policies would benefit nearly all producing sectors, monetary policies would mainly affect export-oriented manufacturing sectors negatively.

The coronavirus pandemic outbreak has caused a severe shock to China's economy due to strict social distancing and massive shutdown of economic activities. The previous studies projected that the COVID-19 would cause significant losses to China's GDP (Ayittey et al., 2020; Duan et al., 2020; McKee and Stuckler, 2020; McKibbin and Fernando, 2021; Guan et al., 2020) , which were verified by the official statistics that China' GDP grew by 2.3% in 2020, much lower than the projected growth (6%) without the pandemic (IMF, 2019) . While the shutdown of economic activities would significantly cut down the enterprises' production, the strict social distancing would cause consumers to decrease their expenditure (Guerrieri et al., 2020; McKibbin and Fernando, 2021) . McKibbin and Fernando (2021) employed a dynamic computable general equilibrium (CGE) model and found that China's GDP in 2020 would decline by 0.4-6.2%, depending on the population mortality and duration of the pandemic. Duan et al. (2020) used the input-output model, indicating that China's GDP would fall by 0.40-0.72%. Moreover, the COVID-19 pandemic would severely impact the output value of producing sectors, especially those exposed directly to the pandemic, such as tourism, hotel, restaurant, and retail (Guan et al., 2020; Duan et al., 2020) . The statistics show that domestic tourism revenue decreased by 3.5 trillion yuan in 2020 compared with that in 2019 (a sharp decline of 61.1%), and the revenues of catering and retail decreased by 16.6% and 2.3%, respectively (National Bureau of Statistics of China, 2020).

Globally, countries have adopted lax monetary policies widely to increase financial liquidity and alleviate financial market fluctuations for mitigating the pandemic's economic losses. For example, the Federal Reserve of the USA cut down the range of the target interest rate for federal funds by one percentage point and launched quantitative easing (QE) of 700 billion US$ (Federal Reserve of the USA, 2020). Similarly, the European Central Bank temporally enforced targeted longer-term refinancing operations (TLTRO) and an additional asset purchase program (APP) of 120 billion EUR (European Central Bank, 2020) . At the same time, proactive fiscal policies have been introduced around the globe. The federal government of the USA, for instance, put together the most ambitious stimulus package in American history, amounting to 2.2 trillion US$ (U.S. Department of the Treasury, 2020). European countries also widely adopted fiscal policies, such as the stimulus package of 30 billion GBP launched by the UK (HM Treasury, 2020) and infrastructure investment of over 122.5 billion EUR by Germany (Federal Ministry of Finance of Germany, 2020).

Since the pandemic outbreak, China's government has also initiated a series of pro-active fiscal policies and sound monetary policies to cope with the severe economic shocks. As for monetary policies, for example, the People's Bank of China launched a 1.2 trillion yuan reverse repurchase operation in the open market on February 3, 2020. The one-year loan prime rate (LPR) was reduced by ten basis points to 4.05%, and the LPR over five years by five basis points to 4.75% on February 20, 2020. The People's Bank of China also set up a targeted re-lending of 300 billion yuan for enterprises fighting against the COVID-19 pandemic, keeping the actual financing cost of these enterprises below 1.6%. A 500-billion-yuan re-lending and rediscount package was launched for agricultural and small enterprises, with a lower interest rate (2.5%). These sound monetary policies would release liquidity and reduce the funding cost in the capital market, which is expected to ease enterprises' fund gap and stimulate investors' confidence.

As for fiscal policies, several comprehensive countermeasures are implemented in China. For example, the value-added tax for small-scale taxpayers was exempted in Hubei province for the entire year of 2020, and the tax rate for small-scale taxpayers in other provinces was reduced from 3% to 1% for three months (from March 1, 2020 to May 31, 2020). The enterprises' income from transportation of goods and materials for epidemic prevention and control was exempted from value-added tax since January 1, 2020. The social insurance over five months was exempted for all companies in Hubei province and small, medium, and micro-enterprises in other provinces. All vehicles were exempted from road tolls nationwide since February 17, 2020. Furthermore, the central and local governments increased the fiscal expenditure for epidemic prevention and control by 110.48 billion by early March 2020. The pro-active fiscal policies aim to provide funds for epidemic prevention, minimize enterprises' production costs, and expand governmental expenditure, which is expected to stabilize employment and economic growth.

While several existing studies have analyzed the economic damages caused by public-health emergencies (Avery et al., 2020; Chang and Velasco, 2020; Coyle, 2020; Keogh-Brown et al., 2020; Stock, 2020; Deriu et al., 2021; Duan et al., 2021) , the effectiveness and trade-offs of various countermeasures to deal with the economic fallout are still poorly understood. The countermeasures were mostly neglected in early studies that assessed the direct impacts of public-health emergencies on the economy from both the supply and demand-side (Keogh-Brown and Smith, 2008; Verikios et al., 2011; Keogh-Brown, 2014 ). An increasing number of studies assessed the effectiveness of fiscal and monetary countermeasures in buffering economic damages caused by natural disasters and financial crises (Beyrer et al., 2006; Keen and Pakko, 2007; Xie et al., 2013; Flessa and Marx, 2016; Guerrieri et al., 2020) . For example, Porsse et al. (2020) projected the economic impacts of the COVID-19 outbreak and fiscal countermeasures on the Brazilian economy. They found that the government fiscal stimulus partially mitigates GDP losses from 3.78% to 0.48% and 10.90% to 7.64% in projections under the COVID-19 outbreak with different severity. They mainly evaluated the role of a single policy, employing econometric or input-output approaches (Hallegatte, 2008; Sangsubhan and Basri, 2012) , but ignored underlying side effects that might over-estimate the role of countermeasures in mitigating economic damages (Haldane, 2020; Bigio et al., 2020) . Although several studies noted that fiscal and monetary policies produce inflation risk, they rarely assess policy benefits against the costs (Kollmann et al., 2013; Fan et al., 2015; Huang and Hosoe, 2015; Gadatsch et al., 2016; Kunimitsu, 2018) . Moreover, the linkage between countermeasures and the sources of economic damages related to public health emergencies was seldom considered (Meng et al., 2010; Liu et al., 2015) . One can sum up by saying that despite the substantial contributions to assessing the economic impacts of public health emergencies by previous studies, the effectiveness and trade-offs of various countermeasures are still inadequately understood, which are especially pertinent for the current pandemic-related economic crisis.

Using a multi-sectoral computable general equilibrium (CGE) model, this study analyzes the effectiveness and trade-off of China's monetary and fiscal countermeasures to economic damages caused by the COVID-19 pandemic. Thus, we establish two illustrative scenarios, specifying the shocks related to the COVID-19 pandemic and monetary and fiscal countermeasures. We contribute to the existing literature in the following perspectives. Firstly, a framework of general equilibrium analysis is introduced to assess both the direct and indirect impacts of countermeasures on various economic indicators, such as GDP, employment, export, and consumer price index (CPI). The general equilibrium analysis is capable of revealing the unintended side effects of countermeasures. Secondly, these monetary and fiscal countermeasures are assessed for synergistic effects by coupling them to the supply and demand-side shocks of the pandemic. Appropriate countermeasures are chosen based on the source of pandemic damages because the source of economic damages differs across countries. Finally, we assess the effectiveness of countermeasures by comparing their costs and benefits with the arc elasticities of economic indicators. Our research provides valuable references to countries worldwide to effectively take targeted countermeasures with fewer negative externalities to cope with public health emergencies.

The remainder of this study is organized into five sections. Section 2 introduces the scenario settings of COVID-19 pandemic shocks and countermeasures in China. Section 3 describes the methodology. The results for the COVID-19 pandemic's economic effects and the effectiveness and trade-offs of countermeasures are discussed in Section 4. Section 5 concludes this study with discussions. The last section recommends several policy implications on the monetary and fiscal countermeasures.

Based on two illustrative scenarios, we employ the CHINAGEM model, a comparative-static, multi-sectoral, single-country CGE model developed based on the ORANI model (Horridge, 2014) . As the CGE model can capture the direct and indirect effects of exogenous changes in the economy and identify the impact mechanisms across the economy, it provides a valuable tool for a variety of policy-oriented studies related to macroeconomic, trade, and environmental policies (Cui et al., 2020) . The model assumes that the market is fully competitive, and the returns to scale of production remain unchanged. The model data is created based on China's input-output table with the base year of 2017 (National Bureau of Statistics China, 2019), aggregating 149 original sectors to 42 sectors. The model contains three types of primary factors (land, capital, and labor), six types of economic agents (production, investment, households, export, government, and inventory), and three types of margin commodities (No. 28, 29, and 32 in Table A1 ). The parameter and elasticity values are adopted from the corresponding values for China in the GTAP V9 database (Aguiar et al., 2016) . The inventory is fixed, and the total government expenditure is endogenous, determined by the government expenditure on different commodities and services. Six major modules are introduced below.

Producers are assumed to determine the production inputs according to production cost minimization and the output distribution to domestic or foreign markets according to profit maximization. The input of each producing sector is a combination of intermediate goods and primary factors. Like most CGE models, the production structure is described by the nesting constant elasticity of substitution (CES) functions aiming to incorporate the substitution possibilities across all inputs. The composite of intermediate inputs and primary factors is on the top level depicted by the Leontief function (Eq. 1). On the second level, the primary factors are composited by labor, capital, and the land described with the CES function (Eq. 6). The Armington assumption allows an incomplete substitution between domestically-produced (dom) and imported intermediate input ( 

1_ ( , ) = 1 * 1 0 ( , ) + 1 * , 1 0 ( , ) + ( ) -

Where 1 ( ) is the output value of sector . 1_ ( , ) refers to the demand of sector for intermediate input , which is composited of domestic and imported ones.

( ) denotes the demand for primary factors, which is a composite of capital, labor, and land.

( ) is the demand for other inputs. 1( ) refers to the parameter on total factor productivity, and 1 s are the parameters on input-augmented technology change. 1_ ( , ) denotes the price of the intermediate input c purchase by sector , that equals to the basic price excluding taxes 1_ ( , ), plus the indirect tax 1_ ( , ). 1_ ( , ) is the average of domestic price 1 0 ( , ) and import price ( 1 0 ( , ) + ( )), weighted by the proportions of the values of domestic and imported commodity c ( 1 and 1 ). The latter includes the import tariff ( ( )). 1 ( ) denotes the labor price for sector , equals to the basic price excluding taxes 1 ( ) plus the income tax 1 ( ).

Investors determine the quantity of capital goods according to cost minimization. Moreover, the total investment of each sector is determined by the rate of return (Eq. 7).

2_ ( , ) = * , , : 2( , , ) 2( , , ) + , = * , +

Where 2 ( ) refers to the total investment of sector . 2_ ( , ) is the demand for investment good c sector , which is composited of domestic and imported ones (Eq. 8).

The current return rate of investment for each sector ( 0 ( )) is determined by the rate of depreciation ( ( )), the rent of the capital ( ( )) and the cost of the capital ( ( )) in Eq. 9.

The expected return rate of investment for each sector ( 1 ( )) is determined by the current return rate of investment ( 0 ( )) and the ratio of future capital stock ( 1 ( )) to current capital stock ( 0 ( )) in Eq. 10. The _ is the shift variable to the expected return rate of investment for all sectors, and ( ) are the shift variables to the expected return rate of investment for each sector.

The future capital stock ( 1 ( )) is the summation of current capital stock ( 0 ( )) and the investment ( ( )) minus the depreciation.

Household consumption is determined by the disposable income and commodity price according to utility maximization. The linear expenditure system (LES) is used to describe household consumption (Eq. 13), derived from the first-order optimization of the Klein-Rubin utility function (Eq. 12). Similar to intermediate inputs and capital goods, the Armington CES function is employed to composite the consumption commodities with different sources (Eq. 14). 

3_ ( ) = 3 * 3 0 ( ) + 3 * , 3 0 ( ) + ( ) -

Where represents household utility. is the deposable household income. is the population. 3_ ( ) is household consumption for the good , which is a composite of the domestic and imported ones. 3 ( ) is the subsistence demand for goods . 3_ ( ) denotes the price of consumption good . ( ) is the marginal propensity to consumption goods . 3_ ( ) plus the consumption tax 3_ ( ). ( ) is the marginal propensity to consumption goods . 3 ( ) is the parameter for the subsistence demand. 3( , ) represents household demand for the goods from the source . 3( , ) is the parameter for consumer's tastes.

The tradable and non-tradable goods are distinguished. For the non-tradable goods, such as electricity, China's supply accounts for a much small share in the global market, or its export quantity and price are mainly determined by inter-governmental agreements. The export demand for non-tradable goods is irrelevant to its price. In contrast, the curve of export demand for tradable goods is downward sloping as a negative function of its price (Eq. 17).

Where 4( ) is export demand of goods c. 4 0 ( ) is the FOB price in local currency excluding taxes, and the 4( ) is the export tax. denotes the nominal exchange rate. 4 ( ) and 4 ( ) are shift variables to the position of the export demand curve.

_ ( ) is the export elasticity of price.

In the general equilibrium, the commodities and primary factors markets are cleared. The condition of zero profit is held that the sale of producers is equal to the production cost, and consumers' expenditure is equal to the summation of purchase cost, tax, and margin cost. Besides, the balances of investment and savings, governmental revenue and expenditure, and payments in international markets are balanced. The balance of investment and savings is achieved by adjusting foreign savings.

This study focuses on the impact of the COVID-19 pandemic on China's economy in the short run and thus employs a short-run closure for the simulations. The capital is fixed in the producing sectors with the sectoral-differentiate rate of return. The real wage is held unchanged while allowing for unemployment. Tax rates and technology are exogenously determined.

The effect of each pandemic shock and countermeasures could be calculated with a decomposition analysis approach. The method, developed by Harrison et al. (2000) , can be used to decompose the simulated results for the endogenous variables, such as GDP, employment, and sectoral output, to the contributions of exogenous shocks. The impacts of the COVID-19 pandemic and countermeasures are attributed to five types of pandemic shocks and ten types of countermeasures (introduced in Section 3). Here, we employ the decomposition analysis approach to assess the impacts of each pandemic shock and countermeasures on economic variables.

The supply-side shocks of the COVID-19 pandemic include the extension of spring festival vacation by 3-16 days (Vacation extension) and insufficient operation of enterprises after the holiday (Insufficient operation). Table 1 shows the shocked variables for the shocks of the COVID-19 pandemic.

(a) Vacation extension. While the nationwide vacation extension hindered the spread of the pandemic, it also reduced sectoral outputs by a significant amount. Following Dixon et al. (2010) and Verikios et al. (2010) , we gauge vacation extension by the rate of the workday loss to yearly workdays and introduce it as the shock to total factor productivity of producing sectors (shocking A1 in Eq. 1). The workday loss in China is calculated by the average workday loss for each province, weighted by the provincial GDP. Hence the nationwide workday loss of 6.9 days accounts for 2.75% of the yearly workdays (251 days). Considering the uneven time distribution of economic activity, the share of GDP in the first quarter (21%) is used to transform the workday loss rate to 2.31% (2.31%=2.75*21%/25%). Moreover, the industries that are important to the national economy and the people's livelihood, including AGR, FOD, CMC, ELE, GAS, WTS, and HPS sectors (the sectoral abbreviations are described in Table A1 ), are excluded.

(b) Insufficient operation. Even after the vacation, the sectoral production was still below the normal level due to labor deficiency caused by home isolation, labor flow lockdown, and peak-shifting return to work. Insufficient operation is gauged by the rate of enterprises returning to work, as the shock to labor productivity of producing sectors (shocking A1LAB in Eq. 2). Based on the official statistics, we assume that the rates of large-scale enterprises returning to work were 70% from February 10 to 

The demand-side shocks of the pandemic refer to the economic damages caused by the reduction in household consumption (Consumption reduction) and total investment (Reduction of investment), as well as the export changes (Export change).

(a) Consumption reduction. Accompanying the decrease in total consumption expenditure, households' consumption structure is also affected by the pandemic, cutting down their consumption for FOD, CSL, TRD, TWP, HTD, HPS, and CSR sectors. According to the National Bureau of Statistics of China (2020), households' total expenditure decreased by 3.7%, and the change in expenditure varied largely between sectors. The consumption for CSR (-26.2%), HTD (-26.2%), TRD (-17.1%), CSL (-11.9%), TWP (-7.0%) and HPS (-5.8%) were all reduced, while that of FOD (4.9%) was increased (shocking A3 in Eq. 14).

(b) Reduction of investment. As the pandemic raised the investment risks, the investors require the higher risk premium added to the rate of return, which raises the price of capital in China. We use the investment shock from Mckibbin and Fernando (2021) and Cui et al. (2021) , reducing the risk premium of investment by 1.97% for the entire year (shocking FR_I in Eq. 10).

(c) Export change. On the one hand, China's export would be negatively affected due to limited production and transportation of goods and services during the pandemic months. On the other hand, as the pandemic more severely hit other countries, China's exports would rise, accompanied by increasing competitiveness in the global market. According to the National Bureau of Statistics of China (2020), the export of AGR (-13.8%), CMP (-31.8%), COG (-31.8%), MTM (-31.8%), NTM (-31.8%), FOD (-2.2%), SMF (-2.2%), PRC (-31.8%) were decreased, while those of TEX (4.6%), CSL (4.6%), PPC (4.6%), CMC (4.6%), NMP (6.8%), MTS (6.8%), MTP (6.8%), GEQ (4.6%), SEQ (4.6%), TEQ (4.6%), ETM (4.6%), CMC (5.3%), MOE (5.3%), OMF (5.3%) were increased (shocking F4Q in Eq. 17). Simultaneously, following Duan et al. (2021) , we assume that the pandemic reduces global commodity prices by 2% (shocking F4P in Eq. 17), shifting the export demand curve to the left in the single-country CGE model.

This study incorporates two types of monetary countermeasures and eight types of fiscal countermeasures adopted in China by the end of March 2020 (Table 2) . Although China's government rolled out some other policies during the pandemic's peak quarter, they cannot be quantified into specific types.

The People's Bank of China (PBC) released the liquidity through the open-market operation to reduce the LPR by ten basis points. Affected by the policy, the LRP fell from 4.15% to 4.05% on February 20, 2020. Hence, the open-market operation cut down the producing sectors' financing cost by 2.4% (shocking FR_I in Eq. 10), calculated by the following equation (Eq. 18).

Where PFC t is the percent change of producing sectors' financing cost in the period t. LPR t and LPR t-1 represent the LPR in the period t and t-1, respectively. 

The PBC announced a specially utilized re-loan (SUR) of 300 billion RMB, lending to the enterprises involved in the pandemic prevention, and a specially utilized re-loan of 1,500 billion RMB, lending to the agricultural and small enterprises as well as the financial institution, for the entire year. We calculated the decreases in financing cost caused by the two re-loans for the covered producing sectors. The SUR can cut down the financing cost of the supported producing sectors by 0.11-0.97% with an average of 0.24% (shocking FR in Eq. 10).

On one hand, the specially utilized re-loan of 300 billion RMB would lower down the sectoral financing costs by 0.44% for FOD, 0.46% for CMC, 0.28% for SEQ, 0.70% for TRD, 0.56% for TWP, 0.60% for CTS, and 0.57% for HPS, as calculated below.

The interest rate of this specially utilized re-loan is 1.28%, 74.4% lower than the interest rate of the commercial loan (5%). However, the specially utilized re-loan only accounts for 1.2% of the commercial loan of specific supported sectors in the 149-sector IO Where R s and R c are the interest rates of the specially utilized re-loan and commercial loan, respectively. SUR represents the value of specially utilized re-loan, and TCL j represents the value of total commercial loan of specific supported sector j in the 149-sector IO table. PGV j is the gross value of sector j, and PGV i is the gross value of aggregated sector i in our model.

On the other hand, the specially utilized re-loan of 1,500 billion RMB would lower the financing costs of supported sectors by 0.12-0.30% with an average of 0.15%. The interest rate of this SUR is 4.55%, 27.8% lower than the interest rate of the small-scale loan (6.3%). On the one hand, the re-loan of 375 billion RMB is allocated to AGR, accounting for 1.1% of the loan to agriculture enterprises (375/35,190=1.1%).

The changes in the financing cost of AGR are estimated by multiplying the decrease in interest rate with a share of SUR in AGR's total loan. On the other hand, the rest 1,125 billion RMB is allocated to small enterprises of producing sectors, which accounts for 3.0% of the nationwide loan to small enterprises (1125/36,900=3.0%).

The changes in the financing cost of producing sectors are estimated by multiplying the decrease in interest rate with a share of SUR in each sector's total loan in 2010, which is updated with the gross value of small enterprises in 2013.

The Ministry of Transport relieved the value-added tax for the transportation of prevention materials, such as food, pharmaceutical products, and clinical equipment, for the entire year, which reduced the indirect tax rate of sectors' utilization of FOD, CMC, and SEQ. We assume that the policy will continue for the whole year, making the rate of value-added tax for transportation to fall from 11% to 0%. The changes in the indirect rate of different prevention materials are calculated by multiplying the decrease in the tax rate with the share of value-added tax for transportation of different prevention materials in each sector's indirect tax. Hence the indirect tax rate of sectors' utilization of FOD, CMC, and SEQ will decrease by 0.12%-11.82%, 0.09%-12.74%, 0.04%-6.47% for different producing sectors (shocking T1_S in Eq. 4).

The Ministry of Transport relieved the value-added tax for private transportation and express postal services, which reduced the tax rate from 11% to 0%. We also assume that the policy will continue for the entire year. Therefore, the indirect tax change rate of residents using transportation, storage, and express postal services is equal to the indirect tax change rate of residents using public transportation and express postal services (100%) multiplied by the proportion of residents' consumption of these two sub-sectors (shocking T3_S in Eq. 15).

For Hubei province, the value-added tax was relieved for the small-scale taxpayer from March 1 to December 31 in 2020. For other provinces, the value-added tax rate for the small-scale taxpayers was reduced from 3% to 1% (shocking T1_S in Eq. 4). Weighted by the shares of the tax revenue of the small-scale taxpayers in different regions, the change in the national value-added tax rate is calculated by Eq. 20.

Where vat is the change in the national value-added tax rate. vat H and vat NH represent the change in the value-added tax rate of Hubei province and other provinces, respectively. VAT H and VAT NH represent the revenue of value-added tax in Hubei province and other provinces, respectively. The change in the value-added tax rate is transformed into the entire year by using the share of the policy-covered months (10/12). Then we obtain the sector-specific changes in indirect tax by multiplying the change in value-added tax rate with the share of each sector's value-added tax in its indirect tax.

The 5% of electricity fees were relieved by the National Development and Reform Commission for all manufacturing and services enterprises except for high-energy consumption sectors from February 1, 2020, to December 31, 2020. The electricity consumption from February to December of 2019 accounted for 91.54% of annual electricity consumption. Hence the electricity price of producing sectors declined by 4.58% (5%*91.54%= 4.58%), except for high-energy consumption ones (shocking T1_S in Eq. 4).

(e) Expanding expenditure on clinical and residential materials (GCM) By the end of March, the government expanded the expenditure by 400 billion RMB in purchasing clinical and residential materials, including FOD, CMC, SEQ, WTS, RSH, and HPS sectors. We can use two methods to calculate the changes in government expenditure on clinical and residential materials. In the first method, the ratio of incremental spending to the fiscal budget on these sectors in 2019 is 8.49% (400/4,709.6=8.49%), but it may underestimate the expenditure changes. The other is the ratio of the incremental expenditure to the fiscal spending on these sectors estimated by the share of the expenditure on these sectors in total expenditure with total fiscal budget (400/(22,090.4*9.6%)= 18.86%). However, it may overestimate expenditure changes. Hence the mean value of the two ratios (13.68%) is used as the percentage change in government expenditure on clinical and residential materials.

The social insurance expenses of 660 billion RMB were exempted by the Ministry of Human Resources and Social Security for small enterprises. It is expected to exempt 510 billion RMB of endowment insurance, unemployment insurance, work injury insurance, and 150 billion RMB of medical insurance. The total labor cost in 2020 is estimated to be 50,697 billion RMB, which is obtained by updating the total labor cost in 2017 by the annual GDP growth rate from 2017 to 2019. Affected by ESI, the labor cost of producing sectors decreases by 1.30% (660/50,697=1.30%) (shocking T1LAB in Eq. 6).

The import tariff was exempted for the prevention materials from January 1 to March 31 in 2020, for medical reagent, disinfectant, protective respirator, ambulance, epidemic prevention vehicle, disinfection vehicle, and an emergency command vehicle. By mapping the specific prevention materials with sectors of our model, the percentage changes in tariff revenue are multiplied by a 100 percent decrease in import tariffs with the share of each prevention in the import of the mapped sector based on the data of the UN Comtrade (2020). Finally, the percentage change in tariff rate, as the shock to the CGE model, is obtained by multiplying the percentage changes in tariff revenue with the tariff rate in the database. The import tariff rates of CMC, CSL, SEQ, and TEQ sectors declined by 2.10, 0.27, 0.05 and 0.01 percent points, respectively (shocking T m in Eq. 5 and Eq. 16).

The road tolls were exempted from all vehicles nationwide from February 3 to May 5 in 2020. The revenue of road tolls in 2017 was 513.02 billion RMB, accounting for 12.18% in road transportation expenditure of producing sectors and households (4.21 Trillion RMB). Moreover, road transportation expenditure accounts for 47.3% of China's total transportation expenditure in 2017. Hence the percent changes in transportation expenditure of household and producing sectors could be calculated by Eq. 21.

Here, SRT represents the percentage changes in transportation expenditure of household and producing sectors. RT is the revenue of the national roads toll in 2017. PRTE and IRTE represent the road transportation expenditure by household and producing sectors, respectively. TTE denotes the total transportation expenditure of China in 2017, and YPI represents the number of months covered by the ERT (shocking T1_S in Eq. 4, T3_S in Eq. 15, and T4 in Eq. 17).

The COVID-19 pandemic has severely damaged China's economy through both supply and demand-side shocks. Our results indicate that the national GDP would fall by 4.42%, accompanied by a 5.96% decrease in employment (or 46.17 million job losses) 1 for the entire year (Panel a, Figure 1 ). The pandemic not only negatively affects China's internal demand by hindering investment and household consumption but also interrupting import supply chains (Panel a, Figure 2 ). Compared with demand-side shocks, supply-side shocks would cause even more severe damages to China's GDP, employment, and investment. While demand-side shocks would impact GDP, household consumption, and employment, they would also lower the CPI and thus improve the competitiveness of exports.

Considering the projection of China's GDP in 2020 before the pandemic (6.0%) by IMF (2019), our results indicate that China's GDP is projected to increase by 3.1% in 2020, which is moderately higher than the official statistics (2.3%). The result indicates that our approach underestimates the negative impact of the pandemic on China's GDP. We attribute it to the fact that this study does not consider the trade and travel restrictions imposed by other countries and the changes in behavior patterns of China's citizens. 

The COVID-19 pandemic has far-reaching and heterogeneous impacts on the output value of producing sectors (Panel b, Figure 1 ), which are mainly driven by supply-side shocks. The output of producing sectors would decrease by an average of 4.58%. Several industries are not directly shocked by the pandemic, yet their production is affected indirectly through upstream and downstream production chain effects, thus leading to an overall macroeconomic recession. The simulation results suggest that the output value of manufacturing sectors would decrease by an average of 3.92% compared to a decline of 5.93% in service sectors. Services are not only affected by supply-side shocks, which lead to reductions in productivity and rising production costs but also negatively affected by lower household consumption caused by demand-side shocks. Surprisingly, the manufacturing sectors are positively affected by demand-side shocks, as the reduction of labor employment in service sectors would lower the labor costs of manufacturing sectors, giving them stronger competitiveness in the global market. However, the supply-side shocks would negatively affect the output value of manufacturing sectors. Damages to most producing sectors are mainly derived from supply-side shocks caused by vacation 2 The COVID-19 represents the total effect of the pandemic, considering both demand and supply-side shocks. The other annotations represent the effect of each of demand and supply-side shocks, which are decomposed from the total effect. extension and insufficient operation, while some service sectors like hotel and catering industries are mainly affected by demand-side shocks caused by consumption reduction. 

Although monetary and fiscal policies can effectively buffer the damage of COVID-19 to China's economy, the trade-off of GDP/employment against CPI emerges in the effects of these countermeasures. Stimulating economic growth and employment potentially lead to an increase in prices, which is consistent with the relationship shown by the Phillips Curve. The 4.42% GDP contraction caused by COVID-19 without countermeasures would be reduced to 2.91% after adopting the stimulus policies (Panel a, Figure 3 ), accompanying with a greater net employment loss (3.51%). Compared with monetary policies, fiscal policies would bring larger increases in GDP and employment. However, at the same time, affected by the countermeasures, the CPI would increase by 1.63%, around 64 percent of which is attributed to monetary policies. Hence, these countermeasures would also raise the prices of production factors and consequently lift the CPI. It is worth noting that compared with fiscal policies, the monetary policies would raise the CPI by a larger amount, even if both of them increase GDP by the same quantity. The estimated arc elasticity of CPI to GDP for monetary policy of ~6 indicates that if the monetary policy raises GDP by 1 percent, the CPI will rise by over 6 percent. In contrast, the arc elasticity of CPI to GDP for fiscal policy is only 0.42. Meanwhile, the arc elasticity of CPI to employment for monetary policy is estimated to be 5.29, which is also several times larger than that for fiscal policy.

Monetary and fiscal countermeasures could buffer various aspects of economic damages caused by the economic shocks triggered by the pandemic. The monetary policies could mitigate the damage to national GDP and employment. However, they will further exacerbate inflation and reduce exports. Comparably, the fiscal policies could buffer reductions in GDP and employment but stabilize CPI and benefit the recovery of exports. As discussed in section 4.1, the supply-side shocks would cause damages to GDP and employment and lead to a sharp rise in CPI, while demand-side shocks will lower CPI. Hence, it is better to use fiscal policies to mitigate the damages caused by supply-side shocks, as the monetary policies would exacerbate the inflation and deteriorate the exports further. Because the demand-side shocks could lower the CPI, creating ample space for utilizing the monetary policies, it is better to use monetary policies to mitigate the damages of the demand-side shocks. They could raise national GDP and employment by a larger amount at the expense of a slight increase in CPI. Since countermeasures' role is much different in buffering various aspects of economic damages caused by supply and demand-side shocks, the sources of the pandemic-induced recession should be identified before making countermeasures.

Although monetary and fiscal policies raise the CPI accompanying economic growth, several specific policies, including relieving value-added tax for private transportation and express service (VRT), exempting import tariff for prevention materials (ETM), and exempting social insurance expenses of small enterprises (ESI), can keep price stable while increasing GDP and employment. Among the monetary countermeasures, the open market operation (OMO) will induce the largest increase in GDP (0.15%) by directly lowering the financing costs of downstream enterprises. However, it also raises CPI by the largest amount (0.90%) due to a relatively high arc elasticity of CPI to GDP. Although the specially utilized re-loan (SUR) seems to increase CPI slightly, its arc elasticity of CPI to GDP is much higher, close to that of the OMO. While exempting road tolls (ERT) and relieving value-added tax for transportation of prevention materials (VTM) can increase employment by 0.14% and 0.36% by reducing transportation costs of producing sectors, they slightly raise CPI by 0.06% and 0.09%. Hence, their arc elasticities of CPI to GDP are relatively small.

Interestingly, the VRT, ETM, and ESI would cut down enterprises' production costs, which reduce the GDP damages and lower down the CPI. They could buffer the increase in prices caused by other monetary and fiscal policies. As showed in Panel C of Figure 3 , the arc elasticity of CPI to GDP of the ETM is estimated to be -2.27, indicating that if the policy raises GDP by 1%, it will cut down CPI by over 2%.

Although the elasticity of CPI to the GDP of the ESI is much small, it could largely buffer the GDP damages by 0.66%. Hence, several specific policies can alleviate the increases in price caused by radical monetary and fiscal countermeasures.

The trade-off between internal and external demand emerges in the effects of the countermeasures to the social and economic restrictions associated with curbing the pandemic. The countermeasures can effectively raise investment and consumption by 4.49% and 1.63%, respectively; they can also stabilize China's internal demand. However, external demand is negatively affected by countermeasures. The CPI would rise by 1.62%, reducing the competitiveness of China's commodities in the global market, which consequently cuts down exports by 3.66%. The trade-off between internal and external demand is vital for choosing the right mix of countermeasures to the pandemic.

While the monetary policies effectively improve investment and foster consumption, they also have severe negative impacts on commodity exports. The OMO, lowering the financing costs of enterprises, raises investments by 1.57%, and the SUR lifts the investment by 0.27% by narrowing the funding shortfall of small enterprises. The monetary policies have positive but smaller impacts on consumption (0.56%) as the up-surging CPI may hinder an increase in consumption. However, the risk of inflation would also emerge in monetary policies in that the OMO would raise the CPI by 0.90%, and the SUR would raise the CPI by 0.15%. As a result, monetary countermeasures would reduce export by 2.61% as a side effect.

While the fiscal policies have smaller positive effects on investment than monetary policies, they would also increase the CPI by a smaller amount. Fiscal policies together would increase investments by 2.65%, over a third of which is attributed to the ESI that reduces social insurance expenses of producing sectors, consequently stimulating production and investment but with potentially adverse effects on demand. The bundle of fiscal policies would increase consumption by 1.32%, around half of which is caused by ESI. Compared with monetary policies, fiscal policies have somewhat smaller negative impacts on external demand, reducing exports by 1.05%, as they would raise the CPI only slightly (0.57%). Moreover, ETM and ESI policies would benefit China's exports slightly, mitigating the export reduction caused by the pandemic and associated lockdown. 

Although the average output value loss of producing sectors is reduced to 3.19%, the inter-sectoral variation in damage to sectoral output value is further amplified by the countermeasures, especially monetary countermeasures (Panel a, Figure 4 ). Our results show that all the producing sectors would benefit from fiscal countermeasures with an average increase in the output value of 1.30%. The impacts of monetary policies are much more uneven among the sectors, 52% of which are negatively affected with an average decline in production of 0.35%, and the remaining sectors show a 0.38% average increase in output value. In summary, while countermeasures can potentially improve the performance of most sectors, monetary countermeasures may further contribute to the decline in output value caused by the COVID-19 pandemic for some sectors, mainly exported-oriented manufacturing sectors, by raising the prices of their production inputs and worsening exports.

The trade-off among producing sectors emerges in the effects of countermeasures, i.e., the countermeasures would not benefit all producing sectors (Panel b, Figure 4) . (1) The positively affected sectors by both types of countermeasures include capital goods and their upstream industries and services. The countermeasures effectively stimulate total investment, consequently reducing the output value damages of capital goods and their upstream sectors caused by the pandemic. The fiscal countermeasures benefit the output value of productucing industries along the supply chain of the target sectors, which surpasses the slight negative impacts resulting from the rising prices of primary factors, which in turn lead to a much larger net increase in the output value of production.

(2) Production of several sectors benefits from fiscal countermeasures but is negatively affected by monetary countermeasures. As for exported-oriented manufacturing sectors, the negative impact of monetary countermeasures would exceed the positive effects of fiscal countermeasures, resulting in a net decrease in their output value. These sectors include textiles and wood products (TEX), electrical machinery and equipment (CMC), and general and special equipment (ERC). In contrast, for the remaining sectors, including agriculture, coal mining products, food processing, and services, the positive impacts of fiscal countermeasures surpass the adverse effects by monetary countermeasures Table A1 .

China has enforced a series of pro-active fiscal policies and sound monetary policies to mitigate the severe economic damages caused by the COVID-19 pandemic. However, the effectiveness of different countermeasures to economic crisis from the public health emergency is still inadequately understood in previous studies. To our knowledge, the trade-offs of countermeasures in mitigating economic damages are seldom analyzed. We establish an illustrative scenario, specifying five shocks of the COVID-19 pandemic to China's economy from the demand and supply-side, as well as two types of monetary policies and eight types of fiscal policies adopted by the government. Then a general equilibrium model of China is applied to analyze the effectiveness of countermeasures with a particular focus on trade-offs in the impacts of monetary and fiscal policies.

We find that (1) the national GDP to fall by 4.42%, accompanied by a 5.96% decrease in employment for the entire year of 2020. Compared with demand-side shocks, supply-side shocks would cause even more severe damages to China's macro-economy. The COVID-19 pandemic has far-reaching, heterogeneous impacts on the output value of producing sectors, which are mainly driven by supply-side shocks. Except for increases in the chemical industry, the output value of producing sectors would decrease by an average of 4.58%.

(2) Both monetary and fiscal countermeasures could effectively mitigate the economic damages to GDP and employment. The GDP contraction caused by COVID-19 would be reduced to 2.91% after adopting the stimulus policies. Compared with monetary policies, fiscal policies would bring larger reductions in GDP and employment losses.

(3) The monetary and fiscal countermeasures would also produce negative side-effects such as an increase in inflation by 1.05% and 0.57%, respectively, and a decline in exports by 2.61% and 1.05%, respectively. We also found that monetary policies would exacerbate the damages to external demand by supply-side shocks of the pandemic, but they are suitable for mitigating demand-side shocks. Whereas fiscal policies would benefit nearly all producing sectors, monetary policies would mainly affect export-oriented manufacturing sectors negatively.

We also perform a sensitivity analysis to check the robustness of our simulation results to different input parameters (Appendix B). We vary each parameter by ± 50% in turn, observing the effect of countermeasures to the economic crisis caused by the COVID-19 pandemic. The robustness of simulation results is confirmed by the small standard deviation resulting from the sensitivity analysis.

Our study also has some limitations that should be focused on in future research. First, our model requires yearly data of China covering the whole country. However, the policies are differentially enforced in different regions of China. For example, to quantify the VST, we re-scale the tax relief by scaling up the provincial shocks to the whole economy based on the regional trends. However, the regional trends in 2017 are somewhat different from current ones, which may cause estimation inaccuracy. Moreover, our method does not cover the over-shooting of economic activity after the pandemic. The share of small taxpayers is based on the input-output table of the year 2010, which is another source of estimation discrepancy. Second, the spatial and sectoral coverage is not clearly defined for several mitigation policies in the official documents. For example, although the GCM expands the governmental expenditure on food, pharmaceutical products, water supply, research, and clinical services, the official documents do not specify how incremental spending is allocated to these products. The third limitation is that this study did not consider the informal work for calculating the labor supply shocks due to the unavailability of reliable data on the informal work. In our opinion, the COVID-19 pandemic will reduce the formal work but increase the informal work, so our estimation may over-estimate the reduction of the labor supply shock.

Given the scale of economic losses, the sources of the "pandemic crisis" should be identified for designing effective countermeasures. Although the countermeasures could stabilize the country's internal demand, the measures will deteriorate the external demand by damaging exports. As the fiscal policies have relatively even effects on manufacturing sectors, the fiscal policies adversely affect the output value of export-oriented manufacturing sectors. Hence, the enforced countermeasures are not free from the side effects. Therefore, due diligence should be carried out while implementing the right mix of countermeasures to the pandemic-induced economic recession.

Due to the multiple trade-off relationships, appropriate countermeasures should balance the costs and benefits of countermeasures under the background of the policy target and the economic condition of each country. Considering trade-offs between GDP, employment, and CPI, monetary policies are recommended to a country dealing with deflation rather than one with stagflation, while the fiscal policies are suitable for countries with high inflation. Due to the trade-offs between internal and external demand, fiscal policies are more suitable for countries that depend highly on international markets, as monetary policies would deteriorate exports further. Due to the trade-offs among producing sectors, careful consideration should be given to other economic, social, and sectoral targets and impacts on stakeholders and society while choosing the right mix of countermeasures.

This paper also provides important policy implications to other countries. Facing the economic losses caused by the pandemic, most countries have enforced fiscal policies by giving a tremendous amount of emergency aid and fiscal stimulus to individuals and corporations and monetary policies by lowering the interest rates. We believe that the countries should implement fiscal and monetary policies depending upon their economic conditions. For example, Japan should use monetary policies more than fiscal policies, as the monetary policies could buffer the impacts of the COVID-19 pandemic, simultaneously addressing the problem of the continued deflation.

A. The sectors of the CHINAGEM model We concentrate sensitivity analysis on the behavior parameters in the CHINAGEM model and vary only six groups of elasticities, including the substitution elasticities between domestically-produced and imported commodities used by producing sectors, investors, and households respectively, the substitution elasticities between primary factors used by producing sectors, the transition elasticities of various output produced by producing sectors, and the price elasticities of export demand for tradable commodities.

The method requires information on the distributions of those parameters for estimating the standard deviation of simulation results. We assume independent, symmetric, triangular distributions with the mean/mode at the value of parameters of the CHINAGEM model, solving by Stroud's Gaussian quadrature (Arndt, 1996) . Each parameter varies by ± 50% in turn. We then draw a discrete distribution, sets of points, and associated weights, whose moments match the moments of the assumed true multivariate distribution up to the order of three. The standard deviations obtained will be very accurate (Preckel and DeVuyst, 1992; Preckel and Liu, 1994) .

The mean and standard deviation of macro-economic variables and sectoral output values are shown in Table A2 and Table A3 , respectively. The standard deviations of macro-economic variables are much small, less than 0.5, which indicates the stimulation results for macro-economic variables are not significantly affected by the value of the parameters of the CHINAGEM model. Although the standard deviation of changes in output values are much different for producing sectors, they are also much small, confirming the robustness of our simulation results. b. Ensuring the production and supply of fertilizers: The coal supply should be prioritized to coal-based fertilizer production enterprises. Besides, the natural gas supply should be ensured for gas-based fertilizer production companies after meeting the residential demands (in spring plowing). 2 c. Keeping foreign investments stable: The foreign enterprises should be assisted in resuming work and production in the epidemic period. All support policies for the domestic corporate sector also apply to foreign-owned companies. 3 d. Ensuring agricultural firms can resume production: The policies such as re-loan, interest subsidy, tax reductions, or exemption should be implemented for agrarian firms. Companies that cannot operate normally due to lack of labor should be assisted in implementing the plans of labor replacement (In spring plowing). 4

a. Ensuring employment of college graduates: One-time employment subsidies should be given to medium-sized, small, and micro firms for recruiting college graduates. State-owned enterprises will expand the scale of hiring graduates for two consecutive years. Available jobs in Hubei public institutions can be specially provided for graduates born in Hubei or those graduating from universities in Hubei (2020-2021). 5

b. Providing living assistance to the unemployed group: The application period will be extended to people who meet the condition of receiving unemployment insurance benefits. Six months' unemployment subsidies will also be paid for persons who do not meet the requirement of receiving unemployment insurance benefits and have not been employed since the expiration date of the pension. 5 c. Preventing workers from employment discrimination: The employers shall not release recruitment information that refuses to recruit workers in affected areas during the epidemic period. They shall not abolish the labor contract for those who cannot return to work temporarily because of the epidemic. 5 d. Expanding rural labor employment: The temporary jobs such as janitors should be prioritized for the poor household laborers, and subsidies should be given according to the standard of local rural public welfare subsidy during the epidemic period. Supportive policies like one-time entrepreneurship subsidies and guarantee loans should be provided to workers who return to the countryside. 6 The official documents on these policies: 1. Ministry of Transport. Notice on ensuring transportation services for migrant workers returning to work, February 11, 2020. 2. National Development and Reform Commission. Notice on ensuring the production and supply of fertilizers in spring plowing, February 17, 2020. 3. National Development and Reform Commission. Notice on furthering reform to

foreign investments stable

Notice on ensuring agricultural firms resuming work and production in the epidemic period

Taking various measures to keep employment stable

Notice on expanding rural labor employment and promoting farmers' income to response to the impact of the epidemic

Writing-Original draft preparation

An overview of the GTAP 9 data base

Revisiting The Fallacy of Free Trade

An Introduction to Systematic Sensitivity Analysis via Gaussian Quadrature

Policy Implications of Models of the Spread of Coronavirus: Perspectives and Opportunities for Economists

Economic impacts of Wuhan 2019-nCoV on China and the world

Responding to AIDS, tuberculosis, malaria, and emerging infectious diseases in Burma: dilemmas of policy and practice

Transfers vs. Credit Policy: Macroeconomic Policy Trade-offs during Covid-19

Online

Economic Policy Incentives to Preserve Lives and Livelihoods

Economists must collaborate courageously

Economic and climate impacts of reducing China's renewable electricity curtailment: A comparison between CGE models with alternative nesting structures of electricity

The impacts of COVID-19 pandemic on China's transport sectors based on the CGE model coupled with a decomposition analysis approach

The economic impact of Covid-19 pandemic in Sardinia

Effects on the US of an H1N1 epidemic: analysis with a quarterly CGE model

The hit of the novel coronavirus outbreak to China's economy

Supplementary Information Coronavirus: limit short-term economic damage

Impact of Financial Crisis and Policy on China's Macro Economy: A Simulation Analysis Based on

Federal Reserve of the USA

Ebola fever epidemic 2014: a call for sustainable health and development policies

Fiscal policy during the crisis: A look on Germany and the Euro area with GEAR

Global supply-chain effects of COVID-19 control measures

Macroeconomic Implications of COVID-19: Can Negative Supply Shocks Cause Demand Shortages? NBER Working Paper

Numerical Evaluation of Multiple Integrals

To set coronavirus policy, model lives and livelihoods in lockstep

An adaptive regional input-output model and its application to the assessment of the economic cost of Katrina

Decomposing simulation results with respect to exogenous shocks

Online

ORANI-G: A Generic Single-Country Computable General Equilibrium Model

Investigating Fiscal and Social Costs of Recovery Policy: A Dynamic General Equilibrium Analysis of a Compound Disaster

World Economic Outlook Database

Monetary policy and natural disasters in a DSGE model: how should the Fed have responded to Hurricane Katrina? FRB of St

The impact of Covid-19, associated behaviours and policies on the UK economy: A computable general equilibrium model. SSM -Population Health

Macroeconomic Effect of Infectious Disease Outbreaks. Encyclopedia of Health Economics

The economic impact of SARS: how does the reality match the predictions? Health policy

Fiscal policy, banks and the financial crisis

Effects of restoration measures from the East Japan earthquake in the Iwate coastal area: application of a DSGE model

A financial CGE model analysis: Oil price shocks and monetary policy responses in China

If the world fails to protect the economy, COVID-19 will damage health not just now but also in the future

The global macroeconomic impacts of COVID-19: Seven scenarios

The impact of the 2008 world financial crisis on tourism and the Singapore economy and policy responses: a CGE analysis

The economic impacts of COVID in Brazil based on an interregional CGE approach

Efficient Handling of Probability Information for Decision Analysis under Risk

Efficient, Independently Weighted Multivariate Quadratures

Global Financial Crisis and ASEAN: Fiscal Policy Response in the Case of Thailand and Indonesia

Data Gaps and the Policy Response to the Novel Coronavirus

The global economic effects of pandemic influenza

H1N1 Influenza in Australia and Its Macroeconomic Effects

Modeling economic costs of disasters and recovery involving positive effects of reconstruction: an analysis using a dynamic CGE model