702 A. Galinis and M. J. van Leeuwen
fuel oil. This shows the orimulsion burning at the Marbach power plant in Germany test data Stoubler and Maier, 1993. Therefore,
additional measures for reduction of emissions are required. The INPP plays a crucial role for the future of the Lithuanian
power sector. But safe operation of the INPP remains an important issue. The State Nuclear Safety Inspectorate VATESI in 1993
approved the plants’ Safety Improvement Programme. To support implementation of this program 13 Western European countries,
through the EBRD, allocated 33 million ECU for this purpose. The number of unplanned reactor shutdowns in 1990–95 was 28,
and number of nuclear events in IAEA scale was 8, of which 7 were according to scale 1 and 1 according to scale 2.
There was the National Environment Protection Strategy pre- pared and approved in 1995. Lithuania also signed several interna-
tional conventions: Geneva Convention on long-range Trans- boundary Air Pollution, Vienna Convention, on Liability for
Nuclear Damage Nonproliferation Treaty, Safeguards Agree- ment, Joint Protocol, Law on Liability, Convention on Protection
of Nuclear Material, etc.
2. GENERAL EQUILIBRIUM MODELING
The starting point for the CGE-LI was a “Computable Equilib- rium Model for Poland” see Hille, 1993, Van Leeuwen, 1997
and the “Bergman Computable General Equilibrium Model” see Bergman, 1990.
The CGE-LI is a static computable general equilibrium model of an open economy. In some respects it differs from the standard
CGE modeling. First, it includes emissions and emission control activities, as well as markets, and market prices, for tradable emis-
sion permits. Second, some of the tradable-producing sectors are taken as price takers on international markets in the standard
Heckler-Ohlin fashion,
2
while others to some extent are price makers in the accordance with the so-called Armington assump-
tion.
3
The sector classification is adjusted for the Lithuanian situation, and for the analytical purpose of the model. For the same reason
2
In Heckler-Ohlin models it is assumed that products are homogeneous across countries.
3
The Armington assumption Armington, 1969 treats similar products produced in different countries as different goods. This formulation also makes it possible that a country
is both importing and exporting the same goods.
A CGE MODEL FOR LITHUANIA 703
a division is made between government and family consumption. In Figure 4 an overview is given of the model. The different parts
of the model are explained in more detail in the next section.
3. SPECIFICATION OF THE LITHUANIAN MODEL
In this section we present a general overview of the CGE model for Lithuania. Only the most essential elements of the model are
presented and explained. Equations are only given when they have a positive explanation value.
4
The model consists of four major blocks: production, consump- tion, foreign trade, and environment. The energy sectors and the
production and consumption and trade of energy products are intertwined with all the blocks of the model. The main inputs and
outputs of the model are an aggregated input–output table io table. The model uses the information from the io table of 1994
see the Appendix as starting point and generates, based on the accounted equilibrium situation, an io table for the projection
year and detailed energy, and environmental indicators.
3A. Sector Classification
The sector classification of the Lithuanian CGE model was made by aggregation of different economy branches into 15
groups. Ten sectors were specified for the production sector of the Lithuanian economy and five sectors for energy. The sector
classification used in the Lithuanian CGE model is presented in Table 5.
3B. Production
The production block forms the core of the model, and is related to all the other blocks. The production block is fed with informa-
tion from the input–output table, and with information per sector and the costs of emission abatement. The technology of production
is represented by a nested constant elasticity of substitution CES and Leontief production function in each production sector. The
structure of the production function is the same for all sectors, but the elasticities of substitution between various inputs interme-
diate inputs, labor, capital, energy may differ across sectors. This
4
For a detailed description of the equations and structure of the model, see Bergman 1990, Hille et al. 1993, and Van Leeuwen 1997.
704 A. Galinis and M. J. van Leeuwen
Figure 4. Diagrammatic overview of the CGE model.
A CGE
MODEL FOR
LITHUANIA
705
Table 5: The Sector Classification of the Lithuanian CGE Model
Model code Economy branches
Share in GDP
T1 Tobacco industry
0.5 T2
Chemical industry 2.0
M1 Agriculture and forestry, other branches
13.0 M2
Food industry 12.0
M3 Metallurgy, plastics, equipment, paper industry, furniture, clothing, etc.
18.3 M4
Building materials, construction 6.6
M5 Trade, restaurants
15.9 M6
Commercial services 15.9
M7 Transport and communication
14.6 N
Public services 1.2
Total nonenergy branches 100
E1 Electroenergetics conventional, Heat
15.0 E2
Gas industry 11.8
E3 Coal industry
1.5 E4
Refinery 49.2
E5 Electroenergetics nuclear
22.5 Total energy branches
100 Source: LEISEO and Lithuanian Department of Statistics 1995.
706 A. Galinis and M. J. van Leeuwen
part of the Lithuanian CGE model was significantly changed with the Polish version. A change related to new energy carriers was
also introduced in the Lithuanian model. Because significant over- capacity in the Lithuanian power system is one of the peculiarities
of the Lithuanian economy—domestic electricity demand can be covered either from nuclear or fossil fuel—electricity generation
in the model is split into conventional and nuclear electricity. Furthermore, natural gas is introduced. Because of those changes
the mixed CES-Leontief production structure was extended.
Obviously, electricity generated from fossil fuel or from nuclear fuel is for consumers with exactly the same product, only the
prices are different. An approach of weighted price calculation is used to determine the price of composite from both electricity
types:
P
e
5 t
1
P
1
1 t
2
P
2
, 1
where P
e
is the price of composite from electricity generated from fossil fuel and nuclear fuel;
P
1
is the price of electricity generated from fossil fuel; P
2
is the price of electricity generated from nuclear fuel; and t
1
and t
2
are the share of electricity produced from fossil and nuclear fuel.
In fact, the price of electricity in the model is calculated using a general specification of the implicit price equation:
P
fx ,j
5 d
j
P
1
2s
j
f 1
1 1 2 d
j
P
1
2s
j
f 2
1 1
2s
j
, 2
where j is the index of production sectors; P
fxj
is the implicit factor price composite of factor 1 and factor 2 of sector j;
P
fi
is the price of production factor P
i
; s
j
is the price elasticity of substitution between factor 1 and 2 of sector j; and
d
j
is the distribution parameter of sector j. Price elasticity of substitution between electricity generated
from fossil fuel and nuclear fuel in this case is set equal to zero. In other words, s
j
5 0 converts Equation 2 to Equation 1. From
the weighted electricity price and the price of coal the implicit price of composite electricity–coal is determined. Next, the price
of fuel is added and the implicit price of electricity–coal–fuel is calculated. The total price of energy is determined by adding the
price of gas that takes a significant share in the Lithuanian energy
A CGE MODEL FOR LITHUANIA 707
balance. Then, the price of capital and later the price of labor are added to calculate the composite energy–capital–labor price. In
a final step, the composite factor input is combined with intermedi- ate sector input, using fixed Leontief coefficients based on the
1994 input–output table.
A sector uses both capital, energy electricity, gas, coal, fuel, nuclear, and intermediate demand to produce output. The total
supply of these production factors is predetermined, and is called the factor endowment. The proportion in which the different pro-
duction factors are used is based on the relative prices of produc- tion factors and the technical substitution possibilities. Each sector
has its own substitution elasticity. The substitution elasticities are adapted to the Lithuanian situation based on expert opinions.
3C. Consumption
Consumers are assumed to maximize their utility, subject to the budget constraint. Total utility is a function of the utility derived
from the different consumer goods. In this model it is assumed that the consumption demand of both households and the govern-
ment can be described by a Linear Expenditure System LES in which the constants are equal to 25 percent households and 75
percent government of the base-year 1990 consumption levels, and the marginal expenditure shares are equal to the base-year
average expenditure shares. The LES uses disposable income INC, output prices, import prices, and factor prices P
x
, the marginal expenditure shares b
x
and exogenous demand D
x
to calculate domestic final demand D
x
The parameters of the LES b
x
and D
x
are derived from the io table. The general specification of the final demand of the goods and services of sector j reads:
D
j
5 D
j
1 B
j
P
j
INC 2
o
i
D
i
P
i
,
o
j
B
j
5 1,
3
where i, j are the index of production and energy sectors; D
j
is the final demand for sector j; P
j
is the producer price of sector j; D
j
is the minimum required constant quantity of good j; B
j
is the marginal expenditure share in the base year; and INC
is the total income.
3D. Foreign Trade
Foreign trade is subdivided into imports and exports. Imports results from the LES and are, in turn, input for both the factor
708 A. Galinis and M. J. van Leeuwen
supply energy and the supply per sector. Exports for the produc- tion sectors depend on domestic and world market prices, and
are input for the demand per sector and production. Exports per energy sector in the prototype model is set at the base-year level.
This approach was not suitable for Lithuania, especially for elec- tricity. Before the transition period, Lithuania exported a lot of
electricity to its neighboring countries. Exports in the base year of the model 1994 was rather small because of economy crisis
in the region. Now, electricity exports are increasing again. It is also expected that exports will continue to grow, because the
existing nuclear power plant can compete on the international energy market. Thus, it was necessary to introduce in the model
some changes in the estimation of the future electricity export. The level of the future electricity export in the Lithuanian version
of the CGE model was defined exogenously using a parameter of early exports growth.
Z
t
5 Z
o
1 1 a
100
t
2
T
, 4
where Z
t
is the electricity exports level in year t; Z
o
is the exports level in the base year; a
is the percentage electricity exports growth per year; T
is the base year; and t
is the year of the defined time horizon. This approach does not reflect the real situation in the electricity
market because it does not deal with local electricity price and electricity price in the electricity market of neighboring counties.
The actual situation is more complex, because two types of electric- ity are distinguished in the model. The proposed principle of
export evaluation is only correct if the electricity exports growth rate is confirmed with information from additional studies. Elec-
tricity export level from the Ignalina NPP is about 6 TWh. Thus, electricity export growth rate is calculated from this amount of
exported electricity. In future versions of the model it is necessary to include a more precise description of electricity trade.
3E. Environment
As mentioned in the Introduction, the effects of economic growth on environmental issues is an important topic in the politi-
cal and scientific debate. Our model distinguishes three types of emission gases CO
2
, NO
x
, and SO
2
, and calculates their emission
A CGE MODEL FOR LITHUANIA 709
for each sector. The sectoral level of emissions resulting from combustion is proportional to the amount of energy used in the
sector. For each sector, the use of coal, gas, and fuel is multiplied by an emission coefficient to calculate total emissions for each
pollutant. Emission coefficients in the CGE model are expressed in physical terms per monetary unit of energy consumed. The
actual emissions in economy sectors, actual fuel consumption per economy sectors and branches, as well as its monetary equivalent
were used for calculation of emission coefficients see Table A2 in the Appendix.
A maximum emission level for each pollutant can be imposed in the model. Environmental policies can be simulated by lowering
these levels. To meet more restrictive environmental goals, each sector has the possibility to either implement abatement technolo-
gies or to switch to a cleaner input mix. In practice, a government can control total emissions by setting regulations and standards.
In the model, however, the concept of tradable emission permits is used. This means that the maximum amount of emissions set
by the government is distributed between the producers by means of permits. The total number of permits a producer owns deter-
mines the maximum level of emissions the producer is allowed to release into the atmosphere. Emission permits can be traded
among producers. In this way a market of emission permits is established, and a price of permits is determined. Ultimately, the
price of a permit becomes equal to the marginal cost of abatement in each sector. Mathematically, the concept of tradable emission
permits can be described as follows:
PEM
em
TOTEMP
em
2 CLEAN
em
2 EMLIM
em
5 0 5
where em is SO
2
, NO
x
, CO
2
; TOTEM
em
is the total emission of pollutant em; CLEAN
em
is the total abatement of pollutant em; EMLIM
em
is the exogenous emission limit for pollutant em; and PEM
em
is the price of an emission permit for pollutant em.
4. SCENARIO RESULTS
