Macroeconomic VS' CGE Models

1
Macroeconomic VS. CGE Models

• CGE models incorporate production at a level
  of aggregation that permits the analysis of
• structural change and also captures the
  essential transmission mechanism of production,
• demand, and trade. They also incorporate
  market mechanisms and policy instruments that
• work through price incentives. Therefore,
  resource allocation theory is always the
• fundamental framework of analysis.
• In contrast to macroeconomic models which are
  often difficult to trace the causal
• mechanisms at work, the mechanisms driving
  CGE models are clear and easy to grasp
• because their structure is rooted in received
  theory. Many economists believe that
• transparency is a key characteristic that an
  empirical general equilibrium model must have
• if it is to provide a framework for policy
  analysis.
• Empirical general equilibrium models that can
  be solved numerically are thus useful
• to provide a bridge between the theorist, the
  planner and the practical policy maker.

2
Macroeconomic VS. CGE Models

• Consider, for example, a disagreement about the
  appropriateness of a major
• devaluation in a country facing a foreign
  shortage. Is the problem more
• macroeconomic, with the price level seen
  tied to the exchange rate by strong
• cost-push factors that make a real devaluation
  impossible? Or, is the whole
• problem that the opposition to exchange-rate
  adjustment is based on income
• distribution consideration? What is needed is an
  economy-wide framework that
• permits an explicit specification of an
  economys operation where each of these
• views can be evaluated.
• Based on macroeconomic model, the central focus
  of analysis include the level
• of foreign debt, inflation, government revenue,
  the aggregate level of economic
• activity, unemployment, and perhaps the
  distribution of income to broadly
• defined factors of production (labor, capital).

3
Macroeconomic VS. CGE Models

• Most countries today, however, work within the
  environment of a mixed
• economy in which the market plays a central
  role. The exchange rate, taxes,
• tariffs, subsidies and other policy variables
  that affect relative prices and
• incentives through the market mechanism have
  become important. It is crucial
• to understand how incentive policies affect the
  allocation of resources and the
• structure of growth by using the general
  equilibrium model.
• Wages, prices, and the exchange rate are viewed
  not in terms of their impact on
• the aggregate flow of funds and the various
  macro balances, but in terms of
• their impact on relative factor returns
  across sectors and by type of factor (labor
• by skill and sector, capital by sector,
  etc.)

4
Normal Step in CGE Modeling

• 1. Specify dimensions of the model
• No. of goods and factors, consumers, countries,
  and active markets
• 2. Select functional forms for production
  (Leontief, Cobb-Douglas, or CES), transformation,
  and utility functions
• 3. Construct micro-consistent dataset
• Data satisfy zero profit conditions for each
  activity (e.g., perfectly competitive market)
• Data satisfy market clearing conditions
• The initial data for this economy is SAM
• 4. Initialization and Calibration
• Parameters are chosen such that functional forms
  and data are consistent.
• 5. Replication
• Once all the parameters are specified, the model
  is solved to reproduce the benchmark data.
• 6. Counter-factual experiments

5
Structure of the Model
6
?????????????????????????????????????????
7
Demand for Primary and Intermediate Inputs
8
Demand for Domestic and Export Commodities
9
Demand for Final Consumption
10
Equations for Normal Profit Conditions
11
Equations for Equilibrium in the Economy
12
Equations for Income and Expenditure of
Households and Government
13
Models Initialization and Calibration

• Initialization assigning the base value to each
  variable before solving for the base-case
  solution. Most values for initialization are
  obtained from SAM table. However, some values,
  particularly those of parameters in production
  equations, are computed by using calibration
  technique.
• Before doing initialization and calibration, you
  need to know the location on the SAM table
  containing the values related to each equation in
  your model.

14
Models Initialization and Calibration

• Production Block
• Assign the value of 1 to PD, PE, PM, PQ and PX.
• On SAM table, locate cell containing the value of
  PDD, PEE, PMM, PQQ, PXX. Dividing those
  values by PD, PE, PM, PQ and PX to obtain the
  value of D, E, M, Q and X.
• Now we need to apply the Calibration technique to
  compute the value of at, aq, bt and bq

(Eq. 1) Production Function (TotalProd)
(Eq. 2) CET Transformation (CETEQ)
(Eq. 3) ARMINGTON Composite Supply (ARMING)
15
Models Initialization and Calibration

• (4) There are 3 steps in calibration.
• Step (4.1) - Elasticity parameter (?) Mostly
  obtained from using econometric method (e.g. OLS,
  GLS, ML).
• Step (4.2) - Parameter of input share (b)
  Computed by using the re-arranging equation of
  price ratio. For example, we re-arrange the E/D
  ratio equation to calculate the value of bt
• Step (4.2) - Parameter of input share (a)
  Computed by using the re-arranging the production
  function. In this example, the CET equation is
  re-arranged, enabling us to obtain the value of
  at.

Assign to value of elasticity parameter (?)
Compute the share parameter (b)
Compute the technology parameter (a)
(Eq. 10) E/D Ratio (EDRAT)
(Eq. 2) CET Transformation (CETEQ)
16
Models Closure Rule

• Typically, the models leave 3 sets of
  variables for closure rule.
• (1) Savings Investment Balance
• Option 1 Saving is investment driven
  Option 2 Investment is saving
  driven
• Investment variable
  fixed Investment
  variable free
• Saving rate free
  
  Saving rate fixed
• (2) Exchange rate and Foreign Savings
• Option 1 Exchange rate is floated
  Option 2 Exchange rate is
  fixed
• Exchange rate
  free
  Exchange rate fixed
• Foreign saving
  fixed
  Foreign saving free
• (3) Factor Markets
• Capital market
  Labor Market
  
• Option 1 Capital is mobile and fully
  employed Option 1 Labor is mobile and fully
  employed
• WFDIST(cap)
  fixed
  WFDIST(labor) fixed
• WF(cap) free
  WF(labor) free
• QF(cap) free
  QF(labor) free
• QFS(cap)
  fixed QFS(labor) fixed

17
Modeling Techniques and Concerns

• Technique Construct the initialization and
  calibration by automatically obtaining initial
  value from SAM.
• Concerns There are indicators for checking the
  correctness of model
• (1) WALRAS is zero
• (2) SAM is balanced
• (3) Base-cases solution is identical to the
  original data
• However, obtaining the base-case solution
  satisfying these 3 conditions does not guarantee
  that your model is correct.

18
Social Accounting Matrix (SAM)

• A SAM is a square matrix that builds on the
  input-output table, but it goes
• further.
• A SAM considers not only production linkages,
  but also concerns on income-
• expenditure (institutions are introduced).
• A SAM is consistent data system that provides a
  static image or a snapshot of
• the economy.

19
Accounts of the 2005 SAM for Thailand

• (1) Labor (2) Capital
• (3) ??????? 42 sectors (4)
  Domestic Commodity 42 sectors
• (5) Foreign Commodity 42 sectors (6) Trade and
  Transport Margin (TTM)
• (7) Direct Taxes (8) VAT
• (9) Excise Taxes (10) Import Duties
• (11)Other Indirect taxes. (12) Subsidies
• (13) Household 5 groups (14) Government
• (15) Private Enterprise/State Enterprise
• (16) The Rest of The World (ROW)
• (17) Capital Account (KA)

20
A GAMS Tutorial

• Structure of a GAMS model
• Sets
• Data
• Variables
• Equations
• Objective functions
• Model and solve statements
• Display statements
• The .lo, .l, .up, .m database
• GAMS output

21
Structure of a GAMS Model
22
Example GAMS representation of the transport
problem

• Given 1) supplies at several plants and demands
  at several markets for a single
• commodity
• 2) unit costs of shipping the
  commodity from plants to markets
• Question how much shipment should there be
  between each plant and each market
• so as to minimize total
  transport cost?
• Indices i plants, j markets
• Given data

23
Example (Continued)
24
Sets

• Sets
• i canning plants / seattle, san-diego /
• j markets / new-york, chicago, topeka /
• i Seattle, San Diego
• j New York, Chicago, Topeka.
• Set i canning plants / seattle, san-diego /
• Set j markets / new-york, chicago, topeka /
• Set t time periods /19912000/
• Set m machines /mach1mach24/
• t 1991,1992,1993, ....., 2000
• m mach1, mach2,......, mach24,
• Alias statement
• Alias (t,tp)

25
Data

• Data entry by list
• Parameters
• a (i) capacity of plant i in cases
• / seattle 350
• san-diego 600 /
• b (j) demand at market j in cases
• / new-york 325
• chicago 300
• topeka 275 /
• Data entry by table
• Table d (i, j) distance in thousands of miles
• new-york chicago topeka
• seattle 2.5 1.7
  1.8
• san-diego 2.5 1.8
  1.4
• Data entry by direct assignment
• Parameter c (i, j) transport cost in
  thousands of dollars per case
• c (i, j) f d (i,
  j) / 1000

26
Variables

• Variable
  
  x (i, j)
  shipment quantities in cases z total
  transportation costs in thousands of dollars
• Once declared, every variable must be assigned a
  type. The permissible types are
• free (default), positive, negative, binary, and
  integer.
• The variable that serves as the quantity to be
  optimized must be a scalar and must
• be of the free type. In our example, z is kept
  free by default, but x (i, j) is constrained
• to non-negativity by the following
  statement.
• Positive variable x

27
Equations

• Equation Declaration
• Equations
• cost define objective function
• supply (i) observe supply limit at plant i
• demand (j) satisfy demand at market j
• GAMS summation notation

28
Equations (Continued)

• The dollar condition
• a(b gt 1.5) 2 is equivalent to if (b gt 1.5),
  then a 2
• dollar on the left no assignment is made unless
  the logical condition is satisfied.
• rho(i) (sig(i) ne 0) (1./ sig (i) ) 1.
  
• rho(i) sig (i) (1./sig (i) )
  1.
• dollar on the right an assignment is always
  made.
• x 2 (y gt 1.5) is equivalent to
  if (y gt 1.5) then (x 2), else (x 0)
• or it can be re-written with an
  explicit if-then-else as
• x 2 (y gt 1.5) 0.5 (y le 1.5)

29
Equation Definition

• The name of the equation being defined
• The domain
• Domain restriction condition (optional)
• The symbol , ,
• Left-hand-side expression
• Relational operator l, e, or g
• Right-hand-side expression
• Example
• cost .. z e sum ( (i, j), c (i, j) x (i, j)
  )
• supply (i) .. sum (j, x (i, j) ) l a (i)
• demand (j) .. sum (i, x (i, j) ) g b (j)

30
Objective Function

• To specify the function to be optimized, you must
  create a variable, which
• is free (unconstrained in sign) and
  scalar-valued (has no domain) and which
• appears in an equation definition that equates
  it to the objective function.

Model and Solve Statements

• model transport /all/
• model transport / cost, supply, demand/
• solve transport using lp minimizing z
• The format of the solve statement is as follows
• 1. The key word solve
• 2. The name of the model to be solve
• 3. The key word using

31
Model and Solve Statements (Continued)

• 4. An available solution procedure. For
  instance,
• lp for linear programming
• nlp for nonlinear programming
• 5. The keyword minimizing or maximizing
• 6. The name of the variable to be optimized

Display Statements

• display x.l, x.m

The .lo, .l, .up, .m database

• .lo lower bound
• .l level or primal value
• .up upper bound
• .m marginal value or dual value

32
The .lo, .l, .up, .m database (continued)

Assignment of variable bounds and/or initial
values
x.up (i, j) capacity (i, j)
x.lo (i, j) 10.0 x.up
(seattle, new-york) 1.2 capacity
(seattle, new-york) These statements
must appear after the variable declaration and
before the Solve statement Transformation and
display of optimal values After the
optimizer is called via the solve statement, the
values it computes for the primal and dual
variables are placed in the database in the .l
and .m fields. We can read these results and
transform and display them with GAMS statements.
parameter pctx (i, j) percentage of market
js demand filled by plant i pctx (i, j)
100.0 x.l (i, j) / b (j) display pctx

33
Introduction for the Stochastic CGE Model

• CGE modeling has become a popular tool to
  examine the economy-wide impacts of economic
  policies, but uncertainty about the values to
  assign to parameters can be a major limitation.
• In most cases, there are no econometric
  estimates of the majority of model parameters.
• The usual procedure is to do the models
  sensitivity by varying one parameter at a time,
  while keeping others at their base values.
• However, this procedure ignores the possibility
  that two or more parameters could act in
  combination to yield unusual or unexpected
  results
• The objective of this study is, therefore, to
  investigate the role of parameter uncertainty in
  the model by performing the Monte Carlo analysis.

34
Monte Carlo Analysis

• In this analysis, the variability and uncertainty
  of each input parameter is represented by a
  frequency distribution.
• The user needs to provide the distribution type
  along with the mean, standard deviation and
  minimum and maximum values of each input
  parameter.
• Base on the frequency distribution of the input
  parameters, the Monte Carlo simulation selects a
  randomly generated input data set and calculates
  the corresponding output.
• Then, a new input data set is generated at
  random, and the corresponding new output is
  calculated. This process is repeated until the
  statistical distribution of the model output
  reaches a stable state.

35
Example Monte Carlo Simulation
36
????????????????? Stochastic CGE

• ??????????????????????????????????????????????????
  ??? CGE

CES
CET
CES
CES
CES
37
????????????????? Stochastic CGE

• ??????????????????????????????????????????????????
  ??????????????????????
• ???????????? ???? Armington elasticity
  ???????????????????????? ???????????????
• ????????????????????????????????????? ???????
  ??????????????????????????
• ???????? ????????????????????????????? (degree
  of substitution) ?????????????
• ????????? ???????????????????????????????????????
  ?????????????????????????
• ????????????
• ??????????????????????????????????????????????????
  ???????????? ??????????
• ?????????????? ??????????????????????????????????
  ??????????????????????????
• ?????????????????????????? ???????
  ???????????????????????????????????
• ??????? ?????????????????????????????????????????
  ????? ?????????????????????
• ?????????????????????????????????????????????????
  ??????

38
????????????????? Stochastic CGE

• ??????????????????????? Monte Carlo Simulation
  ?????????? stochastic CGE
• ????????????? ????????????????????? ???????
  ?????????????????????????? ???????????????????????
  ?????????????? (normal distribution)
  ????????????????????????????????? ??????? 50
  ?????
• ?????? Monte Carlo ????????????????????????????
  ? ? ???????? ?????????????????????????????????????
  ??? ? ???? mean ??? standard deviation
  ????????????????????????????? (distribution of
  results) ????????????? ? ??????????
• ??????? ???????????????????????????????????????
  ? ??????

39
????????????????? Stochastic CGE

40
??????????????????????

• ????????? ???????????????????????????????????????
  ??? ???????????????????????????????????? 2.50
• ?????????????????????????????????? 3 ??????????
  ??????
• ???????????????????????????? Deterministic CGE
• ????????????????????????? Sensitivity Analysis
• ????????????????????????? Monte Carlo Simulation
  ??
• ???????? Stochastic CGE

41
??????????????????????

• ????????????????????????????? Deterministic CGE

• Real GDP ?????????????? 0.1
• ??????????? Exports ???????????? 0.74
• ??? Imports ??????? 1.43

42
??????????????????????

• ?????????????????????????????????? Deterministic
  CGE
• 1) ???????????????????????? 2.5 ?????
  ??????????????????????????????????????
• ???????????? ???????????????????????????????
  ????????????? ???????????????
• ???? ? ???????????????? 0.86 0.74 ??? 0.58
  ???????? ????????????????????????????
• ??????????????????????
• 2) ????????????? ?????????????????????????????????
  ???????????????????????
• ???????? ????????????? ??????????????????? ?
  ??????????? 1.40 1.31 ???
• 1.97 ???????? ???????????????????????????????????
  ??????????????????????????????
• 3) ???????????????????????????????????????????????
  ??????????????????????????????
• ????????????????? ????????????????????????????????
  ??????????????????????? ??
• ??????????????????????????????????????????????????
  ?? ?????????? ????????????
• ???? ? ???? 0.03 0.20 ??? 0.29 ????????
• 4) ???????????????????????????????????????????????
  ?? ??????????????????????
• ??????????????????? ????????????????????????????
  0.05 ??? 0.16 ????????

43
??????????????????????

• ???????????????? Sensitivity Analysis
  ??????????????????????????????????????????????????
  ????????????????????????????
• ?????????????????????????????? 25 ??? 50
  ??????????????? ??????????????????
• ?????????????????????????????????????????????????
  ????????? ????????????????????????
• ?????????????????????????????????????????????????
  ????????????? ???????????????
• ?????????????????????????????????????????????????
  ??????????????? ?????????????
• ??????????????????????????? real GDP
  ???????????????????????????????????????????
• aggregate demand ????????????
• ???????????? ?????????? ??????????????????????????
  ???? 50 ???????????????
• ????????????????????????????????? -1.40 ????
  -1.51 ???????????????????
• ????????????????????? 1.10 ???? 1.32
  ?????????????????????????????????????????? -
• -0.03 ???? 0.03 ???????????????????????????????
  ??????? 0.05 ???? 0.09
• ???????
• ???????????? ????????????????????????
  ????????????????????????????

44
??????????????????????

• ???????????????? Sensitivity Analysis
  ??????????????????????????????????????????????????
  ???????????????
• ?????????????????????????????? 25 ??? 50
  ??????????????? ?????????????????
• ?????????????????????????????????????????
  ???????????????????????????????????? ??? Real GDP
• ???????????? ??????????????? ?????????????????????
  ????????? 50 ????????????
• ??? ?????????????????????????????????????? 0.74
  ???? 0.81 ??????????????
• ?????????????????? 0.16 ???? 0.19 ??? Real GDP
  ???????????? 0.10 ???? 0.15
• ???????
• ???????????? ????????????????????????
  ????????????????????????????

45
??????????????????????

• ???????????????? Monte Carlo Simulation
  ?????????? Stochastic CGE

?????????????????????????????????????????????????
??????????????????????
46
??????????????????????
???????????????? Monte Carlo Simulation
?????????? Stochastic CGE

• ?????????????????????????? (distribution
  property) ???????? Monte
• Carlo Simulation ?????????????????????
  ?????????????????????????????? (the
• most volatile variable) ?????????
  ??????????????????????????????? (volatility)
• ?????????????????????????????????????????????????
  ?????????????? ?????????
• ????????????????????????? ???????????????????
  (degree of volatility) ??????????????? percentage
  of standard deviation per mean
• ??????????????????????????? ???????????
  ?????????????????????????????
• ????????????????????? ????????
  ???????????????????? ???????????????????
• ???????? ?????????????????????
  ????????????????????????????????????????
• ???????? ?????????????????????????
  ??????????????????? ???????? ?????????
• ?????????????????????????????????????????????????
  ????????????????????????
• ????????????????

47
??????????????????????
???????????????? Monte Carlo Simulation
???????????????????????
48
??????????????????????
???????????????? Monte Carlo Simulation
?????????? Stochastic CGE

• ????????? ????? Monte Carlo Simulation
  ?????????????????????????????????????
• ????????????????????????????????????
  ???????????? ???????????????????????????
• ???????????????????????????? ?????? crude oil
  and coal (????????????? 4 ??? S.D./Mean
• ??????? 0.24) unclassified (????????????? 42
  ??? S.D./Mean ??????? 0.15) other
• transportation (????????????? 32 ??? S.D./Mean
  ??????? 0.14) ??????? ????????????
• ????????? crude oil and coal ????? S.D./Mean
  ??????? 0.03 unclassified ???????????? 0.02 ???
  other transportation ???????????? 0.02
• ???????? ???????????????????????????????????? ???
  ??????????????????????????? ??????
• ??????????? ? ????? S.D./Mean ??????????? 0.52
  ??? 0.046 ?????????????????????????????
• ?????????????????????????????????????
• ??????????????????????????? ????? Nominal GDP
  Real GDP ??? GDP deflator ?????
• S.D./Mean ??????? 0.04 0.03 ??? 0.03 ????????

49
??????????????????????
???????????????? Monte Carlo Simulation
?????????? Stochastic CGE

• ???????? Stochastic CGE ???????????????????????
  degree of volatility ????
• ?????????????????????????????????????????????????
  ?????????????????????
• ?????????? ? (shock) ????????????????????????????
  ?????????? ????????????
• ?????????????????????????? (distribution of
  results) ?????????????????????
• ????????????????????????????? ?
  ???????????????????????????? ????????????????
• ?????????????????? CGE ??? deterministic
  ??????????????????????????????????
• ?????

50
??????????????????????
???????????????????????? Monte Carlo Simulation
???????????????????????