世界銀行_新冠疫情對GDP和貿易的潛在影響-初步評估.pdf

Policy Research Working Paper 9211 The Potential Impact of COVID-19 on GDP and T rade A Preliminary Assessment Maryla Maliszewska Aaditya Mattoo Dominique van der Mensbrugghe East Asia and the Pacific Region Office of the Chief Economist ii the increase in costs of international transactions; iii the sharp drop in travel; and iv the decline in demand for services that require proximity between people. We consider two scenarios a global pandemic and an amplified global pandemic. In the case of the global pandemic, it is assumed that countries bear only one-half of the impact of the full China shock. In the case of the amplified global pandemic, the shocks are uni across all countries. A baseline global pandemic scenario sees GDP of the world fall by 2 percent below the baseline, of developing countries by 2.5 percent, and of industrial countries by 1.8 percent. The declines are nearly twice as large in an amplified pandemic scenario in which containment is assumed to take longer. It is still too early to make an assessment of the impact of the virus based on full statistical evidence. High frequency data are providing some indicators, but it is hard to assess the depth and the breadth of the pandemic as it spreads, and to precisely estimate how long it will take countries to return to normal activity levels. This paper seeks to illustrate the transmission channels and heterogenous impact of COVID19 on output and trade in different scenarios. The results presented here should be regarded as scenario analyses, not as projections. The implemented shocks are illustrative and based on previous episodes of global epidemics or on preliminary data. The assumptions on the spread of the disease are not grounded in epidemiological projections, they do not take into consideration the quality of the health systems in the affected countries, transport connections to affected countries, and health policy responses to the outbreak. The model incorporates the decline in demand due to reduced production and incomes but does not fully capture the independent contraction in demand, except for the reductions in tourism and other services that require close human contact. It also does not include the decline in investor confidence and any financial repercussions. We capture some aspects of global value chains trade, but a fuller analysis will require a richer data set. This analysis will evolve as we fine tune assumptions in line with early impacts and uate potential scenarios of the spread of the virus. - 3 - 2. ology, transmission channels, and scenarios 2.1 Global computable general equilibrium model Envisage The quantitative findings in this paper are based on simulations using a version of the Envisage model calibrated to GTAP Version 10A Aguiar et al. 2019; see Annex 1 for aggregation mappings. The latter has a 2014 reference year and the model is being used in its comparative static specification. Envisage is a relatively standard computable general equilibrium CGE model. 2The model has been configured for a short-term closure with the following assumptions Production elasticities have been reduced to near zero, so there is little substitution possibility across s in production. In order to capture the typically durable relationship within global value chains, trade elasticities for goods have been reduced from their standard values to represent the short-run inability to replace imported components and final goods with products from other countries. The elasticity between domestic and imported goods has been set to 0.4. The elasticity of substitution across import sources has been set to 0.8. Labor supply is exogenous, while wages adjust to equate demand and supply of labor. The return to capital is fixed, while supply of capital is endogenous. 2.2 Transmission channels The shocks have been divided into four sets, but all are assumed to occur simultaneously, i.e. the final shock encompasses all shocks. 3The duration of the shocks is currently unknown, though, based on prior events, it is likely to last from 8-12 weeks and most likely unsynchronized across countries. 1. The first shock is a drop in employment by 3 percent below the baseline. With lower availability of labor, we would expect wages, ceteris paribus, to rise, while return to capital is unchanged under our assumptions. Lower labor also means lower demand for capital, as firms need a combination of labor and capital to produce goods and services. Underutilization of capacity takes place due to factory closures workers stay home, leaving capital and natural resources idle as well as social distancing forcing workers to stay at home. Due to higher rates of contagion, immediate unemployment consequences of COVID-related business closures and negative demand shock, we conservatively assume the underutilization of the labor force to be 3 on average over the whole year across all sectors of the economy. 4There is a lot of uncertainty surrounding these assumptions, and the country-specific employment effects will depend on the duration and intensity of the pandemic and containment measures, the sectoral composition of employment, and the flexibility of the labor market. 2A full description of the Envisage model is available at https//mygeohub.org/groups/gtap/File/uploads/ENVISAGE10.01_Documentation.pdf. 3The shocks are scaled down as compared with the shocks derived for Liberia under the Ebola epidemic, as in Evans et al. 2014. 4This is a conservative estimate. Some estimates put potential reduction of employment at the annual level at 10, assuming unemployment of over 30 in Q2 and returning to pre-crisis level in Q3 and Q4. https//www.stlouisfed.org/on-the-economy/2020/march/back-envelope-estimates-next-quarters- unemployment-rate - 4 - 2. The second shock cumulative with the supply shock raises the international trade costs of imports and exports by 25. The shock is applied across all goods and services. Trade costs arise when goods cross borders. The assumed increase in transport and transactions costs in foreign trade is driven by additional inspections, reduced hours of operation, road closures, border closures, increases in transport costs, etc. Evans et al. 2015 estimate that the outbreak of Ebola could lead to an increase in trade costs of 10. Since COVID-19 is affecting more countries and the containment measures seem more severe due to the efforts to contain the virus, we amplified the shock increasing international trade costs of imports and exports to 25. 3. The third shock entails a sharp drop in international tourism. This is captured via a 50 consumption tax on international tourism-related services, such as transport, accommodation, etc. This generates a typically small revenue for the relevant countries that is rebated back to households with a lump sum. 5The export tax is applied to both outbound and inbound tourist services that include accommodation, food and service activities; water, air and other transport; and recreational and other services. The effects of COVID-19 in the tourism, hospitality and recreation sectors have been unprecedented. In the accommodation and lodging sectors, quarterly revenues are down 75. Travel agents saw a slowdown in bookings of 50 in March of 2020. Airlines worldwide are expected to lose 113 billion in revenues for 2020. In the peak of the outbreak, 70 of scheduled flights in China have been canceled. As of mid-March 2020, international travel has ground to a halt, with the World Travel and Tourism Council WTTC estimating that global travel would decline at least 25 percent in 2020. To capture the effects of the drop in tourism, hospitality and recreation services, we implemented a 50 tax on the export of trade-related services, resulting in a drop in exports of tourism services at a global level of 20-32. 4. The fourth shock represents a demand switch by households who purchase fewer services requiring close human interaction, such as mass transport, domestic tourism, restaurants, and recreational activities, while redirecting demand towards consumption of goods and other services. Demand for the targeted services is assumed to drop by 15. This results in a reallocation of household demand across sectors, while total expenditures are still driven by previous shocks and relative prices of goods in the consumption basket. It is difficult to estimate the impact of social distancing and overall decline of economic activity on those selected sectors, but anecdotal evidence suggests that it is likely to be significant. With social distancing measures and closures of nonessential businesses, the bookings through Opentable network declined by 100 in the second half of March data the United States, the United Kingdom, and Germany. Depending on the length of the business closures, the annual impact could vary drastically. The decline of 15 at an annual level seems like a middle of the road estimate. 5There are a number of ways to affect demand choices by increasing the cost of purchasing the relevant good. The solution in this case has been to impose export taxes that directly affect the price of the targeted services. The revenues generated by this tax are rebated back to households. - 5 - Figure 1. Implications of the COVID-19 as implemented in the Envisage model. 2. 3 Scenarios We start by considering the effects of COVID-19 on world supply capacity, trade costs, international tourism, and demand switching, as discussed above. Then we study the consequence of similar shocks under the “amplified global pandemic” scenario. “Global pandemic” scenario In the global pandemic scenario, we aim to capture relatively rapid recovery and limited contagion, where the shocks are implemented to the full degree in China, but other countries experience shocks amounting to only half the shocks described below ? Underutilization of labor by 3 percent across all sectors in the global economy results in declining capital usage. ? Trade costs of global imports and exports increase by 25, applied across all goods and services. ? Sharp drop in international tourism captured via a 50 tax on inbound and outbound tourist- related services such as transport, accommodation, etc.. ? Reallocation of demand away from sectors requiring human interaction.