D-3517-1
362
ENERGY DEVELOPMENT AND ECONOMIC GROWTH
Is INDIA
P. RAGHAVERDRAN |
Indian 011 Coporation Ltd.
Qifen Vang
Shanghai Institute of Mechanical Engineering
: and
Visiting Scholar
Sloan School of Management
Massachusetts Institute of Technology
July 1983
(revised in December 1983)
D3517-1
EWERGY DEVELOPMENT AND ECONOMIC GROWTH
‘IN INDIA
P. RAGHAVERDRAN
Indian 041 Coporation Ltd.
Qifan Wang
Shanghai, Institute of Mechanical Engineering
and
Visiting Scholar
Sloan School of Management
Massachusetts Institute of Technology
ABSTRACT
A dynamic simulation model of the Indian econosy has been developed
which captures the important linkages between econoaic growth and the
development of various foras of energy. Non-commercial forms of energy
vhich supplied the bulk of total energy requirements of the econosy so fer
have clearly reached their saturation lisits. Capital costs for coal and
petroleum increase with resource depletion. The cost of hydroelectricity
increases as the cheaper and more accessible resources are exhausted. The
coste of renewable energy sources such as solar, wind and biomass decrease
with cumulative production due to technical progress. Such sources of
energy become more inportant sources in the future though their current
share of the total energy production is negligible.
The thesia examines the dynamics of the transition to the new ere as
well as responses of the econoay to energy shocks such as steep increases
in international 011 prices. It investigates the possibility of an interia
crisis if the domestic energy industry is slow to develop or if the
response of energy demand to rising energy prices is sluggish. Such a
difficult transition may be marked by persistent import dependence, high
energy prices and high outlays in the energy sectors that reduce the
Tesources available to the non-energy sectors for consumption and grovth.
An aggregate production function utilising capital, labour and energy
as factor inputs has been utilised for the economy along vith a
neo-classical formulation for consumption and savings in the economy. The
model generates the energy demand of the economy endogenously and
incorporates the adaptation of energy intensity to rising real energy
prices through sore efficient new cepitel equipacnt ae well as retrorite of
inefficient equipaen'
The model has been calibrated using Indian data. Where parameters or
assumptions are based on uncertain facts, sonsitivity tests have been
carried out. The effect of goverment policies such as taxation of energy
or emphasis on conservation have been investigated.
ACKNOWLEDGEMENTS: 363
Many colleagues and friends have made valuable contributions to this A
effort. We shall attempt to acknowledge the more important contributions.
Prof. John D. Sterman’s Ph.D. thesis provided the inspiration for this
project. Our colleagues in the System Dynamics Group at M.I.T. provided
valuable support at critical junctures. George Richardson, Bob Eberlein,
Jim Hines and David Kreutser were especially helpful when We developed
blind spots. Bob Eberlein found time to offer useful comments on drafts of
this paper.
1.0 InTRODUCTION
In this paper we describe a model which ie intended to examine the
links between economic growth and energy development. We consider various
forms of energy and the impact of exogenous changes in international of}
prices and technology on the various forms. The model uses Indian date,
but the features represented have general relevance for many developing
countries which are undergoing rapid modernisation and transition from
non-commercial to commercial forms of energy.
Most previous studies of the energy problem in developing countries
have focussed on energy supply and energy demand issues in isolation
(19,25). However, the interactions between supply and demand are very
important in determining the reaction of the systea to exogenous shocks.
The model developed ia a highly endogenous model. It considers issues
of capital investment, and the effect on energy consumption of the type of
capital invested in. The model also allows for modification of existing
capital equipaent for more energy efficient operation through retrofitting.
The rate of investment is constrained by the availability of savings to
meet investment needs.
The model is intended to examine shifts in the relative shares of
different forms of energy, depletion of fossi) fuels and the change in
renewable fuel costs. ‘To accomplish ‘this the model has four commercial
energy production sectors, representing the production of coal, oil,
hydroelectricity and renewables. In addition to commercial energy
production the mode) takes into account the abundant sources of
non-commercial energy currently represent a major proportion of energy
consumption in developing countries.
The emphasis of the study is on understanding the impact of structural
features such as delays in the perception and reaction to events, physical
delays such as long construction times for coal and hydroelectric projects
and to identify forces that may oppose or dilute policy measures.
The macroeconomic effects of the increase in the real price of
imported oil im the 1970's have been examined. The low share of energy in
the national output implies that the long-run effect of even significant
Anereases in the real price of energy on OEP vill be small. Such orises
may take the form of excessive high cost oil imports as well as stagnating
of investment flows into non-energy sectors due to high outlays in the
energy sectors.
‘The model has been utilised to explore the impact of taxes and
subsidies on the development of various forms of energy. Taxes on an
energy source reduce its attractiveness to consumers and hence its share of
total energy demand. This ultimately reduces production of that fora of
energy.
It 49 seen that conservation policies such as information campaigns
and energy audits have a beneficial impact in the short run in reducing the
energy demand during the difficult transition, but in the long run, the
economy would have made the required adjustments even in the absence of
such policies.
‘The model ignores certain potentially significant concepts such as
economic cycles, nominal prices and wages and foreign exchange constrainta
on imports of energy in the short run. The model also ignores some
long-term constraints on growth such as environmental pollution and
scarcity of non-energy resources. Such excluded issues are important, but
are outaide the scope of this study.
2.0 MODEL OVERVIEW
There are five production sectors portrayed in the model, a main
production sector and four energy sectors representing coal, petroloun,
hydroelectricity and renewables. The main production sector generates the
energy demand of the economy which is allocated among the four domestic
energy sectors and imports. Actual energy consumption equals energy denand
with any gap between energy demand and domestic production being met
through imports. The output of the main production sector is allocated
among investment in the five production sectors, exports and domestic
consumption. The consumption (household) sector generates the savings in
the economy which are invested in the fora of capital in the five
production sectors. Figure 1 shows the key flows among the sectors.
2.1 Bxogenous variables: International oil prices are treated as
exogenous in the model. Because of the small role that developing
countries play in determining world demand domestic developments would have
any significant influence on international oil prices. The model also
treats as exogenous the labour force in the economy. While energy prices
364
and availability could influence the employment potential in the econony,
this influence ia weak in a labour-surplus economy. Technical progress is
‘also exogenous. Sensitivity analysis has been carried out on a plausible
range of possible future values of all these exogenous variabl
2.2 Excluded variables: ‘The model deliberately excludes the following
concepts and issues to make the task of model building mansgeable. Since
the purpose was to create a simple model to investigate the links between
economic growth end energy development, their exclusion was considered
justified.
2.2.1 Cycles: The focus of the project is on long-term fundasental
Linkages between energy development and economic growth. Short run’(4-7
year) business cycles, and long waves. (40-60 year cycles) (12) have been
ignored. Short-lived inventories which have been shown to pley a role in
business cycles (15) have also been excluded.
2.2.2 Prices and Wages: All prices and costs in the model are expresned
in constant 1970-71 prices. Strong government control over the price and
wage setting processes in the key sectors of the economy tend to make
prices and costs adjust to real changes in exogenous variables such as
international oil prices or technical progress.
2.2.3 International Trade: ‘The model assumes trade balance at all tines
with the economy exporting enough goods and services to pay for the
imports of oil. That the economy can and will import all the energy
that it needs 1s certainly. a strong assuaption. In the short run,
foreign exchange constraints may develop as & result of the inability of
the economy to divert output from dangerously low consuaption levels to
exports or to develop an export market in pace with its energy need
In the long run, the needed adjustments vould take place.
2.2.4 Non-energy Constrainta: Limits to growth may be imposed on the
economy by environmental pollution or by shortages of non-energy resources
such as fresh water. Pollution in particular may be closely linked to
the rate of energy development and hence could conceivably be
important in the long run. Because the work is intened to deal with energy
ecoconosy interactions these variables have not been included.
3-0 PRODUCTION SECTOR
3.1 Produétion Function
Production in this sector constitutes the national output
The production function for this sector is hence a key
deterainant of overall behavior. This sector uses capital, labour and
energy as factor inputs. The production function portrays the
combination of these factors to produce the output of the economy. While
labour is specified exogenously, capital and energy are endogenously
determined in the model. The model uses a two-level nested CES production
function (see Figure 2) for the long-run potential production: (1)
PP, = npe(eLr, /at)"™* (ex, /10)°E?
-ESP ESP] (~1/8SP)
BK, = 108 [(1-Rvs80)(c,/10)
(1-BSE)/B8B
-- Potential Production
Reference Production
Effective Labour Force
Reference Labour
Effective Capital
Initial Capital
Share of Labour in Output
Share of Capital in Output =,
Share of Energy in Output
Elasticity of Substitution between Energy and Capita
Energy Requirements of Capital
Reference Energy Consuaption
~~ Capital
+ RVSEC(ERC /REC)
‘A Cobb-Douglas formulation implying unit elasticity of substitution
between effective capital and effective labour is used in the equation for
potential production (PP). Empirical investigations have tended to
support near-unit elasticities betveen these inputs. (7) There is
however considerable uncertainty with regard to the elasticity of
substitution betwe capital and energy (ESE), one of the most
important parameters in modelling energy-economy interactions. (8) A CES
formulation has hence been chosen in the equation for effective capital
(EK) so that the sensitivity of results to the elasticity can be
tested.
365
woziint
3.2 Capacity utilization in production.
PRY, PPL Cs,
cu,, = £(DP,,/PP,,) (see Figure 3)
Production in sector
PP Potential Production in sector
DP -- Desired Production in sector
Since it takes time to change potential production, the model
allows for variable utilization of existing capacity in all production
sectors. Capacity utilization is determined, by comparing desired
production with potential production. When desired production (DP)
exceeds potential production (PP), capacity utilization increases due to
use of overtine and extra shifts, but at a diminishing rate because of
constraints on such incresses in capacity utilisation. Similarly when
DP drops below PP, capacity utilisation is reduced, but Jess than
proportionately, reflecting a desire for stability of output.
3.3 Investment Function
t
CAPITAL,= CAPITALS* f (KAR, -KDR,) dt
DR, = ‘CAPITAL 7 ALK
where
‘ALK -- Average Life of Capital
The investment function seeks to capture the pressures,
constraints and decision rules that lead firms in the five production
sectors to adjust capital stocka towards percieved optimal levels. Figure
4 depicts the formulation for the investaent function (25) used in the
model. Capital stock in the sector (CAPITAL) is increased through
acquisitons (KAR) and reduced through discards (KDR).
A thira order exponential delay is assume for construction with a total
construction period of KCT years. Capital construction initiation rate
(KCIR) 10 given by the ratio of the backlog of orders of capital(UOK) to
the delivery delay in the economy (DDG). .
Backlog (UOK) is increased by capital order rate (KKOR) and reduced
by KCIR. .
wR, = DELAY3P(KCIR ,KCT)
KCIR, = UOK,/DDG,
. t
UoK, vor, +f," (xKor,-KorR,) at
Capital Order rate (KKOR) equals the discard rate of capital (KDR)
corrected for the following factor:
a)correction for growth (Xx¢K)
b)correction for the stock under construction (XKUC)
e)correction for the backlog of orders in the supply line (XBK)
4)correction for desired capital (XK).
KKOR, =KDR,*
ByKDR,* f (TXK,)
Tx, = (KOK +XKUC, +KBK, #XK, )/KDy,
KKOR is however constrained to be positive regardless of how negative
the pressure to adjust production (TXK) becomes.
The managers obtain their signals to increase or decrease capital
intensity by comparing the marginal revenue product of capital with ite
marginal cost. Information on marginal productivity can, hovever, be
gained only through experimentation and comparison of operating results
with different factor proportions.
RPC = MPC/McC = (8(PP)/2c )/(IR+(1/ALC))
where
RPC -- Relative Productivity of Capital
- Marginal Productivity of Capital
Marginal Cost of Capital
PP -- Potential Production
Real Interest Rate
ALC -- Average Life of Capital
‘The effect of relative productivity on desired capital as well
desired energy intensity is specified as a non-linear function (NTERPI).
4.0 ENERGY DEMAND
4.1 Desired Energy Consumption
‘The energy consumption of the economy is tied to the existing capital
stock. The model includes a mechanism for. retrofits that permits partial
woovins
upgrading of the energy efficiency of existing capital equipment to the
level of efficiency of new equipment.
DEC, = ERC,* SOU,
ERC, = ERC eft (1ERC, +RERC,-DERC,) at
&
.
BIC,® CDR,
RPFE * BINT, + (1-RPFE) RBIC,
& 8
BS
werc,/ Cy
t
NERC, = NERC) +So (teRc,-NEIC,*cDR,) at
TERC, = BICUC, * CAR,
g
Ey
RUC, /cUC,
g
2
a
.
t
ERCUC, +f f° (BINT,CCIR,-TERC,) at
BINI, = SMOOTH (DEIC, ,NTABII)
DEIC, = BIC, *ERPE,
BIC -- Energy Intensity of Capital
- Potential Energy Intensity Through Retrofit
- Energy Intensity of Capital Under Construction
- Energy Requirements of Capital Under Construction
- Capital Construction Initiation Rate
- Capital Acquisition Rate
- Capital Discard Rate -
STAEII -- Hanofecturera’ Time to Adjust Energy Intensit of New
Equipment
Desired energy einseapiifon in the economy (DEC) is determined by the
capital equipment in use in the economy (C) and increases with capacity
utilization (SCU). The energy requirements of capital (ERC) increase with
new investment (IERC) and are reduced through capital discards (DERC) as
well as retrofits (RERC).
It ie assumed that existing equipment cannot be upgraded
economically to match the energy intensity of new equipment. The
Retrofit Potential RPFE (set to 0.25 in the base case) represents the
fraction of the gap between the exergy intensity of new investeent
(BINI) and the original intensity of existing equipment (NSIC)
that can be economically realised through retrofits. RPFE determines
the ex-post flexibility for changing the energy intensity of existing
capital. If the retrofit potential ie sero, energy intensity can be
Ww
changed only through capital turnover (a putty-clay approach). If the 367
retrofit potential is one, existing capital offers no constraint
to retrofits (a putty-putty approach) though the necessary adjustuents
through retrofits is with a delay. To keep track of NBIC , the original
energy requirement of capital (NERC) are also computed. Figure 5 shows the
formulation for energy requirements of capital used in the model.
4.2 Energy Prices
UP yg BS, 4 / PP,
. * «
RS, , = (19PT, D(C, .#TRe CUC, *IR +0DR,.)
Price Per Unit of Energy
Required Revenues from, Sales
Potential Production
Fractional Advalorem Tax
Capital employed in the sector ‘
Capital Under Construction
Real Interest Rate
CDR -- CAPITAL DISCARD RATE
The model endogenously generates the costs and prices per unit of
energy endogenously. The price of imported energy (PIE) is hovever
specified exogenously.
Capital (C) is the sole factor input in the energy sector and thus
the major determinant of peoductionenergy costs. Energy prices are
regulated by the government on the basis of average costs.
The government can influence the market's reaction to the various
forms of energy through taxes or subsidies. Since the energy sectors are
under government control, the model treats the difference between
market prices and the cost of capital as an effective tax or subsidy.
DPE, ¢ BSiy +PIE,*IMP,) ‘/08Cq IF IMP)
= ( BS), )/(DPRytECEP,) IF IMPKO
Were IT Ry HCEP
PIR -- Price of Imported Energy
IMP -- Quantity of Imports
DEC -- Desired Energy Consuaption
DPR Domestic Production of Energy
CEP Production of Non-Commercial Energy
BS Revenues from Sales
The Domestic Price of Energy (DPE) is the marginal cost of energy
faced by the main production sector. It deteraines the productivity of
energy and hence is crucial for the factor balancing mechanism. The
formulation adopted reflects the current government policy of pooling
domestic costs and international prices to obtain an average price of
energy (10). If the economy were to become self-sufficient however,
the domestic price of energy would become the weighted average cost of
total domestic production.
4.3 Honcommercial Energy Production
Honcommercial sources of energy such as firewood, agricultural waste
and animal dung have in the past been the dominant sources of eneray
in India.
Table 1:Production of commercial & non-commercial Energy
(4m million tons of coal equivalent)
Year Commercial © ‘oncommercial Share of noncommercial.
1953-54 41.30 133.63 0.76
1960-61 64.60 155.18 0.71
1965-66 85.60 170.28 0.67
1970-71 99.60 183-70 0.65
1975-76 134.90 207.44 0.61
(derived from Tables 2.5 & 11.3 of the Report of Working Group)
From Table 1, it is however seen that while commercial energy grew at
5-4% p-a., non-commercial energy grew only at 2% p.a.Total energy
consumption grew at the rate of about 3% p.a. The total energy-CDP
elasticity is thus just 0.82 which compares favourably with comparable data
for developed nations. (18,19) This provides # logical explanation for
the high commercial energy-GDP elasticity observed in developing countries
in a period of transition from non-commercial to commercial forus of
energy. Production of non-commercial forms of eneray is specified
‘exogenously in the model.
5.0 PRODUCTION IN THE ENERGY SECTOR
Potential production in the energy sectors is determined by the
capital stock of each sector and the productivity of that stock.
5-1 Coal and Petroleum
The resource bases of coal and of) are represented by the sane
structures with different parameters. Reserves (RES) in these sectors in
the model are the ultimate geological recoverable reserves. They include
not only proven reserves but also reserves expected to be
estabilished in future with furthur exploration. (27) Accordingly capital
in these sectors include investments in exploration and development
as well as production.
- RPC, #1
PPyg 7 NPEK, Q°RPC, (Cy
. af *
RES,,* RES, =f," PR, ,at
RPC, ,~ f (RES, ,)
where
FP,, -- Potential Production ef i (coal/petroleum)
RPC -- Relative Productivity of Capital
RES -- Reserves
PR Production
NPEK -- Productivity of Energy Capital (27)
f == a functional relationship, which is linear in this sector
The Linear functional relationship above implies that as reserv.
are depleted marginal cost of production rises gently at first. and then
at an increasing rate as the industry is forced to exploit the costlier
deposits.
5.2 Hydroelectric Production
of?
HPP, = HPP,* {.* HPAR, at
HPAR © (HCAR,. HCDR,)* HRPC,* HRPEK,
Hydroelectric Potentisl Production
Hydroelectric Potential Acquisition Rate
Hydroelectric Relative Productivity of Capital
Hydroelectric Capital Acquieition Rate
HODR -- liydroelectric Capital Discard Rate
HNPEK-- Normal productivity of Hydroelectric Investwent (27)
‘Au the more economical sites are exploited, the cost of
hydroslectricity will increase. One can conceive of a theoretical
potential for hydroelectric generation based on the runoff over all
heads without taking into consideration the economic viability of the
projects.
368
woo
5.3 Production of Renewable Energy
RPP, = RPP, +{o° (RPAR,-RPLR,) at
RPAR,* RBCAR,® ‘RNPEK)* RRPC,
RRPC,= (acunp,/acune,) PP
RCUMP,= RCUMP,*{," RPR, at
RPLR,= RCDR,*RPP,/RC,
where
RPP -- Renewable Potential Production
RPAR --- Renewable Potential Acquisition Rate
Renet Potential Loping Rate
Renewable Capital Acquisition Rate
Productivity of Energy Capital.in starting year
Renewable Relative Productivity of Capital
Renewable Cumulative Production
RPR -- Renewable Production Rate
RCDR -- Renevable Capital Discard Rate
RC -- Capital in Renewables Sector
PEXP -- Progress Blasticity Coefficient
‘Though. Renewable sources of energy such as solar, wind and bicasss
currently supply a negligible share of the energy demand of the
econony, they could conceivably become important in the future as the
costs of depleting resources such as coal and. petroleum rise and as
costs for Renewable technologies decrease with time due to the learning
effect. The Productivity of capitel in this sector is expected to
increase with cumulative production due to economies of scale in
development and manufacture and rapid diffusion of innovations. A
progress function is used to capture thie effect in the mode) similar
to Hirsch’s Progress function for labour (11).
6.0 CONSUMPYION OF OUTPUT
6.1 Domestic Consuaption and Savings
DINC,* PR,-NEIP,
CONS,= SINC, -SAV,
SINC = SMOOTH(DINC, ,TAT) (es POr,*TAI)
SAV, = TDR,* POI, 7,» (((DSC#STEC )-70,)/m8)
AG, = CONS, + NEXP,
0G, = AC,* Por,*v0G,+ ((DDDeAG,)-UoG,)/TAG
t
wos, = voc, +f\* (oc,-aG,) at
where
TAI -- Time to Average Income
psc Desired Savings Coverage (years)
TAS -- Time to Adjust Saving
Consumption of output in the model is based on standard
cro-economic theory. The formulation includes the formation of income
expectations, as in the permanent income hypothesis, as well as the
influence of wealth in the determination of consumption, as suggested by
the life-cycle consumption theory. Output in the economy (PR) is
partially spent abroad (WEXP) to pay for imports of oil. The balance
constitutes disposable income (DIEC). Consumers spend a portion of
their income (CONS) and save the rest (SAV), Hormal savings cover
depreciation of all capital equipment (TDR) as well as new investment
needed to maintain growth of capital at the perceived growth rate in
income. (PGI)
Te total capital stock in the five production sectors constitutes
the wealth of the economy (TC). Correction for any deviation of
actual savings from desired wealth is hence assumed to be made over a
period TAS (assumed to be 25 years) through additional savings. During
growth, orders for goods (0G) placed by the domestic consumers as well
‘as external importers are higher than the actual consuaption and exports
by an amount needed to sustain the backlog of orders (U0G) at its
optimal value so that the delivery delay of goods (DDG) is at its
desired level (DDG). In a period of growth, the backlog will have to rise
at the growth rate to maintain delivery delay at normal value.
7-0 CALIBRATION OF THE MODEL
7.1 Key Parameters
The model has been calibrated using Indian data. The model ie
initialised so as to start in equilibrium in the year 1900. It takes a few
years for the model to reach steady-state growth conditions. Calibration
seeks to make important model variables such as Production (PR), Desired
Energy Consumption (DEC) and production of Coal (CPR), Petroleum (PPR) and
369
Hydroelectricity (HPR) close to their actual values in recent years.
Tables 3 and 4 in Appendix 1 provide values of the paramoters used. in the
base run of the model along with sources and referenc
7.2 Analysis of the Base Run
Figure 6 and Table 2 show the results of simulation using base case
parameters till the year 2050. The base case provides the standard
against which changes in parameters and key policies are
investigated.
Economic Growth in the model is generated by growth of the labour
force (2.5% per annum) and labour augumenting technical progress (2% per
annum). The resulting growth in the national output (PR) is approximately
4.5% per annum in the absence of changes in energy prices. The real cost
of capital is constant in the model: while the average cost of energy
increases. The cost of oi] and coal rise due to depletion and the cost of
hydroelectricity rises due to the increasing scale of operations. Though
the real cost of renewable energy falle due to technical progress, the
share of renewable energy in total energy production is small. The decline
in the productivity of energy due to the increase in average coste causes
the mix of capital, labor and energy to be rebalanced, slowing the rate of
economic growth as capital and labour are substituted for energy. As
Production of non-commercial energy is assumed to be constant beyond 1980,
growth in commercial energy demand and domestic energy production vill
however be higher than growth in total energy demand.
The level of imports over the years as shown in Table 2 is
interesting and illustrates some of the important mechanieas within the
model. ‘The steep five-fold increase in the price of of] in the 1970's
leads to a drop in the level of imports to near-rero levele during the
1980's. After some subsequent fluctuations, imports eventually grow
faster than coal, petroleum and hydroelectricity (see Figure 6). The
merease in the price of imported of] in the 1970's reduces the
attractiveness of imported ofl and causes an increase in demand for the
domestic energy sources. However, long construction times and
perception delays prevent an immediate increase in domestic production.
Imports which are the residual source of energy in the model hence continue
at @ high level for nearly a decade before domestic production
increases sufficiently to reduce imports Excessive high-cost imports
in the interim period reduces the productivity of energy. Substitution and
conservation now reduce the growth in total energy demand (see Figure 7).
After 1985, energy demand grows again as conservation opportunities are
exhausted.
‘The base case assumes rero growth in thé real price of imported oi]
beyond. 1980. The costs of coal, domestic petroleum and hydroelectricity,
however, increase with time due to depletion and scale effects. Inports
or than domestic sources, and becomes a significant
are forced to grow ft
source of energy once again beyond 1990.
Figure 8 shows that the oil price increases do not have any
perceptible effect on national output or consumption. In response
to the increase in average cost of energy in the 1970's, (see Figure 9),
the energy efficiency of the economy {measured by the ratio of onergy
consumption to national output) improves through retrofits as vell as
replacement of capital stock by more efficient equipment. This
compensates partially for the increase in the cost of energy. The higher
cost of energy also leads to substitution of energy by capital. Beyond
1980, increase in the production of the lower cost domestic energy
leads to a decrease in the average cost of energy. By 1985, imports
have been adequately reduced and the average cost of energy increases
slowly thereafter due to conl and oil depletion and the increase
in the cost of hydroelectricity. The overall impact on national
output and consumption of the steep increase in oil prices during the
1970's i8 hence negligible.
The high import of] prices however lead to high investments in the
As shown in Figure 10, during the transition,
domestic energy sectors.
tment in
energy investments become a significant fraction of the total im
Diversion of capital from the non-energy sectors of the
the economy.
economy reduces potential growth in the economy.
This example illustrates how the mode) captures some of the important
mechanions that are at work in the economy and the utility of the
model in tracing the major transmission channels. It also illustrates an
important issue regarding energy price shocks. Since the share of energy
in the national output is small, the effect of an increase in the real
price of energy will be small in equilibrium. However, during the
370. transition, the ;
. economy may need’ s:
shares and factor balances, ignificant ad.
in a disequilibrius
framework such as ha;
© been used in
‘Tal i
ble 2: Base Run Simulation Results
cons cop
gous DEC kor y
9 #09, 706 B15 ets ets DPE DODP ace
Me ez da PAE 190 aa
BE os ae EHS ak lee os 18
$12 214 185 429 fare 78?
. 6 Las
2157 9920 42
3372 16131 gag 31°94 17-96 2012 45 68
30.69 26. 817.4 Bs
5261 25750 rine Sse aa 22.28 65 ae
~- Production ( in bi;
Lion
Gonsusption (in bi21t0n rapeent
ip 1, fapttel (in dilaton rupess
bour Force (40 millions)
tion of Ene:
~ Energy-Net Nati, rey (4n quads)
ace __ onal Prodi °
CE -- Average Cost of Boeray (rope taste Coop 2r0/rapee)
peee/ai2lion Bry;
8.0 PoLIcr ANALYSIS ap CORCLUSIONS
1ly anticipated. For
of
eimilar and generally sore complex compensating mechanioms are
typical for most policy interventions, some policies have higher
leverage than others.
The effects of taxes and subsidies as well as emphasis on conservation
have been assessed with the use of the model. -
8.1 Taxes and Subsidies
Taxes on an energy source reduce its attractiveness to the consumers
thereby reducing its share of totel energy demand. This ultimately
reduces production of that for of energy. By increasing the average cost
of energy and hence reducing the relative productivity of energy, ® tax
also reduces total energy demand. Figure 11 shows the effect of a 50%
advalorem tax on petroleum coupled with a 50% subsidy on renewable
energy. As expected, petroleum development is slowed and renewable
energy develops faster than in the base case.
In the initial stages when the share of petroleum is high compared
to production of renewables, such a policy creates a net inflow into the
government treasury. However when renevable energy production becomes
large enough, continuation of this policy will be constrained by the
government budget. A more realistic representation would involve a
gradual reduction of subsidies for renewable energy over # period of
time, eg. subsidy of renewable energy linked to tax revenues from
petroleum. Policies such as these could easily be investigated if taxes
and subsidies are treated as variables in the aodel.
Figure 12 (see section 8.2) shows the effect of the above policy on
the average cost of energy (ACE) and the energy-NNP ratio (ratio of
energy demand to net national product at factor cost). The average
cost of energy tends to reduce due to the direct effect of subsidies
on renewable energy as well as due to reduction in unit costs
ciated with the accelerated technical progress in this sector. This is
compensated fully until about year 2020 by the increases in the cost of
petroleum due to the taxes. The greater production of renewables in later
years compared to the production of petroleum leads to a reduction of
average energy costs compared to the base cai
8.2 Emphasis on Conservation
A natural response of the Government to energy price shocks as in the
1970's ie to emphasise conservation and substitution. High energy prices
provide a natural incentive to the econoay to reduce energy demand.
Governments could attempt to accelerate this effect through taxes which
would, however, have undesirable political consequences. In fact, the usual
response is to reduce taxes and increase subsidies. In this context,
mpheis on conservation is seen as a desirable alternative to control
energy 4d nd »
The model has been used to investigate conservation policies that
increase the speed of response of the econony to energy price increases.
The time to adjust energy intensity of new investment (NTABII) equals tvo
years in the base case. The time to adjust Retrofit Potential (RATE)
equals four years. Information campaigns that accelerate consuaer deaand
for more efficient equipment would put pressure on manufacturere to
improve energy efficiency of new equipaent at a faster pace and reduce
NTABII. Energy audits, information campaigns and tex credits on
conservation investments would reduce RATE, though there are clearly
limits to the extent to which the reaction times could be reduced.
Figure 13 shows the effect of reducing NTABII an RATE to 1 year as
also increasing thea to 4 years and 5 years respectively. The response
to the ofl price increase of the 70's is compared with the base case for
both scenarios. It is seen that conservation policies have a beneficial
Ampact in the short run in reducing energy demand during the difficult
energy transition. The effects vanish beyond 1990 unless the cost of
energy rises again leading to additional ‘opportunities for improvesent
in energy efficiency.
8.3 Conclusion
1) Te model in its current stage of development can shed light on
the response of the economy to energy shocks such as steep price
increaser In particular, it can identify the transmission channels
that are responsible for the oversll effects.
-2) The decline in the rate of growth of non-commercial energy
accounts completely for the high energy-elasticity coefficient
observed in India compared to developed countries.
a
3) The impact of energy shocks such as oil price increases is seen
to be essentially a medium tera phenomenon. The transition in
response to such a major shock as for the five fold real increase in
prices in the 1970's may last 20-25 years. ‘The long-term impact of
auch shocke will however be relatively small as the economy adjusts
to the new prices through appropriate factor balancing and
conservation investments.
4) Taxes can have a significant impact in reducing the development of
any form of energy. Subsidies accelerate such development but are limited
by government budget constraints.
5) Conservation policies can be very effective within limited
bounds during the difficult transition period following an energy price
shock in modulating energy demand. Their impact in the long-run is however
small unless there are further aajor increases in the cost of energy.
Development of the model in the following directions is likely to
be useful for policy makers and analys'
1) axes and subsidies should be treated as variables in the model to
enable exploration of policies such as subsidies on renevable energy
linked to tax revenues on petroleum.
2) Future research to determine values of elasticity of the
substitution of energy. and technical progress in Renewable energy
would be useful. While the principal model results do not depend on
their specific values, precision would be necessary to evaluate the
numerical impacts of different policies.
3) The model is not currently designed to explore issue
such as the impacts of foreign exchange shortages, economic cycles,
and long-term constraints such as environmental pollution and
ecarcity of non-energy resources on energy development. All these are
fruitful areas for furthur development of the model.
In its current stage of development, however, the model
incorporates widely held assumptions with regard to the key
relationships among energy and macroeconomic variables (derived largely
from Working Group, November 1979). It provides a framework to test
the impact of key policy issues such as emphasis on conservation,
influences of foseil fuel resources on future energy costs, and the role of
taxes and subsidies in regulating energy development along desired: lines.
372
22
(1) Arrow, K.-S. et al., “Capital-labour substitution an Economic
Efficiency,” Review of Economics and Statistics, 43,3, August 1961,
225-25
(2) Backus, George et al., FOSSIL79: Documentation (3 Vols.) DSD
166.Hanover HH: Resource Policy Center, 1979
(3) Berndt, Brost R. and Wood, David 0., “Engineering and Econometric
Interpretations of Energy-Capita Complementarity”, American
Economic Review,69, 3, June 1979
(4) Berndt, Ernst R. and Wood, David 0., “Technology, Prices and the
Derived Demand for Energy,” Revie of Economics and Statistics,57,3,
August 1975, 259-268
(5) Carasso, Meir et al., The Energy Supply Planning Model, 2
»PB-245,382, San Francisco, CA: Bechte Corporation, August
(6) The Energy Decade 1970-1980 A Statistical and Graphic Chronicle,
Ballinger Publishing Company (1962)
Cambridge, MA: MIT
(7) Forrester, Jay W., Principles of Syst
Press, 1968
(8) Forrester, Jay W-, Industrial Dynamics, Cambridge, MA: MIT Preso,
1961 >
(9) Friedman, Milton, A Theory of the Consumption Function. Princeton :
Princeton University Press, 1957
(10) Goodman, Michael R., Study Hotes in Syste Dynamics, The MIT Pres
(1980) :
(11) Hirech, Werner Z., “ Manufacturing Progress Functions”, Review of
Economics and Statistics, May 1952
(12) Kondratiev, ¥.D., "The Long Waves in Economi Life,” Review of
Economic Statistics, 17, November 1935, 105-115
A Statistical Outline of
(13) Kulkarni, V.G., Decennial Census Statistics
Indian Economy (1968)
+ (14) Marcuse, W. et al., A Dynamic Time Dependent Model for the Analysis of
Alternative Energy Policies. BNL 19406, Upton NY: Brookhave
Rational Laboratory
(15) Mass, Rathaniel, Economic Cycles : An Analysis of Underlying Causes,
Cambridge, MA: MIT Press, 1975
(16) Modigliani, The Life Cycle Hypothesis of Savings: Volum 2 of the
Collected Papers of Franco Modigliani, KIT Press, 1981
2
Figure 1: Key Flows between Sectors
373
(17) Monthly Abstracts of Statistics, Central Statistical Organisation,
(Government of India) September 1982.
(18) 011 Prices Comittee (1976) Report of the 011 Prices Committ.
Government of India, Ministry of Petroleum (November 1976:
(19) Pachauri, B.K., Energy and Economic Developaen in India : Praeger 1977
(20) Petroleum Statietics (1975) Indian Petroleum and Petrochemical
ENERGY SECTORS ¢
COAL OIL, HYDRO %
RENEWABLES
ENERGY DEMAND
ALLOCATOR
(21) Statistics, Government of India, Ministry of Petroleum, (1975)
(22) Richardson, George and Pugh III, Alexander, System Dynamics
Modelling with Dynamo, Cambridge MA: MIT Prese, 1981
(23) Senge, Peter M., The System Dynamics National Model Investment
Function : A comparison to the Neo-classica Investment
Function. Ph.D. Dissertation, A.P. Sloan School of Managenent,
MIT, Cambridge, MA: 1978
(24) Sterman, John D., The Buergy Transition and the Econony: A System
Dynamics Approach Ph.D. Dissertation, A.P Sloan Schoo) of
Management, MIT, Cambridge, MA:1981. DEMAND FOR cniuentitay
(25) Tyner, W.E., Energy Resources and Economic Development in India, IMPORTED OIL 5 *
Martinus Wijhoff Social Sciences Division, 1978 “ s
(26) United Mations (1976) World Energy Supplies (1950-1974) United
Hations (1976)
SAVINGS FOR
(27) Working Group (November 1979) Report of the Working Group on Energy . INVESTMENT
Policy, Government of India, Planning Commission, Ne Delhi (1979)
(28) Working Group (February 1979) Interim Report of the Working Group
on Energy Policy, Government of India, Plannin Commission,
(reproduced as Annex II in the November 1979 report
(29) VI Pive Year Plan Government of India, Planning Commission (1981) .
CONSUMPTION
SECTOR
FOREIGN TRADE
SECTOR
S565
Figure 2 : Potential Production
(adapted with permission from Sterman 1981)
BE
8
‘SCHCOULED canaciry UTILILATION
APRAC TICE OF HORMAL
20)
Cr ar a a a a eT)
DESIRED PRODUCTION
PEIENTIAL PROBUSTION
Figure 3: Scheduled Capacity Utilisation
(adapted with permission from Sterman 1981)
374
Figure 4 : Capital.and Investment Function
(adapted with permission from Sterman 1981)
LS
conntcion
EAT
Teepe”
“
(spenb)
Sq4odwy 9 uo}y3npory ‘puewag *61903
375
: Energy Requirements of Capital
(adapted with permission from Sterman 1981)
Figure 5
: Energy Shares in the base run
Figure 6
t
'
3
is 3
; 3
18 :
1B &
5
i ns a a
iz" Er
13 28
18 3
1 o
BAe is
Figure 7
ays we UGE
(spenb)
S]4odwy pue uo1QInpory ‘puewag Ab1au7
tcxssy a ov 6
TOTAL INVESTMENT
Figure 10 : Investment in Energy and Main production
002 ast e00r
(4eak/"sy YOELLNG ) LN3WESBANT
en ie
WITHOUT PRICE (SHO:
(sy uoptLya) vopadunsu0g pur’ uN
HH SUBSIDY
RENEWABLES WIT!
NEWABLES NOi
re
- 02
{(savnv)
STUNVAINTY QNV THO 40 NOTLINGOWL
a!
g:
2
hn we
Pe
an ‘
8!
Se, 2 !
3 =
/ £
3; 8
S$ 4
3 = 3
i -2 - 2
a
1 3 be
& B
¢ & 8
a %
8 we
Bs ere
3: ee
2 ten
3 LES
2 PS
©
a
§
3
&
2
=
2
=. x= ly 2 ==
= = 2 2
+ low a a )
(aadny /Ntd 0001) 98324 aNN AisauR
(N10 VoELiym/sy } AB4aUd yo 4s09 aheraay
Pigure 9 : Energy - NHP ratio’
Figure 11 : Effect of 50% tax and on oil and 50%
subsity for renewable energy
AVERAGE COST OF EHERGY ( Rs. /mittion btu)
ENERGY - NNP RATIO ( 000 OTU/rupee)
ENERGY CONSUMPTION ( quadritiion BTUs)
Figure 12 : Effect of 50% tax and on oi] and 50%
subsity for reneweble energy
09K ‘9000.8
WOE
WF.- - 2-2 2
é :
Figure 13 : Conservation Policies
377
3517-1
32
Appendix 1.
Table 3 : National Parameters
+ Re.130 Billion
2% per annum
110 million
Reference Production (NKP at factor cost)
Future Technical Progress
Reference Labour (13)
Puture growth in Labour Force (13)
Reference Value Share of Energy (26)
Elasticity of Substitution of Energy
Real Interest Rate
Time to Perceive Relative Productivity (sec 3.1)
Time to Adjust Goods
Time to Average Income (9)
‘Time to Adjust Savings (16)
Average Life of Capital (24)
Capital Construction Time
Time to Adjust Capital (23)
Time to Adjust Backlog
Normal Delivery Delay
Time to average orders (25)
Table 4 : Key Energy Parameters
COAL = OIL
Initial Reserves (acts) (zn) 2200 260,
Normal Productivity of (27 160000 34500
Capital (BIU/Rupee
life of capital (years) 15 15
Construction Time(years)(27,28) 5 2
Convenience Factor » 0.3 1.0
Fractional Taxation (18) 0.073 0.15
‘Time to adjust Energy Intensity
of new equipment (STABII)
Retrofit Potential Fraction
Retrofit Adjustment Tine
Import Convenience Factor(ICF)
Allocation Weight Factor (AWF)
Reference Capital in Hydroelectricity (27)
‘Time to Adjust Shares (TASHARE)
Renewable Cuaulative Production in 1900
(section 5.3) 23
2 0.35
10
°
+ 10 years
Renewable Progress Elasticity (PEXP)
Future Growth in non-commercial energy (27,28)
Future growth in international of] prices .
(section 4.2)
