2017年-世界发展银行全球_A_GEM_for_Streamlined_Dynamic_CGE_Analysis___Structure_Interface_Data_and_Macro_Application_66页_1mb

GEM-Core: A Dynamic CGE Model for Policy Analysis

Core Content

GEM-Core is a dynamic computable general equilibrium (CGE) model developed for medium- and long-run policy analysis, particularly in developing countries. It is designed to be a template model, meaning it can be quickly adapted to different countries using an appropriately formatted database. The model is based on a Social Accounting Matrix (SAM) and is equipped with a user-friendly Excel interface called ISIM, which simplifies the process of model application and simulation.

Main Features

• Multi-purpose: GEM-Core can analyze a wide range of policies including fiscal space, public investment, social safety nets, trade, jobs, demography, poverty, and inequality.
• Flexible: It can be applied to databases with varying levels of disaggregation, from highly aggregated to highly detailed.
• User-friendly: The Excel-based interface (ISIM) allows analysts to choose between databases, define simulations, and adjust assumptions without needing advanced programming skills.
• Training-focused: The interface facilitates the organization of training around economic concepts rather than computer programming.

Key Components

1. Model Structure

• Activities: Represent production entities (private and government).
• Commodities: Include private and government outputs, and are linked to markets.
• Factors of Production: Labor and private capital.
• Institutions: Households, government, and the rest of the world (ROW).
• Taxes and Transfers: The model includes detailed tax accounts and transfer mechanisms.
• Capital Accounts: For households, government, and ROW.
• Investment and Stock Changes: Investment in private and government capital, and changes in stock levels.
• Balance of Payments: Influenced by real exchange rates, import/export quantities, and non-trade payments.

2. Macroeconomic Balances

• Government Budget: Balanced through direct tax rate adjustments.
• Savings-Investment Balance: Households’ savings are used to finance private investment, while foreign investment is determined by the balance of payments.
• Balance of Payments: Equilibrated by real exchange rate adjustments, which influence export and import quantities.

3. Labor Market

• Labor demand: Influenced by wages and unemployment rates.
• Unemployment: Endogenous, with a negative relationship between real wage and unemployment rate.
• Capital demand: For private capital, rents are flexible and clear the market.

4. Growth Mechanisms

• Endogenous growth: Driven by employment growth in private capital and labor, and total factor productivity (TFP).
• TFP: May depend on public capital growth and trade openness.
• Population growth: Declines gradually from 2.7% in 2016 to 2.37% in 2030.
• Labor force participation: Maintained at 79.5% throughout the simulation period.

Key Data and Assumptions

1. SAM (Social Accounting Matrix)

• Macro SAM for an archetype low-income country in 2015.
• Defines disaggregation into two activities (private and government), two factors (labor and private capital), and three institutions (household, government, ROW).
• Used to set base-year values for model parameters, including production technologies, commodity supplies, demands, tax rates, and transfers.

2. Elasticities

• Factor substitution: 0.7 (private), 0.7 (government).
• Import-Domestic Output Substitution: 1.5 (private).
• Export-Domestic Sales Transformation: 1.5 (private).
• TFP with respect to trade openness: 0.1.

3. Base-Year Data

• Private employment: 95.1% of total employment.
• Government employment: 4.9%.
• Unemployment rate: 5.5%.
• Capital stock (private): 180.2% of GDP.
• Capital depreciation rate (private): 4.0%.
• Government capital stock: 65.0% of GDP.
• Government capital depreciation rate: 2.5%.
• Poverty headcount ratio: 46.2% (at $1.90 a day).
• Gini index: 0.428.

4. Projections

• GDP growth: 5.01% for the simulation period (2015–2030).
• Population growth: Declines from 2.7% in 2016 to 2.37% in 2030.
• Labor force participation rate: Maintained at 79.5%.
• Population aged 15–64: Increases gradually from 53.8% in 2015 to 57.8% in 2030.

Simulations

1. Base Scenario

• Represents business-as-usual conditions without policy changes.
• GDP growth is endogenously adjusted through labor productivity.
• Taxes and transfers are fixed at base-year rates.
• The model includes three equilibrating mechanisms for the government budget, savings-investment balance, and balance of payments.

2. Non-Base Scenarios

• Used to explore policy impacts and exogenous shocks.
• Includes adjustments in tax rates, government spending, investment sources, and exchange rates.
• Demonstrates how changes in real interest rates, debt stocks, and foreign reserves can be simulated.

Poverty Module

• Based on MAMS (Multinational and Multisectoral Model) by Lofgren et al. (2013).
• Offers three approaches for poverty and distributional analysis:
  • Constant elasticity of poverty with respect to per-capita welfare.
  • Log-normal distribution of per-capita welfare.
  • Real-world household survey data.
• In this application, the log-normal distribution is used.
• The module is linked to base-year poverty and distributional data.
• For single representative household applications, it assumes no change in distribution.

Conclusion

GEM-Core is a versatile and accessible model that supports dynamic policy analysis in developing countries. Its Excel interface and minimal data requirements make it ideal for quick application and training purposes. The model can be used to evaluate the macroeconomic and sectoral impacts of various policies, including fiscal, trade, and demographic changes, while also incorporating poverty and inequality analysis. It is part of a broader initiative by the World Bank to provide open access to research and support development policy discussions.