Transcript Lecture15 - Stanford University

Economics 216
The Macroeconomics of
Economic Development
Lawrence J. Lau, Ph. D.
Kwoh-Ting Li Professor of Economic Development
Department of Economics
Stanford University
Stanford, CA 94305-6072, U.S.A.
Winter, 1999-2000
Phone: 1-650-723-3708; Fax: 1-650-723-7145
Email: [email protected]; Website: www.stanford.edu/~ljlau
Lecture 15
Applied General Equilibrium Models
Lawrence J. Lau, Ph. D.
Kwoh-Ting Li Professor of Economic Development
Department of Economics
Stanford University
Stanford, CA 94305-6072, U.S.A.
Winter, 1999-2000
General Equilibrium Models of the Economy
 Under
the assumptions of:
 (1)
concave technologies;
 (2) quasiconcave preferences;
 (3) price-taking behavior
 (4) profit maximization by producers;
 (5) utility maximization by households.
 Characterization
of a competitive general equilibrium
(Excess demand is less than or equal to zero in every
market):
 Existence
 Uniqueness
 Optimality
Lawrence J. Lau, Stanford University
3
General Equilibrium Models of the Economy
 Welfare
Theorem: A competitive general equilibrium is
efficient
 Converse Theorem: An efficient allocation can be realized
as a competitive general equilibrium
Lawrence J. Lau, Stanford University
4
Why is Partial Equilibrium Analysis not
Enough?
 Everything
depends on everything else
 Other things are not equal
 Example: A given
policy measure may change both the supply
and the demand sides with the outcome on both the equilibrium
price and quantity not easily predictable a priori
Lawrence J. Lau, Stanford University
5
Why Applied (Computable) General
Equilibrium (CGE) Models?
 Analytical
indeterminacy of effects
 Need to know magnitude as well as direction
 Analytical intractability--substitution of numerical
simulation for analysis
 Sensitivity analysis
Lawrence J. Lau, Stanford University
6
A Simple Static Applied General Equilibrium
Model: Specification
 Economic
agents
 Households
(utility functions)
 Firms (production functions)
 Goods
and factors
 Initial Endowments
 Leisure
 Inventory
 Capital
 Behavior
 Utility
maximization
 Profit maximization
 Markets
 Simultaneous
clearing
with zero excess demand of all goods
Lawrence J. Lau, Stanford University
7
A Simple Static Applied General Equilibrium
Model: Specification
 Choice
of a numeraire good (zero degree homogeneity)
 Choice of assumptions on the utility and production
functions
 Choice of functional forms for utility and production
functions
Lawrence J. Lau, Stanford University
8
Specification
 Households
(Preferences)
 Demander
of goods for consumption
 Supplier of labor
 Supplier of saving
 Owner of capital
 Firms
(Technologies)
 Demander
of capital
 Demander of labor
 Supplier of goods for consumption and investment
Lawrence J. Lau, Stanford University
9
Specification
 There
is no government, no external sector, no money and
no financial sector
Lawrence J. Lau, Stanford University
10
The Simplest System of Equations
Households
Demand for consumption =
DC (r*, w*, K-1, SS)
Supply of labor =
SL (r*, w*, K-1, SS)
Supply of savings =
SS (exogenously given)
Firms
Demand for capital =
DK (r*, w*)
Demand for labor =
DL (r*, w*)
Supply of output =
SO (r*, w*)
Lawrence J. Lau, Stanford University
11
General Equilibrium
General Equilibrium
Demand for capital = DK (r*, w*) = Supply of capital = K-1
Demand for labor =DL (r*, w*)=Supply of labor= SL (r*, w*, K-1, SS)
Supply of output = SO (r*, w*) = Demand for consumption+Savings
= DC (r*, w*, K-1, SS) + SS
Lawrence J. Lau, Stanford University
12
Determination of the Parameters:
Calibration versus Econometric Estimation
 The
derivation of the numerical values of the parameters
 The calibration approach
 matching
quantities and prices in the base period
 overly dependent on assumptions on the functional forms
 The
econometric approach
 estimating
parameters on the basis of a time-series of
observations
 permits validation of estimated values of parameters with actual
empirical experience
 functional form and other assumptions can be empirically tested
Lawrence J. Lau, Stanford University
13
Solution of the Model:
The Choice of Algorithms
 Fixed
point algorithms (Scarf)
Lawrence J. Lau, Stanford University
14
Welfare Analysis
 Compensating
variations--the sum of additional consumer
expenditures required in order to achieve the old levels of
utilities at the new prices
 Equivalent variations--the sum of the additional consumer
expenditures required in order to achieve the new levels of
utilities at the old prices
 The social welfare function (interpersonal comparison of
utilities required)
Lawrence J. Lau, Stanford University
15
Extension to Multiple Periods
 A sequence
of static general equilibria linked by
endogenously determined savings and investments
 The rate of time preference (choice between present and
future consumption)
 The assumption of intertemporal separability
 U(C1, C2, …, CT) =  Ut (Ct)
Lawrence J. Lau, Stanford University
16
The Importance of the Terminal Conditions
 For
finite horizon models, it will be optimal to allow the
capital stock to go to zero at the terminal point, which
cannot possibly correspond to a real world situation
 The terminal conditions have a significant impact on the
simulation results
 Solutions:
 Infinite
horizon (steady-state) models
 Ad hoc savings function
Lawrence J. Lau, Stanford University
17
The Role of Rational Expectations
 A rational
expectations general equilibrium implies that the
prices in every period must be ex ante anticipated by the
economic agents
 A backward recursive solution algorithm is required
Lawrence J. Lau, Stanford University
18
Extension to Open Economies
 Trade
(Exports and Imports)
 Foreign direct investment
 Foreign portfolio investment, loans and aid
 Tariffs, quotas, and other non-tariff barriers
 The exchange rate
 Technology transfer
Lawrence J. Lau, Stanford University
19
The Introduction of Government:
Expenditures and Taxes
 Government
expenditure (public consumption) can be
treated as an argument in the utility function
 Government can also be treated as an independent
economic agent, with its own objective function and
behavioral assumptions
 Government expenditures and public capital stocks may
affect both the consumption behavior of households and
production behavior of firms
 Likewise, government taxation may also affect both the
consumption behavior of households and production and
investment behavior of firms
Lawrence J. Lau, Stanford University
20
The Introduction of Money and the Financial
Sector
 The
neutrality of money--the absence of money illusion (Is
it true?)
 Does indexing have an impact? (it may depend on
anticipations/expectations)
 The “Cash-in-Advance” Constraint
Lawrence J. Lau, Stanford University
21
The Possibility of Multiple Equilibria
 Multiple
equilibria are possible
 “flat”
indifference surfaces
 rational expectations equilibria
 Rank-ordering
multiple equilibria
Lawrence J. Lau, Stanford University
22
The Importance of Sensitivity Analysis
 The
robustness of the simulation results must be tested
with sensitivity analysis
Lawrence J. Lau, Stanford University
23
The Role of Uncertainty:
Incompleteness of Markets
 Availability
of futures markets
 Availability of insurance markets
 Availability of contingent markets
Lawrence J. Lau, Stanford University
24