Modeling Issues Related to EDRC Models

1
Modeling Issues Related to EDRC Models
Mushegh TumasyanEconomic Development and
Research Center, Armenia
Ex-ante Poverty Impact Assessment of
Macroeconomic policies International
WorkshopWashington D.C. October 14-15, 2003
2
Contents

• Why EDRC participates in PRSP and modeling
  processes and how it contributed to the
  development of PRSP in Armenia?
• What are the Criteria that a Model Dealing with
  Poverty Issues should meet?
• What Environment EDRCs Models are Developed for?
• Ideas underlying the model Political economics
  of the Model
• The Set of The Models

3
Why EDRC Participates in PRSP and Modeling
Processes and how It Contributed to the
Development of PRSP in Armenia?

• To contribute to a comprehensive understanding of
  poverty and possibilities of poverty reduction in
  Armenia
• To design independent domestic policy programs
• To provide a tool to Civil Society Organizations
  (CSO-s) for policy evaluation
• To create institutional basis for strengthening
  alternative macro-modeling capacities of the
  non-governmental sector
• To promote CSO-s participation in public policy
  making

4
PRSP Processes in Armenia EDRC Contribution

• EDRC conducted the Policy Choice for Poverty
  Reduction project to develop policy
  recommendations for PRSP
• Using EDRCs models, our experts developed the
  macroeconomic strategy and analysis of the
  macroeconomic policy impact on poverty for the
  first draft of PRSP for Armenia
• However in the final PRSP it was replaced by the
  macroeconomic framework of the IMFs PRGF program
  with Armenia
• In the final PRSP the tools that were used to
  derive the policy impact on distribution and
  poverty incidence are not transparent and have
  not been disclosed

5
What are the Criteria that a Model Dealing with
Poverty Issues Should Meet?

• Ability to estimate policy impact on welfare
• The model needs to produce a macroeconomic
  framework, to forecast income distribution and
  poverty incidence
• For participatory policy making, models also
  should be transparent and user-friendly
• Data availability constraint should be properly
  taken into account

6
What Environment EDRCs Models are Developed for?

• Low income transition economy
• High level of transitional polarization
• Corruption and resulting market distortions
• Openness to trade and capital account
  liberalization
• Low level of resource utilization
• Stable prices

7
The Set of the model
Policy Revision
MAGMA Macroeconomic framework
Policy
Growth and Distribution Model Poverty Incidence
Growth
Exogenous variables
GINI
Income Distribution Matrix Income Distribution
and GINI
Household Survey Data
8
Growth and Distribution Model Estimates poverty
incidence for different combinations of Gini
coefficient and economic growth

• Estimation of Lorents curve function
• Calculation of coefficients of Lorents curve for
  different Gini coefficients
• Calculation of poverty incidence for given
  combinations of Gini and income growth index

9
Growth and Distribution Model Estimation of
Lorents Curve Function

• To choose the type of distribution function,
  logarithmic, exponential and combined functions
  and polynomials were observed.
• We estimate Lorents curve by solving the
  following system of equations.

Equation (2) expresses the relation between the
selected function and actual Gini index, and
equation (1) is the equation satisfying the
condition x100 y100 and x0 y0.

• To choose the best function we should take one
  with minimum dispersion from actual deciles
  values. In case of Armenia we select following
  functional form

10
Growth and Distribution Model Estimation of
Poverty Incidence for Different Combinations of
Gini Coefficient and Economic Growth
Income
Change of Gini
Poverty line
Population
Poverty Reduction
11
Growth and Distribution Model Calculation of
Poverty Incidence for Various Combinations of
Gini and Income Growth Index
12
Macroeconomic Adjustments and Growth Model for
Armenia (MAGMA)

• Main Features
• behavior of different income groups and
  institutions
• decomposition of consumption and import functions
  by income groups
• Assuming that higher income groups have higher
  propensity to save and to consume imported goods,
  decrease of polarization leads to economic growth

• Main Outcome
• Endogenous GDP growth
• Forecast GDP by expenditures, production and
  income
• Estimation of employment by sectors

13
MAGMA Main Equations
14
Description of Variables of Main Equations (1)
15
Description of Variables of Main Equations (2)
16
Income Distribution Matrix (IDM)

• IDM assesses the policy impact on income
  distribution
• It produces a distribution table showing the type
  (wages, profit, transfers) and sources of income
  for each income group, finally deriving Gini
  coefficient.
• Analyzes the distribution of 1) assets, 2)
  employment and 3) pensioners for each income
  group
• IDM uses Households surveys Data, outcomes of
  the Macro Model (MAGMA) and some other exogenous
  estimates

17
Income Distribution Matrix
18

• Economic Development and Research Center
• www.edrc.am
• info_at_edrc.am