FOR PARTICIPANTS ONLY

wallsgloomyΔιαχείριση

28 Οκτ 2013 (πριν από 3 χρόνια και 8 μήνες)

89 εμφανίσεις

FOR PARTICIPANTS ONLY





ECONOMIC AND SOCIAL COMMISSION FOR ASIA AND THE PACIFIC



Subregional Seminar on Macroeconomic Policy Analysis and


Modeling in the Economies of Central Asia


20
-
22 June 2001

Tashkent, Uzbekistan







THE EXPERIENCE OF INDIA IN
USING MODELLING FOR

MACROECONOMIC POLICY ANALYSIS















____________


*This paper has been reproduced as submitted. A limited number of copies has been
issued for use only by participants.



This paper was prepared by Mr Shashanka Bhide, Chief Ec
onomist and Head of
Macroeconomic Monitoring and Forecasting Division, National Council of Applied
Economic Research, New Delhi.



The designations employed and the presentation of the material do not imply the
expression of any opinion whatsoever on the p
art of the United Nations concerning the legal
status of any country, territory, city or area, or of its authorities, or concerning the
delimitations of its frontiers or boundaries.

The Experience of India in Using Modelling for

Macroeconomic Policy Analy
sis


Summary


Macroeconomic modelling research in India has a long tradition. While the initial application
of economic modelling was motivated by the planning requirements, econometric approach
which focussed more on description of the economy and policy
assessment also evolved over
the years. The econometric approach has remained largely in the arena of the academic
efforts at the universities and research institutions. The plan models were developed and
maintained by the official agencies. The CGE approa
ch has also been the effort in the
academic arena.


The policy needs of empirical analysis of the economy have increased with greater
liberalization of the economy. There is also a need for assessment of the economy for
commercial activity. The macro econo
mic models now are looked upon to provide these
research inputs for policies both for the government as well as private sector.


India’s economic policy reforms of the 1990s resulted in wide ranging changes in the way the
markets function. The financial ma
rkets including foreign exchange markets as well as the
commodity markets are freed from strict administrative controls. The macro models have
evolved to accommodate these changes in their specification of the economic inter
-
relationships.


This paper pro
vides a review of the development of the macro modelling research in India,
major changes in economic policies that began in the early 1990s and finally the manner in
which the macro models have attempted to incorporate the changed economic policy
environm
ent.





2

The Experience of India in Using Modelling for Macroeconomic Policy
Analysis



By Shashanka Bhide
1


I.

Introduction



Macroeconomic modelling, defined broadly as development of ‘economy
-
wide’
empirical models, has been a significant area of research
in India since the early 1950s
2
.
Besides the empirical models, there have been a number of important contributions by
various researchers to the conceptualization of the Indian economy, which can be termed
theoretical modelling of the economy. For example,

Pandit (1999) notes the two basic strands
of ‘classical’ and ‘Keynesian’ approaches, which have been debated in the Indian economic
literature. The debate has been on the extent to which the Indian economy fits the key
elements of the two basic paradigms
of economic theory. Dutt (1995) provides a review of
the elements of the open
-
economy modelling framework found in the macroeconomic models
of the Indian economy, in the context of opening up of the economy.


The macro modelling research has followed the
evolution of the economy both in
terms of the policy thrust as well as the changing economic structure. For example, while
attention to agriculture and the subsistence nature of production for a majority of the
producers in this sector has continued, great
er attention is now paid to the services, trade and
financial infrastructure. Attention has shifted from analysing merely the implications of
government policies on the economy to such factors as private capital flows from abroad.


Thus, macro
-
modelling re
search has been responsive to the needs of policy analysis in
India, by providing an empirical framework for addressing the issues such as the implications
of the alternative monetary, fiscal and trade policy measures. The macro modelling research,
particu
larly that was not part of the ‘planning’ variety also grew more as a result of work
within the academic circles rather than merely as a policy tool. This feature of research also
enabled the growth of alternative approaches to modelling the economy. The e
mpirical macro



1

Chief Economist and Head of Division of Macroeconomic Monitoring and Forecasting, National Council of
Applied Economic Research, New Delhi


2

In fact the first macroeconometric model w
as estimated by Narasimahan in 1948 under the guidance of
Tinbergen (Krishnamurty, 1992)


3

modelling work has kept pace with similar approaches in many other developing economies
although the level of desegregation and use of high frequency data have been relatively low
3
.

This paper is an attempt to review the experience of using
macro
-
modelling research
for policy analysis in India. We focus only on the empirical modelling of the Indian
economy.


More specifically, the objective of this paper is three
-
fold:

(a)

To briefly review the alternative approaches to the modelling of the Ind
ian
economy in the more recent period of the 1990s;

(b)

To highlight the issues raised by the changes in India’s economic policies of the
1990s and their implications to macro modelling; and

(c)

To discuss the changes in the specification necessitated by the pol
icy changes that
have sought to reduce discretionary role of administrative regulation.

Remaining part of the paper is organised in four sections. The alternative approaches
are reviewed in Section II, issues raised by the economic policy changes are disc
ussed in
Section III, changes in the specification of the models required by the policy changes are
noted in Section IV and the Section V provides the concluding remarks.



II.

Alternative Approaches to Modelling of Indian Economy

The empirical models of the I
ndian economy can be broadly classified into three
categories: plan models, econometric models and the Computable General Equilibrium
(CGE) models. While the models are distinguished by estimation methods, theoretical
underpinnings and time horizon over wh
ich the future view of the economy is provided, the
most important distinguishing feature is the objective of the model. The three approaches
reflect varying objectives either in terms of policy analysis or the need for an assessment of
the performance of
the economy
4
.

The ‘plan’ models aimed to provide estimates of investment requirements to meet the
targeted economic growth rates, the macroeconometric models were essentially meant to
track the evolution of the economy and provide estimates of the impact o
f alternative policy



3

Ichimura and Matsumoto (1993) provide a comprehensive account of the macro econometric models of the
Asian
-
Pacific countries. Indian researchers have participated in

the Project LINK, where models of numerous
economies around the world are modelled.

4

While the policy analysis interest is relatively long standing, commercial interest in forecasts of the macro
environment is more recent.


4

scenarios and the more recent CGE models are meant to provide insights into more detailed
workings of the economy
5
. At the institutional or ‘official’ level, it is only the ‘plan models’
that were developed, maintained and used for pol
icy applications over the years. Surprisingly,
the other two official agencies where the need for model based research inputs is more
obvious, namely, the Central bank (Reserve Bank of India) and the Ministry of Finance, the
initiatives to develop macroeco
nomic models were either indirect or intermittent
6
. In the case
of Planning Commission, the need for model
-
based research inputs may have been greater as
detailed allocation of resources across sectors over time required a rigorous exercise of a
quantitati
ve nature that incorporated not only the physical relationships of production but also
the behavioral relationships such as the consumption pattern of the population. In the case of
finance and the monetary authorities, the need was greater for ‘forecasts’

rather than policy
evaluation as the ‘plan’ priorities were greater than the other policies. In this sense, the
‘plans’ also determined the scope of fiscal and monetary policies.

Besides the models of plan variety, India has had a rich record of research
experience
with macroeconometric modelling. Krishnamurty (2001), in his recent review of the macro
modelling in India, notes that, the ‘environment’ for research on Tinbergen
-
Klein type
macroeconometric modelling may have turned more encouraging in the new

policy regime.
The key to change appears to be the need for research inputs on fiscal and monetary policies
as instruments of policy rather than more detailed management of the economy. The need for
more accurate macroeconomic forecasts of the economy is
also more evident for the business
sector as it is for the policy makers.

Macroeconometric modelling has evolved in India largely in the academic arena
supported directly or indirectly by the official agencies. Marwah (1991) notes that until the
end of 19
80s, about 40 macroeconometric models were estimated for the Indian economy,
almost all of them being the works of individual researchers. Krishnamurty (2001) tracing the
evolution of these models, divides the modelling effort, including the models of the
1990s,
into five broad categories as those from the
first

generation to the
fifth

generation. The first
generation models were the early models of the 1960s and 1970s, which were estimated
under severe data constraints. Consistent data series were availabl
e for only a few years,
starting from the early 1950s and that too for only a limited number of variables. The second



5

We do not provide any discuss
ion of the plan models for India in this paper. A documentation is provided by
Dahiya (1982).

6

Although a there are a number of macro models developed by RBI staffers over the years, there is no official
RBI macro model for India. The Ministry of Finance

supported the macro economic modelling activity at
NCAER during the 1980s and also in the 1990s.


5

generation models were built in the late 1970s and early 1980s and ventured more into policy
analysis than merely attempting to modelling
the economy. With better availability of data,
these models were relatively more detailed than the earlier models. The second half of the
1980s and the 1990s saw more active interest in macroeconometric model building, which
Krishnamurty terms as belongin
g to the
third

generation. The distinction is in the more
detailed coverage of the economy, better estimation techniques and greater focus on policy.
Issues relating to trade, monetary policy and productivity in industry received greater
attention.

The
fou
rth

generation models, under Krishnamurty’s classification, are those
estimated in the 1990s and they begin to reflect the new policy environment. The
fifth

generation models, are those that clearly capture the new policy regime where the prices are
market

determined, role of public sector is limited to a few sectors and monetary policy
becomes independent of the fiscal stance. The above review does not adequately cover the
modelling efforts along the CGE framework
7
. One set of CGE models is macroeconomic b
y
focus, the others ignore the macroeconomic framework and focus on more detailed
specification of sectors, agents and processes. The micro CGE models have successfully
incorporated detailed ‘global’ environment as compared to the macroeconomic models. The

global macro models are not developed in the Indian context.

The developments in macro modelling in India, thus, have followed the policy
evolution as well as development of data and estimation techniques. The paradigm of the
models has become increasingl
y complex from one of simple Keynesian system of
expenditure accounting to set of inter relationships that capture the dynamic linkages between
investment and output, deficits and debt and deregulation and growth. Our attempt in this
paper will be to exami
ne the underlying specification of the working of the economy in the
models that were developed prior to the beginning of the ‘new’ economic policies in the
1990s


III.

Economic Reforms and New Issues for Macroeconomic Modelling

The year 1991 was a watershed i
n India’s economic policies. The balance of payments
crisis of 1991 marked point in history when significant changes in economic policies were
undertaken to tide over the crisis and at the same time to redefine the role of discretionary
economic policy reg
ime. The year saw a beginning of major reforms in the area of foreign



7

Chadha et al (1998) provide a brief discussion of the CGE models for India in their review of the models for
the Indian economy.


6

exchange transactions, industrial policies, fiscal policies, monetary policies and the
international trade policies. The focus of the changes was to give the markets and the private
sect
or a greater role in the allocation of economy’s resources and reduce the administrative
controls on the economy.

The changes also represented a challenge to the modelers of the economy. The need of
the time was to spell out the implications of the policy
changes. The policy changes were
brought about swiftly without the empirical basis for predicting a particular type of response.
In this sense, the policy changes were based on the micro economic reasoning with
considerable evidence on inefficient use of r
esources under a regime of numerous controls on
the economy. The changed policy environment also posed a number of issues to the
macroeconomic modelers.

In order to understand the type of issues which the policy reforms addressed and the
implications of th
e reforms to the functioning of the economy, we present here briefly some
trends in the Indian economy up to the beginning of the 1990s. First we point to the
acceleration in the rate of economic growth during the three decades since the 1950s. There
was c
learly an acceleration in the rate of growth of per capita GDP during the period of the
1990s (Figure 1). India was indeed one of the top 10 economies in the world in terms the
average rate of growth of GDP in the 1990s. While the growth was achieved, ther
e were
growing imbalances in the economy. The ratio of fiscal deficit of the government to GDP
increased to about 10% by 1990
-
91 as compared to less than 5% in 1980
-
81. The increase
was steady and sustained indicating a policy of rising government expendit
ures without a
simultaneous improvement in revenues. The trend in fiscal deficit is also illustrated by the
rise in the level of public debt relative to GDP (Figure 2). The other imbalance in the
economy is reflected in the rising level of current account
deficit relative to GDP. While the
deficit increased, forex reserves decreased by 1990
-
01 (Figure 3).

Thus, acceleration in growth was leading to greater vulnerability of the economy to
internal and external shocks. The internal vulnerability was higher be
cause, government’s
ability to provide resources for development were under stress as the bill for subsidies,
interest payments and support to unprofitable public sector enterprises took larger share of
government revenues (Figure 4). The external vulnerab
ility was greater because, India’s
exports showed no signs of improvement, while the imports rose steadily (Figure 5).

The policy changes, therefore, were meant to address these issues: reduction in fiscal
deficit, improvement in external balances and at t
he same time sustain the growth
momentum. The fiscal reforms meant first a check on expenditures and then restructuring of

7

the tax regime. The changes in the tax regime were related to the overall view of the reforms:
the tariff rate on imports was reduced

over the years from the average rate of collection of
about 60% in 1990
-
91 to the current level of less than 30%. The tax rate on corporate and
personal income was decreased. The domestic indirect taxes were gradually rationalized by
reducing the number o
f rates and the level of average rate of tax.

Improvement in the external balances was sought through a number of initiatives: the
foreign exchange rate was gradually subjected to the pressures of the market for foreign
exchange for the current account tra
nsactions. The Reserve Bank of India now provides only
indicative rates to the market. The actual rates are determined in the market. The impact of
the policy changes in the case of foreign exchange transactions is indicated in the trend
pattern of the exc
hange rate of the rupee. As Figure 6 shows, the rupee/ US dollar exchange
rate showed only relatively fewer changes till 1991
-
92 but since then, the changes are
continuous and gradual.

As noted earlier, trade regime was liberalized: tariff and non
-
tariff b
arriers were reduced
by lowering custom duties and eliminating other restrictions on imports. Discretionary
restrictions on some of the exports were also relaxed, particularly in the case of agricultural
produce. The regime of controls on the inflow of ext
ernal capital was also relaxed. Foreign
investment in the economy was permitted with assurances on the repatriation of profits and
capital. The changes implied greater competition from imports.

The industrial polices saw radical changes. Industry was no lo
nger required to get
‘licenses’ to set up new production capacities for the vast majority of sectors. The areas that
were once the exclusive domain of the public sector were now open to private investors.
Increasing the scale of operations, foreign collabo
rations, changes in the product lines, choice
of imported machinery were all essentially now the decisions that the entrepreneurs could
make freely without the initial approvals needed on administrative basis. Capital markets
were also freed from pricing c
ontrols. The ‘capital’ was valued and ‘priced’ by the market
rather than by the government agencies. The financial sector saw the entry of new investment
enterprises.

Changes in the monetary policy were also significant. Initially, the changes related to t
he
financing of the government’s fiscal deficit. These policies were a combination of monetary
policy reforms and the reforms in the banking sector. The government had access to cheap
financial resources of the banking sector, which the banks were required

to keep in the form
of statutory reserves. These reserve requirements were gradually reduced providing larger
quantum of financial resources in the market: the government was to compete with the other

8

claimants for these resources. The reforms in the mone
tary, fiscal and the financial sectors led
to changes in the manner in which the fiscal deficit of the Central government is financed, the
manner in which the rates of interest on bank loans to the investors are determined, the
interest rate at which the g
overnment would borrow from the market. The general thrust of
these policy changes was to provide greater role for the markets in the allocation of financial
resources. The relatively greater variation in interest rates is illustrated in the case of the
Pr
ime Lending Rate (PLR) of a major investment funding agency, Industrial Development
Bank of India (IDBI) in Figure 7.

Finally, these changes in policies also had an impact on the inflation rate. In Figure 8, we
have presented the pattern of inflation rate
in the Indian economy for the period 1980
-
81 to
2000
-
01. The pattern highlights the drop in the rate of inflation in the 1990s after the
initiation of the economic reforms from the high levels seen immediately prior to the reforms.

These wide ranging chang
es in the policies during the 1990s and the prospects of more
changes now have posed a number of issues for analysis. The macro models of the ‘planning
era’, emphasized the special features of the period. For instance, the planning process, which
visualize
d a particular growth path and composition of growth, led to different pricing
regimes in different sectors. They also led to the predominance of the public sector in a
number of sectors. The economy had, therefore, sectors in which market clearing was
ach
ieved by prices, and sectors in which market clearing was achieved by quantity
adjustments and rationing. In the more liberal policy regime of the 1990s a number of
changes have emerged changing the basic principles under which the economy functioned.
To i
llustrate the point we note the following changes as a result of the policy changes of the
1990s:

1.

The exchange rate of the rupee can no longer be taken as an exogenous variable in the
macro models which was indeed the case in most cases

2.

Interest rate also
is increasingly determined by the market forces

3.

The monetary authorities use open market operations to influence money supply to a
greater extent than before with the development of a market for government securities

4.

Pricing in the manufacturing sector is
influenced by the international prices which act as a
ceiling, and the mark up is now residual rather than a fixed rate

5.

Private investment is influenced by relative returns across sectors rather than allocations
determined by the government



9

The major area
s where changes have taken place relate to the determination of prices,
interest rate and exchange rate. Pandit (1995) in a review of the conceptual framework for the
macroeconometric models for India notes that, those segments of the models dealing with
(a)
capital formation, (b) financial and capital markets, and (c) external markets are most likely
to undergo major changes quickly, (d) price formation and inflation, and (e) monetary and
fiscal linkages and responses are likely to change more slowly and
(f) sectoral productivity
and overall output, and (g) consumption and saving behaviour are likely to change even more
slowly. This view appears to be vindicated broadly by the changes in the economy that have
taken place except for the fact that changes i
n price determination paradigm have been more
rapid than anticipated.

As the purpose of this paper is to focus on the changes in the macro models in
response to the changed policy environment, we will consider the changes that have taken
place in specific
type of interrelationships, rather than look at complete specification of any
particular model. We will review the specification followed generally before the ‘new policy’
regime was set in motion and the changes the macro models have undergone to reflect
such
changes in policy environment.




IV.
Economic Reforms and Changes in Model Specification

Before proceeding to the particular relationships of the macro models, a brief
overview of the macro
-
modelling framework is useful. The key features of the mode
ls are
given by their structure or ‘closure rules’. The structure may also be interpreted in terms of
the agents in the economy (firms, households, government, the central bank and rest of the
world), sectors (agriculture, industry, services), transactions

(consumption expenditures,
investment, taxes, subsidies, imports and exports) and the processes by which goods and
services are produced, sold and consumed. The clusure rules indicate what are the
‘endogenous’ and ‘exogenous’ variables’. The identificatio
n of ‘exogenous’ and
‘endogenous’ variables for the model is a function of not only the use to which the model is
put but also the role of economic policies in influencing the markets.

The level of desegregation of the economy with respect to agents, sect
ors and
transactions is determined by the objectives of the model, availability of data and the critical
distinction between the sub
-
categories in each of these cases.


10

Most macro models of the macro economy in India have followed five
-
sector
classification

of production: agriculture, manufacturing, infrastructure, government services
and the other services. In fact, Patnaik (1995) points out that agriculture has always received
a distinct attention in understanding India’s macro economy. In the same manner,

one can
also point to the unique nature of the infrastructure sector in the Indian economy. Its
dominance by the public sector enterprises makes the sector unique in its response to demand
conditions and pricing rules. The services provided by the governm
ent are subject to different
forces that determine their output levels than the ones supplied by the private sector. Hence,
macro models have incorporated a separate treatment of the services provided by the
government from those provided by the private se
ctor. It is, therefore, the output and price
adjustment mechanisms that have determined the level of desegregation of the production
sectors. To highlight this point, we have summarized the key mechanisms of output and price
adjustments followed in traditi
onally in Indian macro models in
Table 1
.


The output determination is ‘eclectic’ and neither strictly Keynesian nor classical. In
the case of agriculture and infrastructure, output is supply
-
constrained for different reasons.
In the case of manufacturing,

output is often modelled in terms of a production relationship
but, the supply is not ‘upward sloping’. The prices are influenced by ‘administered’ or
‘government
-
determined’ prices. The impact of international price movements as well as that
of exchange
rate is passed on to the domestic prices.


The impact of trade flows on prices is weak and indirect: export growth may influence
prices through their impact on money supply. Higher imports result from higher domestic
prices and higher GDP but these impo
rts affect prices, again through their impact on money
supply.



Table 1. Traditional Specification of Output and Price Adjustments in the Macroeconometric
Models for India

Sector

Output adjustment

Price adjustment

Agriculture

Supply determined; supply
a
function of natural factors
(rainfall) as well as policy
factors (public investment,
fertilizer price, government
-
determined purchase price
of output)

Market clearing but also
influenced by government

determined or ‘administered’
灲icesK
=
janufact畲i湧
=
pu灰py=c潮straine搠 扡se搠潮=
ca灩tal= st潣欬⁲aw=materials=
E摯destic/= im灯pte搩Ⱐ
infrastructure
=
C潳t⁰=us=a灰p潡ch;=als漠
influence搠 批=摥man搠
灲essures=reflecte搠 批=the=rati漠
潦=m潮ey= su灰py= t漠oeal=dam;=

11

administered energy prices;
import prices also play a

role.


Infrastructure

Supply constrained

Administered

Services

Partly supply constrained
(government services), and
partly demand determined

Cost of living index, ratio of
money supply to real GDP




In Table 2, we have summarized the general approach
to modelling capital formation
in the macroeconometric models for India. The specification of capital formation reflects the
structural characteristics of the ‘mixed economy’. The desegregation into sectors and
institutions is common. Public investment is
exogenous in ‘nominal’ value but endogenous in
real value: inflation may erode the value of public investment in real terms. The relationship
between private and public investment is treated as an empirical issue with aspects of both
‘crowding
-
in’ and ‘cro
wding
-
out’. The specification also reflects the ‘credit
-
constrained’
nature of private investment. Beyond this, neo
-
classical factors such as real interest rate and
taxes also find a role. The specification does not reflect the influence of foreign capital

flows
or foreign direct investment, which is a phenomenon of the 1990s.






Table 2. Traditional Specification of Capital Formation in the Macroeconometric Models for
India


Sector

Public sector

Private sector

Agriculture

Exogenous in nominal
value

Det
ermined by public
sector investment, terms of
trade, institutional credit,
agricultural GDP,
household savings

Manufacturing

Not significant

Private savings,
institutional credit, tax
rates, public investment in
infrastructure

Infrastructure

Exogenous i
n nominal
value

Not significant

Services

Exogenous

Residual in nature;
sometimes dependent on
institutional credit



12



The major features of the external accounts are noted in
Table 3
. Although India has a
miniscule share in world trade even today, export
s are often modeled as ‘demand’ equations.
But it may also be a fair assessment to say that export equations tend to take the form of
‘hybrid’ specification that incorporates both the supply and demand factors. In an economy
subject to severe policy constr
aints in terms of high import tariffs, stiff export quotas, non
-
tariff barriers on imports, such eclectic characterization was generally acceptable. However,
exchange rate remained ‘exogenous’. The invisibles account also was frequently exogenous
as the ca
pital account. However, in some macro models, invisibles flows were specified as
demand relationships.


We finally review the specification of the monetary and fiscal relationships in the
macro models of Indian economy. Money supply is modeled as a functio
n of reserve or ‘high
powered’ money. High
-
powered money is a function of monetized deficit of the Central
government and changes in foreign exchange reserves of the Central bank. Changes in money
supply originating from either of these sources affected pr
ices and inflation rate, which in
turn had several channels of transmission of the shocks to other variables in the economy.
Thus, expansionary fiscal stance of the government did not automatically translate into pure
‘Keynesian’ multiplier effect. Some of

the impact was lost in the form of higher inflation
rate. Higher money supply also had a ‘supply side’ effect: bank credit expansion followed the
increase in money supply and led to higher investment. Again, increased public spending had
a ‘crowding
-
in’ e
ffect through monetary channels also, besides the direct ‘crowding
-
in’ effect
if expenditures were in infrastructure sectors.



Table 3. Traditional Specification of External Accounts in the Macroeconometric Models for
India

Item

Relationship

Exports (
merchandise)

World economic activity level, export
UVI/ world prices, Export UVI/domestic
prices; in some cases supply constraints

Exports invisibles

Exchange rate (nominal), World GDP

Imports Petroleum
-
related

GDP

Imports other merchandise

Ratio of UVI

to domestic price, GDP,
tariffs

Imports invisibles

External debt, GDP

Capital flows

Exogenous

Note: Desegregation of trade flows into sectoral level is also followed in a number of macro
econometric models



13

The fiscal and monetary sector specification
generally treated interest rate as an exogenous
variable. The impact variations in global interest rates had no impact on interest rates in the
Indian economy, unless of course interest rates were altered by policy.


The macroeconomic models of the pre
-
199
0s vintage, therefore, did not incorporate
the features that emerged during the 1990s. As noted previously, the new features relate to
the modelling of exchange rate, interest rate, opening up of the economy to freer trade and
investment flows. Some of the

more recent models have attempted to incorporate these
features. These attempts are reviewed below.


IV.

Response of Macro Modelling Research to the Changed Policy Environment

We have selected some of the main areas where the policy changes have significantly

affected the nature of market mechanism by the changes in the economic policies during the
1990s and how the macro models have and can incorporate these changes. In the final sub
-
section here we also briefly note the implications of data availability to t
he macro modelling
research.



IVa. Modelling Prices and Inflation Rate under a relatively more liberalized trade regime

Reduction in tariff and non
-
tariff barriers on trade flows results in greater competition in the
market place. The increased competiti
on may be captured in terms of a price formation
equation:



Pd = WP * (1+ tar) * (1+ dt)* er
-----------------

(1)

Where,


Pd = domestic price

WP = world price

tar = tariff rate on imports

dt = domestic taxes in the form of countervailing duties

er = exc
hange rate


The above specification is a major departure from the previous specifications of prices which
were predominantly determined by domestic factors. The ‘cost plus’ approach now gets

14

transformed into one where the cost plus is a ‘residual’. The wor
ld prices have a far greater
influence on domestic prices. Clearly, the above specification does not where trade is not a
significant part of total transactions as in the case of services. Even in merchandise trade,
non
-
tariff barriers on consumer goods we
re lowered only as late as in the years 2000 and
2001. Therefore, transformation from the ‘cost plus’ appraoch to ‘competitive pricing’
approach is gradual and the models will continue to incorporate the changes gradually. The
short
-
term macroeconomic mode
l for the Indian economy maintained at the National Council
of Applied Economic Research (NCAER) in New Delhi has used the specification of
equation (1) above for intermediates other than fertilizers and petroleum products (POL), and
(2) machinery. In the
case of agriculture, consumer goods, fertilizers and POL, construction
and services, the influence of ‘administered’ prices and the ‘cost
-
plus’ approach is retained.


IVb. Modelling Interest Rate

The Open
-
economy models of the economy adopt some version
of the ‘uncovered interest
parity’ approach, which states a direct relationship between domestic and international
interest rates:


id = iw + E(

% er) +


---------------

(2)


where,

id = domestic interest rate (say, the lending rate)

iw = interest rate i
n the global capital markets

E(

% er) = expected percentage change in exchange rate



= a measure of risk associated with the performance of the economy

If there is interest rate differential that can not be explained by the expected variations in
foreign
exchange rate or the risk factors, capital flows take place to bring the two interest
rates on par again.


The strict form of the equation (2) above is not applicable to Indian conditions as yet. One
attempt at modelling Indian financial markets at NCAER (
Patnaik, Vasudevan and Sharma,
2000) has adopted the approach more common in developing economies, captured in the
Edwards
-
Khan approach:



15

id =


* io + (1
-


) * ic
-------------

(3)


where,


io = interest rate in the open economy framework as given in e
quation (2)

ic= interest rate in the ‘closed economy’ framework (subject to domestic policies) and



= is the weight (between 0 and 1) depending in the ‘openness’ of the economy

The specification still requires us to specify the expectations regarding ex
change rate
variations which is indicated in the next sub
-
section.


We note below the approach to modeling interest rates taken in another recent macro
econometric model for India (IEG_DSE, 1999). The relationships noted below are only an
approximation of

the actual estimates.


PLR = a0 + a1 BR + a2 PLR (
-
1) + a3 BCG
--------------

(4)

WRGS = b0 + b1 BR + b2 (DEF/GDPMP) + b3 WRGS(
-
1)
-------------

(5)


Where

PLR = prime lending rate of the commercial banks

BR= bank rate charged by the Reserve Bank of
India (RBI) on borrowings by the commercial
banks from RBI

BCG = Gross bank credit (total bank credit to the economy)

WRGS = average interest rate on borrowings by the Central government in the financial
market

DEF = fiscal deficit of the Central governmen
t

GDPMP = gross domestic product at market prices

ai’s and bi’s are positive coefficients


The specification does endogenize the interest rates but do not capture the role of global
capital markets. Again, given the limited opening up of the economy in the

financial markets,
the specification will also continue to evolve. Bhattacharya and Aggarwal (2001) adopt
slightly different specification that included the cash reserve requirement of the commercial
banks rather than the gross bank credit in equation (4)
.


16


IVc. Modelling Exchange Rate

The three common approaches to modelling exchange rate are the ‘elasticity
approach’ that focuses on the current account transactions, the ‘purchasing power parity’
approach and the ‘monetary approach’ that links not only t
he differences in inflation rates in
the domestic and foreign markets but also the output and monetary policies with the exchange
rate variations. As the monetary approach in a sense links both the exchange rate and interest
rate determination, we do not d
iscuss it here specifically
8
. The elasticity approach essentially
solves for the equilibrium in the market for foreign exchange in the ‘current account’ taking
capital flows as exogenous. Such an approach is implicit in the short
-
term model developed
at NC
AER (Bhide and Pohit, 1993). The export equations for merchandise trade and
invisibles earnings provide the estimated supply of foreign exchange for a given rate of
exchange. The import equations provide the estimate of demand for foreign exchange for
curr
ent account transactions at a given rate of exchange. Therefore, an exchange rate can be
found that balances the current account.


The above approach does not capture deviations from equilibrium and essentially
does not capture behavioral rigidities in the

demand and supply of foreign exchange even in
the current account transactions. The role of central bank interventions in the foreign
exchange markets as well as the influence of capital flows is completely exogenous. We note
below the ‘purchasing power p
arity’ (PPP) approach (also called relative PPP approach),
which again is only one alternative to the elasticity approach:



% er =

% Pd
-


% Pw
----------------

(6)


Where



% er = percentage change in exchange rate


% Pd = domestic inflation rate


% P
w = Inflation rate in the international economy (major trading partners)





8

The monetary approach lea
ds to the formulation:


% er = (

% M
-


% M
*
) +


(

% Y
-


% Y
*
) +


((id


iw), where M is the money supply, Y is the level
output, the superscript ‘*’ indicates foreign market and all other symbols are as explained previously. This
formulation can be see
n in Rivera
-
Batiz and Rivera
-
Batiz, 1994).


17

Given the inflation rate in the domestic market and in the international economy, the
exchange rate changes are determined. The specification allows for simultaneous
determination of

domestic inflation rate (from the rest of the macro model) and the exchange
rate (equation 6). The strict form of PPP is unlikely to hold for a partially open economy such
as India’s. However, the approach provides an alternative to the elasticity approac
h. The
limitations of exogenous capital account or interventions by the government or the central
bank in many ways persist.


The approach taken in the IEG
-
DSE model for India is noted below:


er = a0


a1 (CAB/GDPMP)


a2 NFE(
-
1) + a3 er(
-
1)


a4 D8191
-
---------

(7)


Where

CAB = current account balance (revenue minus expenditure)

NFE = net foreign exchange assets of the RBI

D8191 = dummy variable distinguishing the pre 1991 period from the subsequent period in
the data

All ai’s are positive and the lags

are indicated by the negative numbers within parentheses
following a variable.


The equation (7) captures the impact of imbalances in the current account as well as the
impact of net capital flows in the previous year. It does at least partly overcome the

limitations noted in the elasticity approach and the PPP approach as it still does not capture
the impact of capital flows in the current year and the interventions of the RBI in the foreign
exchange market. The equation for endogenously determined exchan
ge rate in another recent
macro model by Bhattacharya and Aggarwal (2000) has similar specification as equation (7)
but for the fact that they look at the foreign exchange receipts and payments separately and
include capital account transactions in definin
g payments and receipts. They also model one
component of capital inflow, the foreign direct investment.




18

IVd. Capturing the Impact of Removing Non
-
tariff barriers, Impact on Income
Distribution and Modelling Regional Variations


The issues raised by th
e policy changes of the 1990s are varied and despite their
significance can not yet be addressed by the macro models. We merely refer here to some of
the attempts to assess the impact of the economic reforms on some dimensions of the
economy through econom
y
-
wide models.

A CGE approach (Chadha et al, 1999) that is set in a global framework has attempted
to assess the impact of reduction in non
-
tariff barriers on the economy: inter
-
sectoral re
-
allocation of resources, as relative prices change. The model sim
ulates the impact by an
implicit reduction in ‘tariff equivalent’ of the non
-
tariff barrier. The economy begins to
export those commodities where India has less of a comparative advantage and shifts
resources to those sectors where the comparative advantag
e exists.

Again, a CGE approach that captures the impact of trade liberalization on inter
-
sectoral allocation of resources and the consequent implications to employment pattern and
household incomes has been attempted to provide an assessment of the impact

of selected
policy reforms on income distribution.

The regional variations in output of agriculture as a result of national level policies are
analyzed in the framework of production frontier using a macroeconometric model for India
(Bhide and Kalirajan,
2000). The approach provides for a framework through which regional
level details can be incorporated in a macroeconomic model. While the model presently
desegregates only the agricultural sector, the approach would seem to have potential for
extension to
the other sectors as well. Whether some regions in an economy would grow
consistently faster than the others leading to growing sub
-
regional inequalities in a national
economy is an issue that is concern for the policy makers.



IVe. Data Issues


The empir
ical modelling approach is evolving to address the issues raised by the
recent policy changes. There is, however, an important issue relating to availability of data
for the period of the ‘new policy regime’. In other words, econometric estimation of the
r
elationships based on time series data, typically the case in the macroeconometric models,
becomes difficult for the period of the new policy regime as the data available is only a few
observations under the new regime. The estimation results are subject t
o the criticism that the

19

estimated coefficients are likely to be less ‘robust’. As we noted earlier, Pandit (1995)
suggests that it is realistic to continue to provide estimates of the coefficients, based on
traditional techniques of estimation as the chan
ges are gradual and their impact is also
expected to be gradual. The CGE approach provides an alternative where reliance is not on
time
-
series data alone. However, the CGE approach has provided only the ‘comparative
static’ type of simulations. The econome
tric approaches or even the structuralist CGE
approaches have tended to be used for forecasting as well and hence, improvement in
estimates of the coefficients of the model equations is important.


An additional alternative to the estimation of model param
eters that has been explored
is the use of higher frequency data in conjunction with the usual annual data. In the case of
financial markets this approach holds greater potential than in the other markets or sectors.


V.

Concluding Remarks

The review of macro

modelling approaches in India for the recent period in the
context of the policy changes of the 1990s has implications to attempts at modelling other
economies in transition. Liberalization of the trade flows, domestic markets, financial
markets as well a
s fiscal policies are the experience of the Indian economy in the 1990s.
Similar experiences are shared by the other economies in the developing world, particularly
those who are switching from ‘plan model’ to ‘market model’. The review shows that while
an
alytical models of open economies exist, their actual application will require that the
features peculiar to a specific economy are not overlooked. The analytical framework will
have to be modified to capture such specific features. The review has also bro
ught out the
issue of data constraints in modelling the transition phase of the economy. The exploration of
alternative approaches in the Indian case can be expected to continue as the demand for
analyses of the economic trends increases both for public po
licy as well as decisions in the
business sector. The experience may also be useful for modellers of the other economies.


20

References


Bhattacharya, B.B. and M. Aggarwal (2000), A Macro
-
Econometric Model for Planning and
Policy Analysis in India, Revised
Version, Mimeographed, Development Planning Centre,
Institute of Economic Growth, Delhi, March 2000,


Bhide,S. and S. Pohit (1993), Forecasting and Policy Analysis through a CGE model for
India, Margin, Vol. 25 (3), Aril
-
June, pp. 271
-
92.


Bhide, S. and
K.P. Kalirajan (2000), Incorporating Regional Details in a Macroeconometric
Model for India: An Application of the Production Frontier Approach, Paper Presented at the
Fall Meetings of the LINK Project, Oslo, September 2000.


Chadha, R., Pohit, S., Deardor
ff, A.V. and R.M. stern, (1998), The Impact of Trade and
Domestic Policy Reforms in India, A CGE Modeling Approach, Studies in International
Economics, The University of Michigan Press, Ann Arbor.


Dahiya, S.B. (1982), Development Planning Models, Volumes
I and II, Inter
-
India
Publications, New Delhi.


Dutt, K.A. (1995), Open
-
Economy Macroeconomic Themes for India, in Patnaik, P. (Editor),
Themes in Economics: Macroeconomics, Oxford University Press, Delhi.



Ichimura, S. and Y. Matsumoto (1993), Econometri
c Models of Asian
-
Pacific Countries,
Springer
-
Verlag, New York.


IEG
-
DSE Research Team (1999), Policies for Stability and Growth: Experiments with a
Comprehensive Structural Model for India, Journal of Quantitative Economics, Vol 15, No.
2, July 1999, pp.
25
-
109.


Krishna, K.L., Krishnamurty, K. Pandit, V.N. and P.D. Sharma, (1989) Macroeconomic
Modelling in India: A Selective Review of Recent Research, in Development Papers No.9,
Econometric Modelling and Forecasting in Asia, Economic and Social Commission

for Asia
and Pacific, United Nations, Bangkok.


21


Krishnamurty, K. (2001), Macroeconometric Models for India: Past, Present and Prospects,
Presidential Address, 37
th

Annual Conference of The Indian Econometric Society,
Administrative Staff College, Hyderab
ad.


Krishnamurty, K. (1992), Status of Macroeconometric Modelling in India, mimeo, Institute
of Economic Growth, Delhi.


Marwah, K. (1991), Macroeconometric Modelling of South East Asia: the Case of India, in
Bodkin, R.G., Klein, L.B. and K. Marwah, (edi
tors), A History of Macroeconomic Model
Building, Edward Elgar, UK.



Pandit, V. (1995), Macroeconomic Character of the Indian Economy: Theories, Facts and
Fancies, in Patnaik, P. (Editor), Themes in Economics: Macroeconomics, Oxford University
Press, Delh
i.


Patnaik, I, Vasudevan, D. and R. Sharma (2000), Modelling of Financial Sector: Exchange
Rate and Interest Rates in the Indian Economy, Mimeo, National Council of Applied
Economic Research, New Delhi.


Patnaik, P. (1995), Introduction: Some Indian Theme
s in Macroeconomics, in and Fancies, in
Patnaik, P. (Editor), Themes in Economics: Macroeconomics, Oxford University Press,
Delhi.


Rivera
-
Batiz, F.L. and L.A. Rivera
-
Batiz (1994), International Finance and Open Economy
Macroeconomics, Prentice
-
Hall, Inc,
New Jersey.



22
















23


















24