Economic Effects of Biofuel Promotion. An
Assessment with a World Dynamic
Computable General Equilibrium Model
V´ıtˇezslav P´ıˇsa a
& Jan Br˚uha b
aCharles University Prague, Faculty of Humanities
U Kˇr´ıˇze 8, 156 00, Prague 5, Czech Republic.
b ˇCNB, Czech National Bank
Na Pˇr´ıkopˇe 28, 115 03 Prague 1, Czech Republic.
The paper is contributed to honor Osvald Vasicek. Osvald has a long and productive career as an academic and a teacher.
One of the authors of this paper (Jan) was a student at the Masaryk University and under Osvald’s continuous support
and encouragement he started to study and apply methods of dynamic system theory, stochastic filtering, stochastic
control, and numerical methods to economic problems. Without his support and encouragement, his beginnings in these
fields were certainly much harder. Now, as Jan get older, he tries to transmit his knowledge to next students, although
Jan’s knowledge and pedagogical skills cannot match that of Osvald.
In this paper, authors apply those skills, especially from stochastic control and numerical mathematics, in which Jan was
initiated under Osvald’s guidance, to an important problem of public finance and environmental economics. As the other
author (Vitezslav) is now a student of Jan, this paper is a small witness of how tireless Osvald’s support as an academic
and a teacher has influenced the economic research in the Czech Republic in various areas. In fact, it would be hard to
overestimate Osvald’s contribution to Czech academia. Let us conclude these paragraphs by expressing the deep feeling
that the authors are much obliged to Osvald and his colleagues as they have been excellent scholars, inspiring teachers,
and great personages.
Abstract
The promotional policies targeted at alternative fuels take different forms and are
currently a very important policy issue integrating the environmental, economic,
and fiscal concerns. While choosing different environmental policy mixes, almost
all important world regions support production and use of biofuels. This paper
introduces the first version of the computable general equilibrium (CGE) model
aiming at the analysis of promotional policies focused on the first generation biofuels
over the world. The model is suited for evaluation of impacts of such policies on the
economy of the Czech Republic. The main features of the model, ie. multiregionality,
intertemporality, and a quasi-fixed land supply (limited area of agricultural land as
one of the production input of agricultural sector) deliberately adress the structure
and nature of biofuel market. Countries are properly aggregated in order to underpin
production and trade of biofuels over the world and also the ability of regions to
affect world price of crude oil. The sectors has been selected in order to know the
links in biofuel production use chain and process of creation of prices in fuel and
agricultural sector. The paper presents exemplar simulations with the core of the
model.
Key words: Biofuels, CGE modeling, Land Use
J.E.L. Classification: C68, Q18, Q42
1 Introduction
The protection of the environment has gradually become an important policy
topic over the world since 1970s. While following the aims of environmental
policy, mainly reduction of GHG emissions, policymakers have recently begun
to promote the production and the use of biofuels. 1
In addition to favor
environmental effects, the most cited economic advantages of growing biofuel
share on the market are creation of outlets for agricultural commodities
and consequent support of rural development, employment in agriculture and
weakening the dependency of economies on fossil fuel sources coming usually
from politically volatile regions.
Taking in mind the above mentioned favor effects, the most important world
regions began to support the production and/or the use of biofuels. The European
Union declared to substitute 10 % of traditional fossil fuels by biofuels in
transport by the year 2020. The U.S., in the American Energy Independence
and Security Act, released in 2007, dictate for placement of 15 bn. gallons of
biofuel on the U.S. transport market by the year 2015 and 36 bn. gallons by
the year 2022. Brazil, as the second biggest bioethanol producer, published its
Brazilian Agroenergy Plan for the years 2006-2011, which stipulates to place
annually between 20 and 25 % of bioethanol and 5 % of biodiesel on the transport
market. Furthermore, China, Australia, Japan and other countries have
chosen some mix of biofuel promotional measures 2
.
In contrast to studies mentioning favorable impacts of biofuel promotion, there
have also appeared voices condemning especially first generation biofuels of
having strong adverse impact on prices of food and propellants together with
insignificant effects on reductions of GHG emissions. The rapidly increased
water demand as well as negative effect on land use and land use change
have also been reminded. In addition, there are also doubts about moving the
production of biofuels abroad with ambiguous impacts on rural development,
especially domestic value added and employment in agriculture.
Therefore, there is a strong need for policymakers to know the impacts of
Email addresses: vitezslav.pisa@centrum.cz (V´ıtˇezslav P´ıˇsa),
jan bruha@yahoo.co.uk (Jan Br˚uha).
1 Firstly, so called first generation biofuels were cited. These biofuels are, however,
made from feedstock that can alternatively be used for production of food. On
the other hand, so called second generation biofuels are made from bio-wastes,
etc. However, these types of biofuels are relatively expensive and currently their
production can not be commercially processed. The most common first generation
biofuels are bioethanol made from sugar cane, sugar beet or wheat or biodiesel, ie.
methyl ester made from rapeseed oil.
2 For more detail, see OECD [34].
2
chosen biofuel support scheme mix, especially on economic variables like employment,
value added and selected prices. With respect to specific characteristics
of production/use chain of biofuels, economic model shall namely
address following issues: the existence of restricted area needed for production
of biofuels, the existence linkages between main sectors in biofuel production/use
chain in the national economy affecting prices in selected sectors and
the existence feedbacks from those production sectors and final markets to
agricultural market. The international dimension of the model should reflect
the ability of some regions to affect world price of crude oil and fossil fuels
derived from crude oil. Also the international dimension should reflect that
the biofuels and their feedstock are largely traded commodities. The nature of
technological development, the need to know right time of adoption of some
type of biofuel technologies together with special treatment of oil reserves call
for incorporation dynamic features into the model. As a result, the multiregional
multi-sectoral dynamic computable general equilibrium model has
been chosen for the analysis.
This paper is organized as follows: Section 2 is devoted to brief summary
of CGE models aimed at assessing biofuel promotional policies. Section 3
describes the current version of the model, Bellman’s equation and solution of
the model for oil producing country. Section 4 shows an exemplar simulation
with the current version of the model, while section 5 concludes by mentioning
future extension of current version of the model.
2 Biofuel Issue in CGE Models
Because the issue of biofuel promotion has gradually become very important,
there can be seen attempts to incorporate the topic into general equilibrium
framework 3
.
Generally, there are several approaches of incorporating land use issue into
general equilibrium models. The straightforward approach consists of incorporation
of the land use issue into the general equilibrium framework, for
example by addressing functional forms of the agriculture sector, for example,
by treatment of land as homogeneous factor of production. Such approach is
chosen by Dixon et. al. [12], who apply an economic model of the U.S. economy
USAGE for evaluation effects of setting mandatory blending quotas in
the U.S. by the year 2020, or by Kretschmer [24] who uses adjusted international
DART model described in Kretschmer [26] for evaluation of economic
impacts of 10 % biofuel target in the EU. Similar approach to land use issue
3 See for example Kretschmer and Peterson [24], Palatnik and Roson [36] or Hertel
et. al. [21] for relatively exhausting review.
3
is chosen by Arndt [4] who analyzes impacts of increased private investments
into biofuel production in Mozambique. To analyze effects on the land use in
the U.S. and the rest of the world, Reilly and Paltsev [41] describe the incorporation
of biomass energy production and competition for land into the world
Emissions Prediction and Policy Analysis (EPPA) model 4
. A relatively new
biofuel technologies are modeled via so called latent technology approach 5
.
Perry et. al. [39] carry out the analysis of response of the Argentine economy
to a boost in the global demand for biofuels and their feedstocks. They run two
scenarios with fixed amount of the land as input of production and scenario
allowing for increase of available land. While they enrich the initial G-Cubed
model 6
McKibbin and Wang [30] also assume homogeneous land.
The issue of restricted area and its utilization can also be addressed by incorporating
the constant elasticity of transformation (CET) function into the
model allowing landowners for choosing the optimal utilization of restricted
land area according to its yields and usage. Such approach use Hertel and Tsigas
[18] who analyze effects of the elimination of farm and food tax preferences
on the US economy in 1997. Hertel [17] also follows this land use modeling
approach and analyzes the effects of agricultural policies on commodity trade
and markets with the help of GTAP model. Banse et. al. [5] use more elaborated
system of constant elasticity transformation functions that take 3-level
nesting structure and extend the GTAP–E model. The model is employed in
analysis of impacts of mandatory blending quotas of the European Biofuel
Directive 7
.
Therefore, more complicated models assume heterogeneous types of land 8
or divide regions into so-called agro-ecological zones or interconnect detailed
bottom-up land use models with top-down CGE models 9
. These models are
usually very data demanding and/or they are limited in dynamics because of
their relatively high disaggregation.
4 In addition to utilization the biofuels in transport, they also consider the biomass
burnt for production of electricity.
5 Latent technologies are usually existent in current time but are too expensive to be
utilized. While the economy develops, the latent technology become profitable. Their
incorporation is usually common in dynamic models. Modeling latent technologies
needs the data on the input and cost structures of the different types of production
of biofuels and mark-up between production costs and prevalent fuel prices. For
example the production of second generation biofuels can be modeled in this way.
6 Multiregional intertemporal world CGE model
7 Directive 2003/30/EC on the promotion of the use of biofuels or other renewable
fuels for transport.
8 See for example Abdegalil and Cohen [1] or Abdula [2]
9 See for example, Ronneberger et. al. [42], [43], Lee et. al.[28], or Bosello and Zhang
[9]
4
To sum up, currently, there exist several approaches to modeling economic
impacts of biofuel support in the world via CGE models. These models are
aimed at the EU, the US economy as well as developing countries. A significant
part of CGE models applied on the issue of biofuels analyze introduction
of some level of mandatory substitution of traditional fossil fuels by biofuels.
The most inspected impacts are price effects, mainly on agricultural commodities
and food but the are also to explain impacts on value added, trade and
employment. The land use issue is treated in different ways, ie. by adoption of
special functional forms in production, incorporation of CET functions with
heterogeneous types of land, incorporation of so called agroecological zones,
or setting sophisticated linkage between bottom up land use models and top
down CGE models. The following model chooses the way of incorporation of
special functional forms into the CGE. The chosen treatment of land use allows
for proper incorporation of dynamics into the model and better addresses
regional data availability in the world.
3 The Core of the Model
The model is divided among four sectors (agriculture, crude oil extraction,
fuel industry and other final good market) and eight aggregated world regions
(the Czech Republic, the rest of the EU-27, the OPEC countries, the rest of
the OECD, the former USSR without Baltic states, the rest of Asia, Africa,
and the rest of Americas) 10
.
3.1 The Social Planner Problem
Each region is endowed with an arable land L, and populated by a representative
agent. The representative agent consumes a final good Cit, food Ait,
public good Git, and supplies an inelastic amount of labor Lit. In each region,
there is a social planner 11
who acts on behalf of the representative agent and
solves following problem:
max E0
∞
t=0
βt
u (Cit, Git, Ait) , (1)
10 For exact division of the countries see Table 5
11 The planner approach can be defended with the reference to the first welfare
theorem, which states that the competitive allocation coincides with a planner allocation
under certain circumstances (which are satisfied here). Negishi [33] shows
that in CGE models it may be easier to solve the planner problem rather than to
find the corresponding competitive allocation and this paper follows this approach.
5
where E0 is the expectation operator and β is the parameter of the intertemporal
rate of substitution. In the formulae above, as well as in all subsequent
formulae, the subindex i refers to regions, while the subindex t refers to time.
The planner maximizes (1) and respects the following set of constraints:
• the GDP identity:
Cit + Git + Iit + φ(Iit, Kit) + Io
it + φo
(Io
it, Ko
it) + κ (Tit) + Xit = Yit,
where Iit are investments, Kit is the capital used in the final good sector,
φ(Iit, Kit) is the investment adjustment function, Io
it and Ko
it are investments
and capital used in the oil extracting sector with φo
being the corresponding
adjustment cost function (for regions, which do not produce oil, we set
trivially Io
it = Ko
it = 0), Xit are net exports of the final good, κ (Tit) is the
cost of using the backstop technology, and Yit is the production of the final
good;
• the production function in the final-good sector:
Yit = Y(Kit, ζitLy
it, Fit),
where Y is the neoclassical production function, Ly
it is the labor used in that
sector, Fit is the fuel used in that sector, and ζit is the labor augmented
technological progress;
• the production function in the fuel-producing sector:
Fd
it = F(Oit, ζb
itBit, ζT
it Tit, Lf
it),
where F is the production function, Oit are the conventional fossil fuels
(henceforth referred to as oil) used for the production of fuel, Bit are the
agricultural inputs, ζb
it is the technology progress, Tit is a back-stop technology
and ζT
it is the corresponding technological progress, and Fd
it is the
domestic production of fuels. The domestic production is related to the
domestic consumption of fuels by the following identity:
Fd
it = Fit + Fx
it,
where Fx
it are net exports of fuels from region i. Similar identities hold for
production, consumption and net exports of food and agricultural products
used for biofuel production:
Ad
it = Ait + Ax
it,
Bd
it = Bit + Bx
it,
where the d superscript denotes the domestic production, and the x superscript
denotes net exports of the relevant variable.
6
• the agricultural production function reads as follows:
Ad
it + Bd
it = A(La
it, Li, ζa
it),
where Li is the agricultural land in region i, La
it is labor employed in agriculture,
and ζa
it is the exogenous technological progress. The agricultural
production function obviously requires that πbt ≡ πat.
• the oil consumption Oit, oil production Od
it, and oil net exports Ox
it
obey the following identities:
Oit = Od
it − Ox
it,
Od
it = χ(Ko
t , Lo
t , Oe
it),
Rit+1 = Rit − Oe
it + ζr
it,
where Rit are oil reserves, ζr
it are shocks to oil reserves (such as new discoveries).
The function χ(Ko
t , Lo
t , Oe
it) captures the notion that it is costly to
extract the oil and that capital and labor should be used, we assume the
sector-specific capital, i.e., the capital used in the oil extracting industry
cannot be easily transferred to the final good sector 12
. The requirements
on variables are Rit ≥ 0, and for oil-nonproducing regions holds: Rit ≡ 0,
Od
it ≡ 0, ζr
it ≡ 0, and thus −Oit = Ox
it ≤ 0.
• The balance of payments reads as:
Xit + πatBx
it + πatAx
it + πftFx
it + πotOx
it + (1 + rt)Wit = Wit+1,
where Wit is the net worth of region i, and πft, πat, πot are relative prices
of fuel, agricultural products, and oil respectively. Recall that πat = πbt by
virtue of the agricultural production function.
• The capital accumulation is standard one:
Kit+1 = (1 − δ)Kit + Iit,
Ko
it+1 = (1 − δo)Ko
it + Io
it,
where δ and δo are depreciation rates in the two sectors.
• The planner respects also the labor-market clearing conditions:
Lit = Ly
it + La
it + Lf
it + Lo
it.
There are multiple sectors in the economy: (i) the sector which produces the
final good, which is both the consumption good as well as the investment
good; (ii) the agricultural sector which produces food and biofuel inputs; (iii)
12 This assumptions is responsible of a sluggish adjustment of oil supply to its prices,
a feature observed in reality.
7
the oil producing sector, which produces oil subject to crude oil reserves; this
sector does not operate in each country, (iv) the fuel producing sectors which
produces fuel from oil and from agricultural inputs, and (v) finally there is a
back-stop technology, which has the potential to replace oil from producing
fuels. The first four sectors are necessary ingredients in our model: the first
sector is needed since it produces consumption and investment goods. The
agricultural sector is needed to generate the biofuel-food trade-off (the land
used for biofuel corps cannot be used for food production). The oil producing
sector is necessary for (a) the model-consistent derivation of the oil price,
since increased biofuel utilization can have effects on the oil price, on the
investments in oil industry (hence affecting the future oil supply and future
oil prices), and (b) modeling the trade effect and the cross-country distribution
resulting from biofuel promotion (as biofuel policies may result in real transfers
from oil producing countries to countries growing biofuel crops).
We include the backstop technology into the model for technical reasons only.
Its only purpose is to have a well defined steady state, where the economy
does not collapse to inaction due to the oil exhaustion 13
. Hovered, in our
calibration, we always set productivity of the back stop technology very low
in near future (so its utilization is negligible), and assume that it will substitute
fossil fuels in very far future. By this we can impose a non-trivial model steady
state without too much affecting the model results for near and medium terms.
3.2 The Functional Forms of Production Functions
The production functions are assumed to have the following functional form:
• The production function in the final good sector Y is assumed to have
standard Cobb-Douglas form, where the shares are derived from the ‘averages’
observed in the economy.
• The production function in the sector producing fuel satisfies:
F(Oit, ζb
itBit, Lf
it) = (αoO−ρ
it + (1 − αo)B−ρ
it )−αf /ρ
(Lf
it)(1−αf )
.
The parameter ρ ∈ (−1, 0) captures the notion that biofuels and fossil fuels
13 The reader may ask whether there exists a parametrization such that the economy
can attain a non-decreasing path of production and consumption despite the
continuous decline in oil supply. We can confirm that such a parameterizations indeed
does exist – basically the share of fuel in the final good production function
has to be low or the agriculture sector must be very productive. Then, it is possible
to substitute fuels by capital or oil by biofuels. Nevertheless, such a property is not
always satisfied and we do not want to ex-ante impose such a calibration on our
model.
8
are substitutes (ρ < 0), although imperfect (ρ > −1). The labor share αf is
calibrated to replicate the labor costs in this industry. In this version of the
model, we abstract from capital (according to the cost structure, its cost
share of capital is much lower than that of fuels, oil, and labor).
• The production function in the agricultural sector is defined as follows:
A = ζa
itLi (La
it)αA
,
where ζa
it models the technological progress, Li is the amount of arable land
(assumed to be constant in this version of the model), and La
it is labor used
in agriculture. The parameter αA models decreasing returns to scale of labor.
In this version of the model, we assume that the agricultural land is
fixed. However, if an extension with variable land use were considered, the
production function should change as this form would imply increasing returns
to scale in agriculture 14
. Still, the fixed land assumption is reasonable
if one wants to model policies constrained by ‘sustainable’ criteria currently
being set on biofuels.
3.3 The Bellman Equation and Solution for an Oil-producing Region
In this section, we deliberately describe the Bellman’s equation and solution
for oil producing region. By setting some assumptions, reader may easily derive
relevant equations for oil-nonprooducing region including, for example, the
Czech Republic 15
.
The Bellman’s equation for oil producers reads as follows:
V(Kit, Ko
it, Wit, Rit) = max u (Cit, Git, Ait|ξit) + βEtV(Kit+1, Ko
it+1, Wit+1, Rit+1)
−λf
t Fit − Fx
it − F(Oit, ζb
itBit, ζT
it Tit, Lf
it)
−λa
t Ait + Bit + A
x
it − A(La
it, Li, ζa
it)
−λl
t Lit − Ly
it − La
it − Lf
it − Lo
it − vitTit,
where a new variable A
x
it = Ax
it + Bx
it has been created to lower the dimensionality
of the problem. Also the GDP identity can be used to substitute
the consumption from the problem. The variable λf
t is the Lagrange multiplier
associated with the constraint given by the fuel production function,
the Lagrange multiplier λa
t is associated with the agricultural sectors, λl
t is
the Lagrange multiplier associated with the labor constraint, and vit is the
14 We thank to Sergey Slobodyan for pointing us this issue.
15 For oil non-producing regions we set Io
it = Ko
it = 0, Rit ≡ 0, Od
it ≡ 0, ζr
it ≡ 0 and
consequently −Oit = Ox
it ≤ 0.
9
Kuhn-Tucker multiplier related to the backstop technology 16
. The resource
constraint Rit = Rit+1 − Oe
it + ζr
it enters directly the value function. We also
eliminate Od
it by Oit = Od
it − Ox
it = χ(Ko
t , Lo
t , Oe
it) − Ox
it and optimize with
respect to Oe
it and Ox
it only.
The condition λf
t FOt = βEtVRt+1 with uct = βEtVWt+1 and with πotβEtVWt+1 =
βEtVRt+1 yield a familiar equality of marginal product with prices πftFOt = πot.
The inter-temporal Euler equation with the envelope condition dictate the
generalization of the Hotteling rule 17
:
πotuctχOt = βEtuct+1πot+1χOt+1
After some simple algebraic manipulation with the first oder conditions 18
, we
get the solution in the following form:
uct = ugt = uat/πat,
YFt = πft,
YLt = πatALt = πftFLt = πftFotχLt,
πftFBt = πat,
πftFOt = πot,
The interpretation of these equations is straightforward. First equation equalize
marginal utility of private consumption with marginal utility of public
consumption (famous Samuelson condition) and also with marginal utility of
consumption of agricultural good evaluated by its price 19
. The second equation
equals marginal product of fuel to its price. The third equation equalize
marginal productivities of labour in selected sectors. Fourth and fifth equation
show that marginal product of biofuels (agricultural products) and crude oil
equals to their prices.
Moreover the Euler equation of the problem equalizes marginal utility of private
consumption in two time periods:
uct = βEt(1 + rt+1)uct+1
The higher the interest rate in next period, the more would consumer like
to invest in current period (causing an increase in marginal utility of current
16 I.e., vit = 0 if the technology is used in year t and vit > 0 if it is not.
17 If the oil were extracted costlessly, this equation would imply the standard Hotteling
rule ( πot = βEt
uct+1
uct
πot+1 ), which dictates that the price of an exhaustible
asset should growth at the rate of intertemporal substitution.
18 Here we transform shadow prices λa
t , λf
t , λo
t into real monetary units using the
following relation: uct = λa
t /πat = λf
t /πft.
19 This is because πct ≡ πgt ≡ 1
10
consumption). On the other hand the increase in consumption in the next
period causes decrease in marginal utility of consumption. Finally, we also get
the equation of the inter-temporal allocation of capital:
uct(1 + φIt) = βEtuct+1 [YKt+1 + φKt − (1 − δ)(1 + φIt+1)] .
3.4 The closure of the model
To close the model, the rest of equilibrium conditions have to be specified.
Mainly, the sum of world net exports should be zero:
i∈I
Xit = 0,
i∈I
A
x
it = 0,
i∈I
Fx
it = 0,
i∈I
Ox
it = 0.
These equations will determine the international relative prices πat, πft, πot and
the world interest rate rt. By virtue of the Walras law, also i∈I [Wit+1 − (1 + rt)Wit] =
0.
3.5 Calibration of the Model
Up to now, the model has been calibrated to reflect:
• average shares of food in consumption are used for calibration of the parameters
of the utility function u;
• the parameters of the utility function u related to the government spending
are calibrated from national account to mimic the ratio of government
spending on GDP;
• the cost structure in relevant industries is utilized for calibration of the
parameters of the production functions;
• the data on international trade in oil and agricultural products replicates
the trade balances;
• the parameter β is calibrated to the conventional value in literature (0.95).
11
4 An Illustrative Simulation
The core version of the model was applied for simulation of the impact of
growing world oil prices. Figure 1 reports the long-run (steady state) effects
of oil price changes on the Czech variables. On the x-axis there is assumed
percentage change in the oil price (normalized initially at 100%) and on the
y-axis there is an implied long-run change in selected variables.
It can be seen that growing world price of crude oil causes a reduction of crude
oil imports. The response is relatively high and will be subject to further revisions
of calibration. On the other hand, the negative response of crude oil
imports and consequently traditional fossil fuel production leads to penetration
of biofuels on the market. This can be seen in the upper right corner
graph where consumption of biofuels increases with the growing price of crude
oil. Again, the response is relatively high and will be subject to revisions.
Nevertheless, the expected relationship is seen, ie. the crude oil dependency
decreases with crude oil prices while the biofuels substitute fossil fuels. It can
also be seen that agricultural products needed for production of biofuels are
being obtained from abroad. Simultaneously, previously exported agricultural
production is used for domestic production of biofuels - therefore net exports
of agricultural goods are decreasing with increasing price of crude oil. The increase
in imports can be explained by the fact that the similar effect applies in
other regions and the biofuel production is boosted worldwide, but relatively
more in less developed regions, which are now willing to serve the demand
in the developed regions including Czech Republic. Still, even the domestic
agricultural production boosts due to price incentives. We also see a slight
increase in domestic agricultural employment 20
.
5 Conclusion
We introduced the description of the world model that is assumed to evaluate
impacts of biofuel promotional policies in the world on the Czech economy.
Because of the nature of biofuel market and biofuel production/use chain, the
multiregional dynamic general equilibrium model has been chosen for the analysis.
The need for proper incorporation of dynamic features and also due to
the data availability our model assumes restricted homogeneous land area as
input into agricultural production. However, the model underpins competition
between biofuels and food on restricted land area with consequent impact on
prices of agricultural commodities. Currently, the static version of the CGE
model has been calibrated and applied for initial simulations. The simulations
20 It can be stressed that the simulation were run with simple core of the model.
12
are gradually being improved but there still exists an area for further improvements
and extensions. Nevertheless, we investigated the effect of changes of
crude oil prices on the main variables in the model. The model results imply
that growing crude oil price causes a decrease in crude oil imports, increase
in domestic consumption of biofuels, with moderate favor impact on domestic
agricultural production and agricultural employment. In order to evaluate impacts
of biofuels promotional measures, the model has to be further enriched.
It is planned to extend it by the existence of two types of households, at least
for the case of the Czech Republic. In this case, the social planner approach
has to be left. The types of households will address the need for more comprehensive
analysis of impacts on agricultural households (and regions). The
associated extension consists of introduction of distortionary taxes like labor
taxes, profit tax, VAT and excise tax.
Acknowledgements
The research on this paper has been supported by the grant of the Grant
Agency of Charles University ‘vs 251 203’ No. 123310. The support is gratefully
acknowledged.
References
[1] Abdelgalil, E. A. and Cohen, S. I.: Policy modelling of the trade-off between
agricultural development and land degradation - the Sudan case. Journal of
Policy Modeling 23, (2001), pp. 847-874.
[2] Abdula, R.: Computable General Equilibrium Analysis of the Economic and
Land-use Interfaces of Bio-energy Development.Contributed paper presented
for presentation at the International Association of Agricultural Economists
Conference, Gold Coast, Australia, August 12-18, (2006).
[3] Adenauer, M.: CAPRI versus AGLINK-COSIMO: Two partial equilibrium
models - Two baseline approaches, presented at international EAAE workshop,
26-29th August 2008, Gent, Belgium.
[4] Arndt, Ch., Benfica, R., Tarp, F., Thurlow, J. Uaiene, R.: Biofuels, Poverty
and Growth. A Computable General Equilibrium Analysis of Mosambique,
International Food Policy Research Institute, Washington, USA (2008).
[5] Banse, M. , ven Meil, H., Tabeau, A., Woltjer, G.: Will EU biofuel Policies
affect Global Agricultural Markets. European Review of Agricultural Economics,
35(2) (2008), pp.117-141.
[6] Bergman, L.: CGE modeling of Environmental Policy and Resource
Management, Handbook of Environmental Economics, 3, Elsevier (2005).
13
[7] Birur, K. D., Hertel T. W., Tyner W. E.: Impact of Biofuel Production
on World Agricultural Markets: A Computable General Equilibrium Analysis.
GTAP Working Paper, 53 (2008).
[8] Boeters, S., Veenendaal, P., van Leeuwen, N. and Rojas-Romagoza, H.: The
potential for biofuels alongside the EU-ETS. Paper presented at the 11th
Annual GTAP Conference. Helsinki, Finland.
[9] Bosello, F., and Zhang, J: Assessing Climate Change Impacts: Agriculture.
FEEM Working Paper 94 (2005).
[10] Britz,
W. (ed.): CAPRI Modelling System Documentation, URL:http://www.ilr1.unibonn.de/agpo/rsrch/capri/capri-documentation.pdf
(2005).
[11] Burniaux, J. M., and Lee, H-L.: Modelling Land-Use Changes in GTAP.
Conference paper. Sixth Annual Conference on Global Economic Anlaysis.
Hague. The Netherlands (2003).
[12] Dixon, P. B., Osborne, S., Rimmer, M. T.: The Economy-Wide Effects in the
United States of Replacing Crude Petroleum with Biomass, Discussion paper
sent to GTAP Workshop, Purdue University, Indiana (2007).
[13] Doumax, V.: Assessing the Economic Impacts of Biofuels : the Use of
Computable General Equilibrium Modeling. Centre d’Analyse Economique
Working Paper (2009).
[14] Giesecke, J. A., Horridge, J. M. and Scaramucci, J. A.: Brazilian Structural
Adjustment to Rapid Growth in Fuel Ethanol, Studies in Regional Science, 39
(1) (2009), pp. 189-207.
[15] Gurgel, A. C., Reilly, J. M. and Paltsev, S.: Potential land use implications
of a global biofuels industry. Journal of Agricultural & Food Industrial
Organization5, Article (2007). (Special Issue: Explorations in Biofuels
Economics, Policy, and History).
[16] Haurie, A., Hung, N. M.: Turnpike Properties for the Optimal Use of a Natural
Resource. The Review of Economic Studies 44(2) (Jun.,1977), pp. 329-336.
[17] Hertel, T. W.: Global Trade Analysis: Modeling and Applications. Cambridge,
Cambridge University Press (1997).
[18] Hertel, T. W., and Tsigas, M. E.:Tax Policy and U.S. Agriculture: A General
Equilibrium Approach. American Journal of Agricultural Economics 70(2)
(1998).
[19] Hertel, T. W., Tyner, W. E. and Birur, D. K.: Biofuels for all? Understanding
the Global Impacts of Multinational Mandates. GTAP Working Paper, 51.
Center for Global Trade Analysis Department of Agricultural Economics,
Purdue University (2008).
[20] Hertel, T. W., Rose, S., and Tol, R. S. J.: Land Use in Computable General
Equilibrium Models: An Overview. GTAP Working Paper 39 (2008).
14
[21] Hertel, T. W., Rose, S. K. and Tol, R. S. J.: Economic analysis of land use in
global climate change policy. Routledge, Abingdon, New York.
[22] Keeney, R., and Hertel, T. W.: GTAP-AGR: A Framework for Assessing the
Implications of Multilateral Changes in Agricultural Policies. GTAP Technical
Paper 24 (2005).
[23] Kongsamut, P., Rebelo, S., Xie, D.: Beyond Balanced Growth. NBER Working
Paper 6159, 1997.
[24] Kretschmer, B. and Peterson, S.: Integrating bioenergy into computable general
equilibrium models: A survey, Energy Economics, 32 (2009), pp. 673-686.
[25] Kretschmer, B., Narita, D. and Peterson, S.: The economic effects of the EU
biofuel target. Energy Economics 31 (2009), pp. 285 - 294..
[26] Kretschmer, B., Peterson, S., Ignaciuk, A.: Integrating Biofuels into the DART
Model, Kiel Institute for the World Economy Working Paper, 1472 December
(2008).
[27] Lee, H. L.: Incorporating agroecological zoned data into the GTAP Framework.
Paper presented at the Seventh Annual Conference on Global Economic
Analysis, The World Bank, Washington DC, USA, June 17 19, (2004).
[28] Lee, H. L., Hertel, T. W., Rose, S. and Avetisyan, M.: An integrated global land
use data base for CGE analysis of climate policy options. Economic Analysis
of Land Use in Global Climate Change Policy. Hertel, T. W., Rose, S. and
Tol, R. S. J. eds., Abingdon: Routledge, (2009).
[29] McDonald, S., Robinson, S., and Thierfelder, K.: Impact of Switching
Production to Bioenergy Crops: The Switchgrass Example. Energy Economics,
28 (2) (2007), pp. 243-265.
[30] McKibbin, J. W., and Wang, Z.: The G-Cubed (Agriculture) Model: A Tool for
Analyzing Agriculture in a Globalizing World. Brookings Discussion Papers in
International Economics 139 (1998).
[31] McKibbin, J. W., Wilcoxen, P. J.: The theoretical and empirical structure of
the G-Cubed model. Economic Modelling 16 (1999), pp. 123-148.
[32] Msangi, S., Sulser, T., Rosegrant, M. W., Valmonte-Santos, R. and Ringler, S.:
Global Scenarios for Biofuels: Impacts and Implications, International Food
Policy Research Institute (2006).
[33] Negishi, T.: Welfare economics and existence of an equilibrium for a competitive
economy. Metroeconomica 12 (1960), pp. 92 -97.
[34] OECD: Biofuel support policies - an economic assessment, OECD, Paris (2008).
[35] OECD: Subsidies: The Distorted Economics of Biofuels. OECD Discussion
Paper 2007 3, Joint Transport Center (2007).
15
[36] Palatnik, R. R and Roson, R.: Climate Change Assessment and Agriculture in
GEM: Alternative Modeling Strategies. CIRCE project paper funded by EC 6th
Framework Programme, CMCC, Italy, April (2009).
[37] Paltsev, S., Reilly, J., Jacoby, H., Eckaus, R., McFarland, J., Sarofim, M.,
Asadoorian, M. and M. Babiker: The MIT Emissions Prediction and Policy
Analysis (EPPA) Model: Version 4, MIT Joint Program on the Science and
Policy of Global Change, Report 125, Cambridge, MA (2005).
[38] Peters, J., Thielmann, S.: Promoting Biofuels: Implications for Developing
Countries. Ruhr Economic Papers 38 (2007).
[39] Perry, M.: Food Production vs. Biomass Export vs. Land-Use Change: A CGE
Analysis for Argentina. Revised version of paper presented at 16th Biomass
Workshop, Valencia, June (2008).
[40] Rajagopal, D. and Zilberman, D.: Review of Environmental, Economic and
Policy Aspects of Biofuels, Policy Research Working Paper 4341, World Bank.
[41] Reilly, J. and Paltsev, S.: Biomass Energy and Competition for Land.
GTAP Working Paper 46. Center for Global Trade Analysis Department of
Agricultural Economics, Purdue University (2008).
[42] Ronneberger, K., Berrittella, M., Bosello, F., and Tol, R. S. J.: KLUM@GTAP:
introducing a biophysical aspects of land-use decisions into a general equilibrium
model-A coupling experiment. FEEM Working Paper 102 (2006).
[43] Ronneberger, K., Berrittella, M. Bosello, F. and Tol, R. S. J.: KLUM@GTAP:
Spatiallyexplicit, biophysical land use in a computable general equilibrium
model. Economic Analysis of Land Use in Global Climate Change Policy.
Hertel, T. W., Rose, S. and Tol, R. S. J eds., Abingdon: Routledge, (2009).
[44] Rosegrant, M. W., Ringler, C., Msangi, S., Sulser, T. B., Zhu, T. and
Cline, S. A.: International Model for Policy Analysis of Agricultural
Commodities and Trade (IMPACT): Model Description, International Food
Policy Research Institute, Washington, D.C., USA (2008).
[45] Stattman, S. L., Bindraban, P. S. and Hospes, O.: Exploring Biodiesel
Production in Brazil: A study on configurational patterns in an evolving policy
domain. Plant Research International B.V. and Wageningen Report 199 (2008).
[46] Steininger, K. and Voraberger, H.: Exploiting the Medium Term Biomass
Energy Potentials in Austria, Environmental and Resource Economics 24(4),
(2003), pp. 359- 377.
[47] van der Werf, E., and Peterson, S.: Modeling linkages between climate policy
and land use: An Overview. Kiel Working Paper 1323 (2007).
[48] Wiesenthal, T., Leduc, G., Christidis, P., Schade, B., Pelkmans, L., Govaerts, L.
and Georgopoulos, P.: Biofuel support policies in Europe: Lessons learnt for the
long way ahead. Renewable and Sustainable Energy Reviews 13 (2009), pp. 789
- 800.
16
Table 1: Division of the world.
Region Countries
Czech Republic Czech Republic
EU 26 Austria, Belgium, Bulgaria, Cyprus, Germany, Denmark, Spain,
Estonia, Finland, France, United Kingdom, Greece, Hungary, Ireland,
Italy, Lithuania, Latvia, Luxembourg, Netherlands, Poland,
Portugal, Romania, Slovak Republic, Slovenia, Sweden, Malta
rest OECD Australia, Canada, Switzerland, Iceland, Japan, Korea, Mexico,
Norway, New Zealand, Turkey, United States
OPEC countries Angola, United Arab Emirates, Algeria, Ecuador, Iran, Iraq,
Kuwait, Lybia, Nigeria, Quatar, Saudi Arabia, Venezuela
former USSR Azerbaian, Belarus, Georgia, Kazakhstan, Russia, Ukraine,
Uzbekistan
rest Asia Bangladesh, China, Hong Kong, Indonesia, India, Lebanon, Sri
Lanka, Malaysia, Oman, Pakistan, Philippines, Singapore, Syrian
Arab Republic, Thailand, Vietnam, Yemen
Latin America Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, Dominican
Republic, Guatemala, Honduras, Jamaica, Panama, Peru,
Puerto Rico, Paraguay, El Salvador, Uruguay
rest Africa Botswana, Cote d’Ivore, Cameroon, Egypt, Ethiopia, Kenya, Morocco,
Sudan, Tunisia, Tanzania, South Africa, Congo, Zambia,
Zimbabwe
17
Fig. 1. Simulation results
18