Land Use in Computable General Equilibrium Models: An Overview*
Thomas W. Hertel
Center for Global Trade Analysis, Purdue University
US Environmental Protection Agency
Richard S. J. Tol
Economic and Social Research Institute
GTAP Working Paper No. 39
*Chapter 1 of the forthcoming book Economic Analysis of Land Use in Global Climate
Change Policy, edited by Thomas W. Hertel, Steven Rose, and Richard S.J. Tol
Table of Contents
Introduction ........................................................................................................ 3
Previous literature .............................................................................................. 4
Methodological issues .......................................................................................... 4
Land-use modeling and climate change ............................................................... 4
Land use in economics ......................................................................................... 5
Geographic models of land use ............................................................................ 6
Economic models of land use ............................................................................... 7
Land use and climate change ............................................................................... 8
Land use and climate policy ............................................................................... 10
Contributions of this Volume and Challenges for Future Work ................. 11
The Spatial Dimension ....................................................................................... 11
Mobility of Land Across Uses and Diversification of Production ..................... 12
Incorporation of the Forestry Sector in Global Models of Land Use. ............... 15
Accessing New Lands ......................................................................................... 16
Biofuels and Land Use ....................................................................................... 17
Agricultural greenhouse gas emissions and mitigation ..................................... 18
The Role of Non-primary Demands for Land..................................................... 19
Conclusion ......................................................................................................... 20
References ........................................................................................................................ 22
Figure 1. Global cropland (a), forest land (b) and grassland (c) projections .................... 30
Figure 2. Baseline land-use change and forestry carbon net emissions. ........................... 31
Figure 3. Cummulative cost-effective agricultural, forestry, and biomass abatement ..... 32
LAND USE IN COMPUTABLE GENERAL EQUILIBRIUM MODELS:
Thomas W. Hertel, Steven Rose, and Richard S.J. Tol
November 22, 2007
Human intervention over the last several centuries has markedly changed land surface
characteristics, primarily through large scale land conversion for cultivation (Vitousek et
al., 1997). The future will not be different. Populations and economies continue to grow,
and climate change and climate policy will affect land use at a massive scale. However,
land has long been neglected in global economics. That is now changing. For the first
time, modeling teams have explicitly introduced global land use into computable general
equilibrium (CGE) models, the work horses of economic policy analysis. In this book, we
present a collection of pioneering papers in the applied economics of land use in CGE
models. This book describes and critically assesses the underlying data, the
methodologies used, and the first CGE applications. The applications are all with respect
to climate change and climate policy, but the methods and data can also be used for other
applications, such as land use change, energy security, and nature conservation.
The policymaking community is trying to determine the potential role for agriculture and
forestry in climate change mitigation, and define implementation protocols, like the
Clean Development Mechanism (CDM). Land-using activities appear to offer
considerable potential for GHG mitigation, with recent studies suggesting that land-based
mitigation could be cost-effective and assume a sizable share of overall mitigation
responsibility in economically optimal abatement and climate stabilization policies. The
non-climate implications are also important—social welfare (e.g., food security, clean
water access), environmental services (e.g., water quality, soil retention), and economic
welfare (e.g., output prices and production). Agriculture and forestry are also considered
to be particularly susceptible to climate change, as the relative productivity of lands
change with changes in plant productivity, weather variability, and disturbances (IPCC,
2007). To date, modeling has not been able to fully account for the opportunity costs of
alternative land-uses and land-based mitigation strategies, which are determined by
heterogeneous and dynamic environmental and economic conditions of land and
economy-wide feedbacks that reallocate inputs, international production, and consumers’
CGE economic models are well suited to evaluate these kinds of tradeoffs. However,
existing CGE frameworks, regional and global, are not structured to model land use
alternatives and the associated emissions sources and mitigation opportunities. This work
has been hindered by the lack of data; specifically, consistent global land resource and
non-CO2 GHG emissions databases linked to underlying economic activity and GHG
emissions and sequestration drivers. Recent development of global land-use and
emissions data, as well as new mitigation cost data, have provided a solid foundation for
advancing global economic land modeling.
The papers in this volume are a compendium of methodological insights and advances for
CGE modeling of land-use. They identify challenges likely to confront most land-use
modelers, and then describe strategies for addressing the challenges. While the papers do
not emphasize new theoretical developments, they are mindful of recent such
developments – for example in economic geography. The chapters presented here are
breakthroughs in model development and application: applying consolidated, well-
established methods, to novel datasets. Only now do we have access to globally
consistent databases of land use, coupled with data on the physical characteristics of the
land and the environment, and integrated with global economic activity. The new data
have facilitated model development and, in so doing, they have unearthed
implementation issues. The chapters describe and address the issues. Despite differences
in focus, many of the chapters are forced to overcome similar obstacles, though
sometimes via alternative approaches.
The book outlines key empirical and analytical advances and issues associated with
modeling land use and land use change in the context of global climate change policy. It
places special emphasis on economy-wide domestic and international competition for
land and other resources; and, the explicit modeling of land management options, versus
land use change. By offering a synthesis and evaluation of a variety of different
approaches to this challenging field of research, this book hopes to serve as a reference
for future work in the economic analysis of regional and global land use, and climate
This chapter begins by reviewing the previous literature, including methodological issues,
the role of land in climate change, and opportunities for land-use in climate policy. It then
provides an overview and synthesis of the important contributions of each chapter and the
challenges they address, as well as identifying directions for future modeling of global
land use within the context of the climate change debate.
2.1.1. Land-use modeling and climate change
Sound economic modeling of land-use is fundamental to improving climate change
policy analysis. Land use is important to climate change in three ways. First, land use
patterns influence emissions of greenhouse gases, the water cycle, and the surface albedo
(see Section 2.2). These matters are increasingly important in climate models, and the
demand for realistic land use scenarios is rising. Spatially explicit land use will be one of
the main issues in the eventual coupling of economic models and biophysical earth
Land use is also important in assessing the impacts of climate change. Changes in
temperature, precipitation, extreme weather, and atmospheric concentrations affect the
productivity of land, overall production opportunities, and carbon stocks (and the carbon
cycle) on managed and unmanaged lands (see Section 2.2). Explicit modeling of land use
is essential for characterizing net damages from climate change that include adaptation
responses and costs such as changes in irrigation, fire management, pest control, and
Finally, land use is important in greenhouse gas emissions reduction, particularly because
of emissions from farming activities, such as livestock and paddy rice production, carbon
sequestration in forests, and bio-energy crops (see Section 2.3). All compete for land and
water, with implications for food crops, nature conservation, energy use, and wood
product supplies. Modeling competition for land requires a characterization of land
qualities, which needs spatial inputs. For most practical purposes, however, this
characterization changes rather slowly so that land heterogeneity can be reduced to a
finite number of uniform classes (as in Chapter 6)1 or a probability distribution (as in
2.1.2. Land use in economics
Economics is Greek for the laws of the household – particularly the traditional farm, for
which subsistence and barter were more important than trade. The first systematic
economists, the Physiocrats, argued that agriculture is the primary sector. We still use
that moniker, but only in a numeric sense. The Physiocrats thought that land was the only
true source of value. The Classical economists similarly placed substantial emphasis on
agriculture and land. David Ricardo’s (1817) notion that land rents reflect land quality is
hotly debated today (Mendelsohn et al., 1994; Darwin, 1999; Schlenker et al., 2005,
Timmins, 2006). Economic textbooks still explain decreasing returns to scale with the
example of additional farmhands on a field – Turgot’s (1793) intensive margin. Steuart’s
(1767) extensive margin, another source of decreasing returns to scale, partly explains
land use patterns. Similarly, externalities are often introduced with the example of the
beekeeper and the farmer. However, as agriculture is a minor economic sector in the
industrialized countries nowadays, it is the vertically integrated, manufacturing sector
that serves as the canvas upon which much of the modern economic theory is painted.
(See Hubacek and van der Bergh (2006) for more historical perspective.)
Land use has attracted even less attention. In the Classical literature, there is Von
Thuenen’s (1826) model of farm specialization as a function of the distance to town, and
Zipf’s Law on the relative size of cities (e.g., Duranton, 2006).3 In economic geography,
there are the location analysis of Weber (1909) and the central-place theory of Christaller
(1933) and Loesch (1940). But land use plays a minimal role in current economic theory.
1 GTAP Working Paper No. 44
2 GTAP Working Paper No. 45
3 Zipf (1935) specified the functional form, but only applied it to the relative frequency of words.
Partly, this may be because land is – for many purposes – a minor problem. The built-up
environment, and hence more than 95% of the world economy, occupies less than 1% of
the land surface (Gruebler, 1994). Distance matters, of course, but can be parameterized
(e.g., as an iceberg; Samuelson, 1954) without an explicit two-dimensional model of the
land surface. Krugman (1998a) offers another explanation: City formation can only be
explained by a combination of congestion and agglomeration externalities. As
agglomeration implies increasing returns to scale, city formation resisted rigorous
analysis before the monopolistic competition revolution (Dixit and Stiglitz, 1977).
This has now changed (Krugman, 1991; Fujita et al., 1999; Brakman et al., 2001). New
economic geography offers micro-founded, general equilibrium models of activity
location (Krugman, 1998b). See Martin (1999) for a defense of the “old” economic
geography of Isard (1954) and Henderson (1974), which is less rigorous but more
empirical. New international economics has its roots in the monopolistic competition
revolution too. Rossi-Hansberg (2005) shows that, in a spatial model, tariffs and transport
costs are different – tariffs are step changes, whereas transport costs are continuous.
Bioeconomics is also gradually going spatial (Sanchirico and Wilen, 1999), particularly
in the investigation of marine protected areas (Smith and Wilen, 2003). General
equilibrium models of ecosystems are now emerging (Tschirhart, 2000, Finnoff and
Tschirhart, 2003) and have recently been extended to include land – as a factor of
production not for humans, but for plants (Eichner and Pethig, 2006). As promising and
exciting as these developments may be, this is theory only (Neary, 2001) – with but a few
empirical tests (Brakman et al., 2006). Although this work is increasingly applied (e.g.,
Stelder, 2005), practical applications cannot be expected in the near future – at least, not
in the sense of a well-calibrated, global model that can be used for numerical questions
about land allocation, either in climate policy or other contexts. Empirical analysis and
operational models particularly suffer from the lack of data. Nordhaus (2006) is a first
step towards spatially explicit economic data. In the meantime, spatial down-scaling
techniques are also being developed to generate gridded socio-economic data from
regional aggregates (e.g., Asadoorian, 2005; Grübler et al., 2007). Validating these
methods against actual spatial data will be necessary for assessing performance and
reducing uncertainties across alternative methods.
2.1.3. Geographic models of land use
Geographers obviously have a keen interest in land use, but mathematical analysis and
numerical models are not core tools for much of this work. Most geographic models are
small-scale, often limited to a small part of a country, and cannot be generalized; indeed,
many geographers resist generalization and large-scale research. Nonetheless, there are a
few large-scale models of land use. Heistermann et al. (2006) distinguish between
statistical models and rule-based models.
CLUE is a prominent example of a statistical model (Veldkamp and Fresco, 1996). Most
of the equations in the model are estimated by multiple regression, but the model is
completed by rule-based competition and transition. The largest scale applications of
CLUE are for China (Verburg et al., 1999) and for tropical South America (Wassenaar et
al., 2007). The latter model works at the impressive spatial resolution of 3x3 km.
SALU and IMAGE are examples of rule-based models. For instance, demand-driven
expansion of agricultural production is met on the basis of a suitability ranking, based on
soil, climate, distance, and so on. The trade-off between infra- and extra-marginal
expansion is modeled in a similar way. The SALU model (Stephenne and Lambin, 2001,
2004) is restricted to the Sahel, but the IMAGE model (Alcamo et al, 1998) is global with
a spatial resolution of 0.5ºx0.5º.
In ACCELERATES (Rounsevell et al., 2003) and KLUM (Ronneberger et al., Chapter
12)4, the rules are derived from profit maximization. In both cases, a risk-averse farmer
maximizes profits given fixed prices of inputs and outputs to the land-using sectors, and a
probability distribution of yields.
2.1.4. Economic models of land use
Partial equilibrium models are based on the same optimization principles as KLUM, but
include both the response of production and consumption to prices, as well as the
adjustment of these prices to attain a global equilibrium between supply and demand for
selected commodities. Examples are IMPACT (Rosegrant et al., 2002) and WATSIM
(Kuhn, 2003) for agriculture; GTM (Sohngen et al., 1999) for forestry; and AgLU (Sands
and Leimbach, 2003; Sands and Edmonds, 2005) and FASOM (Adams et al., 1996;
USEPA, 2005) for both agriculture and forestry. The distinct advantage of a partial
equilibrium model is that it captures price dynamics in the land-using sectors and is able
to model substantial spatial and land management detail. The disadvantage of partial
equilibrium models is that the rest of the economy is ignored. General equilibrium
models do not have this problem.
The first global Computable General Equilibrium (CGE) model with land use
disaggregated by physical characteristics was the FARM model by Darwin et al. (1995).
Land, typically treated as a non-tradable endowment in CGE models, was split into a
number of different land categories, distinguished by length of growing period. Land
endowment by category was an aggregate taken from a spatially explicit bioclimatic
model. The model was used to estimate the impacts of climate change (Darwin et al.,
1995), of sea level rise (Darwin and Tol, 2001), and of nature conservation (Darwin et
al., 1996) – each of which changes the relative land endowments in various regions of the
world. Changes in demand for land were met, in the spatially-explicit biophysical model,
on the basis of rules, but not on the basis of optimal behaviour. That is, Darwin et al.
(1995) brought biophysical realism into their economic model, but they did not bring
economic realism into their biophysical model.
GTAP-L (Burniaux, 2002) extends, the work of Darwin et al. (1995) with explicit
tracking of the transformation of land (from one crop to another); thus, competition
between alternative land uses was introduced. However, the input data were rudimentary.
4 GTAP Working Paper No. 50
The GTAP-AEZ model (Chapter 6)5 extends the earlier work with a more extensive land
use data base, and a more sophisticated representation of land-based emissions and forest
Atmospheric composition and climate are affected by land cover and land use changes
via biogeophysical and biogeochemical mechanisms. Biogeophysical mechanisms, such
as the effects of changes in surface roughness, transpiration, and albedo, are thought to
have had a global cooling effect over the past millennium (Brovkin et al., 1999), while
biogeochemical effects from direct emissions of GHGs into the atmosphere have a global
warming effect. Cumulative emissions from historical land cover conversion for the
entire industrial period 1850–2000 accounted for roughly a third of total anthropogenic
carbon emissions over this period (Houghton, 2003). In addition, land management
activities were estimated to be responsible for over half of global anthropogenic methane
(CH4) and over three quarters of nitrous oxide (N2O) emissions in 2000 (USEPA, 2006a).
Land management activities include, among other things, cropland fertilization and water
management, manure management and changes to forest rotation lengths.
Land use and climate change
However, until recently, projected changes in land use were not explicitly represented in
carbon cycle modeling (Cramer et al., 2004; House et al., 2002; Levy et al., 2004).
Recent studies have shown that land use (e.g., Brovkin et al., 2006; Matthews et al.,
2003; Gitz and Ciais, 2004) and feedbacks in the society-biosphere-atmosphere system
(e.g., Strengers et al., 2004) must be considered for realistic estimates of the future
development of the carbon cycle. Long-run integrated assessment models have begun to
explicitly model land use. One example is offered by Tubiello and Fischer (2007) who
utilize the AEZ model of IIASA and FAO to analyze the impact of climate change on
agricultural productivity. This framework uses detailed agronomic-based knowledge to
simulate potential yields and the availability and use of land as a function of climate.
They develop estimates of changes in productivity at the level of 2.2 million grid cells
and then aggregate these to serve as productivity shocks to the Basic Linked System
(BLS) economic model of world trade. However, there is no interaction between the
economic model and the land use model. In general, economic frameworks for capturing
dynamic and heterogeneous global land use decisions are still rather immature. Much of
the work reflected in the chapters of this book was motivated by the need for theoretically
sound and consistent global frameworks for modeling land use.
Land use is driven by the demand for land based products and services (e.g., food, timber,
bio-energy crops, and ecosystem services) and land use production possibilities and
opportunity costs (e.g., yield improving technologies, temperature and precipitation
changes, and CO2 fertilization). Non-market values will also shape land use outcomes,
both use and non-use values, such as environmental services and species existence values
respectively. Total world food consumption is expected to increase by more than 50% by
2030 (Bruinsma, 2003) with significant structural change in consumption patterns,
including increases in global per capita meat consumption, on the order of 25% by 2030,
5 GTAP Working Paper No. 44
with faster growth in developing and transitional countries of more than 40 and 30%,
respectively (Bruinsma, 2003; Cassman et al., 2003). Additional cropland is expected to
be required to support these projected increases in overall food and livestock feed
Technological change is a critical driver of land use, and a critical assumption in
projections. For example, Sands and Leimbach (2003) suggest that globally 800 million
hectares of cropland expansion could be avoided with a 1.0% annual growth in crop
yields. The Millennium Ecosystem Assessment (MEA, 2005) scenarios project positive
but declining crop productivity growth over time due primarily to diminishing marginal
technical productivity gains and environmental degradation.
Ludena et al. (2007) examine patterns of agricultural productivity growth over the past 40
years, decomposing it into that portion due to outward movement in the technological
frontier, and that due to “catching up” to the global frontier. Typically the latter is more
important for developing countries. This historical analysis of productivity growth in
global agriculture forms the basis for projections to 2040. They anticipate continued
differential growth rates between developed and developing countries, as well as across
sub-sectors. Developing countries are expected to show the fastest productivity growth in
non-ruminant livestock, while industrialized countries are expected to continue with
faster productivity growth in crops. This difference is fueled in part by the massive
private sector R&D investments being made in the rich countries, which now dwarf
public sector investments. Ruminant livestock productivity growth shows a strong
tendency towards divergence, with high growth in developed and some developing
countries and negative productivity growth anticipated in others. Since increasing
(decreasing) net productivity per hectare results in reduced (increased) cropland demand,
this work suggests that the derived demand for land is likely to grow at very different
rates in different parts of the world. To the extent that rapid productivity growth relieves
the pressure for deforestation, this could have very significant impacts on CO2 emissions.
This “footrace” between long run supply and demand for land use is the subject of
Chapter 106 in this volume.
While land use shapes climate change, climate changes will also shape future land use.
Rising temperatures and CO2 fertilization could improve regional crop yields in the near
term, thereby reducing pressure for additional cropland and deforestation. However,
modeling the productivity impacts of climate change is not straightforward with, among
other things, the distributions of precipitation, weather extremes, and disturbances
difficult to model, as well as interactions with other environmental variables, e.g., air
quality, soil erosion (IPCC, 2007). Current land use modeling generally considers, if
anything, only CO2 fertilization and changes in annual average global temperature (e.g.,
Clarke et al., 2007; van Vuuren et al., 2007).
Figure 1 presents recent baseline land-use projections for global cropland, forest land,
and grazing land (IPCC, 2007). Most global scenarios—from integrated assessment,
computable general equilibrium, and sectoral modelling—project significant changes in
agricultural land use caused primarily by regional changes in food demand and
production technology. Scenarios with larger amounts of agricultural land result from
6 GTAP Working Paper No. 48
assumptions about higher population growth rates, higher food demands, and lower rates
of technological improvement. Drawing on the studies represented in Figure 1, projected
cropland changes vary from -18 to +69% by 2050 relative to 2000 (-123 to +1158 million
hectares) and forest land changes range from -18 to +3% (-680 to +94 million hectares)
by 2050. However, most of the long-term scenarios assume that forest trends are driven
almost exclusively by cropland expansion or contraction, and only deal superficially with
driving forces such as global production, consumption and trade in agricultural and forest
products and conservation demands. In these scenarios, biomass crops are not projected
to play a large role in global business-as-usual land cover. Higher long-run energy price
expectations (due to the pricing of carbon, greater economic scarcity, energy security
initiatives, or other forces) would, ceteris paribus, create pressure to expand biomass
Increasing GHG emissions are projected in the near future for CO2 (Figure 2) and over
the long-term for non-CO2 GHGs (IPCC, 2007). Net deforestation pressure is projected to
decrease over time as population growth slows and crop and livestock productivity
increase; and, despite projected losses of forest area in some scenarios, carbon uptake
from afforestation and reforestation results in net sequestration. Recent non-CO2 GHG
emissions baseline scenarios agree that agricultural CH4 and N2O emissions increasing
until the end of this century, potentially doubling in some baselines. However, the
modeling of forest and agricultural emission sources and carbon sinks varies across
scenarios with limited consideration of explicit emissions drivers and the actual
production trade-offs that affect land use, emissions, and production costs. Accordingly,
this issue will be taken up in considerable detail in Chapter 67 of this volume.
Land-use practices can be modified to mitigate GHG emissions, such that they reduce
emissions of CO2, CH4 and N2O, increase sequestration of atmospheric CO2 into plant
biomass and soils, and/or produce biomass fuel substitutes for fossil fuels. Forests alone
have substantial potential to sequester carbon (Watson et al., 1995; Watson et al., 2000;
IPCC, 2001). In addition, current technologies are capable of substantially reducing CH4
and N2O emissions from agriculture (USEPA, 2006b, Section V), while a number of
global biomass energy potential assessments have been conducted (see Berndes et al.
(2003) for an overview).
Land use and climate policy
Recent climate stabilization studies have found that land use mitigation options could
provide cost-effective abatement flexibility in achieving climate stabilization targets
(Rose et al., 2007), accounting for 15 to 40 percent of cumulative required abatement
over the century. Four particular studies found that including land-based mitigation (both
non-CO2 and CO2) reduced the costs for stabilizing radiative forcing (Kurosawa, 2006;
van Vuuren et al., 2006; Rao and Riahi, 2006; and Jakeman and Fisher, 2006).
Rose et al. (2007) showed that annual forestry, agriculture, and biomass abatement levels
are projected to grow over time in stabilization scenarios with relatively stable annual
7 GTAP Working Paper No. 44
increases in agricultural mitigation and gradual deployment of biomass mitigation that
accelerates dramatically in the last half of the century to become the dominant land
mitigation strategy. In some scenarios, increased commercial biomass energy (solid and
liquid fuel) provides 5 to 30% of cumulative abatement, and 1 to 15% of total primary
energy (500 to 9,500 EJ of additional bio-energy above the baseline over the century).
However, there are substantial uncertainties. There is little agreement about the
magnitudes of abatement (Figure 4). The scenarios disagree about the role of agricultural
strategies targeting CH4 versus N2O as well as the timing and annual growth of forestry
abatement, some scenarios suggesting substantial early deployment of forest abatement,
while others suggesting gradual annual growth or increasing annual growth. Furthermore,
while there is some indication that agricultural mitigation is projected to be a larger share
of the developing countries’ total mitigation portfolio; and, developing countries are
likely to provide the vast majority of global agricultural mitigation, it is currently not
possible to assess the regional land-use abatement potential in stabilization scenarios
given the scarcity of published regional results (IPCC, 2007).
Overall, the modeling of global land based climate change mitigation is relatively
immature with significant opportunities for improving baseline and mitigation land use
scenarios and better characterizing the emissions and mitigation potential of land.
Essential to future land modeling are improvements in the dynamic modeling of regional
land use and land-use competition. The cost of any land based mitigation strategy should
include the opportunity costs of land, which are dynamic and regionally unique functions
of changing regional bio-physical and economic circumstances. Subsequent development
efforts should address competition between mitigation options, as well as modeling of the
implications of climate change for land-use and land mitigation opportunities, including
the potential climate driven changes in forest disturbance frequency and intensity. These
challenges provide the motivation for this book, and we now turn to the contributions
made by this volume.
3. Contributions of this Volume and Challenges for Future Work
This book provides a unique reference on recent key methodological advances in the
analysis of global land use as it relates to climate change policy. It also provides a guide
for implementation by others. In this section, we discuss some of the most challenging
issues facing authors in this volume, as well as potential future directions for research in
As noted previously in this chapter, the spatial dimension is at the very heart of land use
modeling. With global land use data now available at the 0.5 degree grid cell level (see
Chapter 28 of this volume) there is ample scope for dramatically increasing the
dimensions of any global model. If such a model also endeavors to cover all economic
The Spatial Dimension
8 GTAP Working Paper No. 40
activity, not just agriculture or just forestry, then the data and modeling requirements
become truly overwhelming at this level of resolution. As a consequence, the chapters in
this book have focused exclusively on land use in the agriculture and forestry sectors.
Furthermore, most of the chapters have adopted a somewhat aggregate level of
resolution, which is more in line with the spatial resolution of economic statistics. For
example, in the MIT, LEITAP and KLUM models (chapters 89, 910 and 1211,
respectively), global results of land allocation across crops are produced at the country
level. In the GTAP-AEZ model (chapters 612 and 1013), land use is aggregated to the level
of Agro-Ecological Zones within countries. Even the Sands-Kim study, which focuses
solely on the US (Chapter 714), aggregates to the level of major watersheds.15
A fundamental problem in modeling agriculture and forestry production at the sub-
national level involves estimation of input usage and production by spatial unit. The
GTAP-AEZ model circumvents this problem, by having a single, national production
function in which land types from different AEZs substitute for one another. This begs
the question: Is this an appropriate approach to modeling land use? In their chapter,
Hertel et al. (Chapter 616) show that this is a legitimate approximation to a model in
which production on each AEZ is modeled separately, provided that: (a) the sub-sectors
(i.e., different AEZs) produce identical products, (b) non-land input-output ratios are the
same across AEZs, (c) common non-land input prices prevail across AEZs, and (d) the
elasticity of substitution between AEZs in a given land use is set very high. These
assumptions, in combination with cost minimization and zero pure profits, mean that land
rents must vary in direct proportion to yields. This is the same assumption that Eickhout
et al. make in constructing their land supply schedule (Chapter 917). In light of the central
role of the national production function assumption in many of the chapters in this
volume, it would be useful to test the requisite maintained hypotheses for key countries,
using disaggregated data on inputs and prices. Of particular interest is the extent to which
non-land input-output ratios vary systematically with AEZs, either due to differences in
choice of technique across different land qualities or due to differing input prices. If this
proves to be the case, then the simple rule of proportionality between yields and land
rents, as well as the capacity of an aggregate production function to capture the impact on
the derived demand for land, are both brought into doubt.
Closely related to the spatial dimension is the issue of the homogeneity of land and its
potential mobility across uses. If the unit of observation is small enough so that for all
practical purposes the land is perfectly homogeneous, then we would expect rental rates
Mobility of Land Across Uses and Diversification of Production
9 GTAP Working Paper No. 46
10 GTAP Working Paper No. 47
11 GTAP Working Paper No. 50
12 GTAP Working Paper No. 44
13 GTAP Working Paper No. 48
14 GTAP Working Paper No. 45
15 In other work, the KLUM model has been calibrated to crop production data at the grid-cell in Europe.
16 GTAP Working Paper No. 44
17 GTAP Working Paper No. 47
on all land within that unit to be equalized. In the absence of risk and uncertainty, and in
the absence of technological interdependence amongst the crops (e.g., benefits from crop
rotation or the sharing of common inputs), we would expect farms to specialize in the
crop with the highest return, net of non-land input costs. However, farms are often
diversified, and certainly most of the larger units of observation (e.g., grid cells or AEZs)
exhibit diversification of production. Explaining this diversification therefore presents the
modeler with a special challenge. The authors in this volume take two different
approaches to reconciling this puzzle. The first approach is to appeal to risk
considerations. This is the approach taken in KLUM (Chapter 1218), where risk averse
producers maximize expected utility and returns to different crops are uncertain. This
combination of factors leads farmers to diversify production. The authors use half the
data in their time series from the FAO to calibrate the risk aversion and cost parameters
for their model. They reserve the second half of the data series for model validation and
find that the model performs reasonably well in this out-of-sample test. Such model
validation is extremely valuable, and should be more widely undertaken by authors in this
Of course, risk aversion is a producer-level issue, not a market-level issue. So when we
move to the level of regions, or indeed countries, the appeal of a risk-based approach to
model calibration is somewhat lessened. For such large areas, it would seem that
diversification likely reflects heterogeneity of the underlying land and climatic
endowments, as well as the heterogeneity of local markets. For example, it may be
attractive to (e.g.) grow certain crops in the valley and others on the hillside. So physical
heterogeneity is a reason why we might observe diversification in crops within a given
This brings us to a group of diverse approaches to reconciling observed patterns of
production with the economic structure of global land use models. There are two
approaches used in the book for dealing with land heterogeneity. The first is to employ a
simple Constant Elasticity of Transformation (CET) function by which an aggregate
endowment of land is transformed across alternative uses, subject to some transformation
parameter that governs the responsiveness of land supply to changes in relative yields.
This approach was first introduced by Hertel and Tsigas (1988) in their agriculture-
focused CGE model of the US economy and it has subsequently been used in the
standard GTAP model (Hertel, 1997) to handle the allocation of land across sectors in the
economy. This approach is also embedded in several of the chapters in this book (Hertel
et al., Chapter 619; Golub et al., Chapter 1020; Eickhout et al., Chapter 921). The problem
with the CET approach is that the “transformation” of land from one use to another
destroys the ability to track the allocation of hectares across agricultural activities.
Instead of constraining the sum of hectares across uses to equal the total availability of
hectares in a given AEZ or country, the CET function constrains the land rental share-
weighted sum of hectares to equal the total endowment of land. In this framework,
differential land rents reflect differences in the effective productivity of a given hectare of
18 GTAP Working Paper No. 50
19 GTAP Working Paper No. 44
20 GTAP Working Paper No. 48
21 GTAP Working Paper No. 47
land across uses and it is these effective hectares that are constrained in the aggregate.
Also, given the lack of an explicit link to yields and the underlying heterogeneity of land,
this model is difficult to validate against the observed data. In short, while it is an
extremely versatile approach to limiting factor mobility across uses, the CET function
“covers a multitude of sins”. A more explicit approach to handling land heterogeneity
would be desirable.
The AgLU model, featured in Chapter 722 by Sands and Kim, also reflects land
heterogeneity in its attempt to reproduce a diversified mix of output in a given AEZ (or in
their case, a given watershed). This methodology, developed by Sands and Leimbach
(2003) and inspired by the work of Clarke and Edmonds (1993) on heterogenous energy
technologies, is isomorphic to the CET approach in its representation of maximum profits
and supply response. However, unlike the CET function, their framework is based on an
explicit model of yield heterogeneity. Indeed, in this model, the variance parameter
(possibly adjusted for the correlation of yields across crops), determines the extent of the
supply response (see the appendix in Sands and Leimbach, 2003). And it can be shown
that there is a direct mapping from these parameters to the transformation parameter in
the CET function. Of course, as with any practical approach to modeling complex
phenomena, this one involves some restrictive assumptions, in particular, the form of the
underlying yield distribution (log-Gumbel). The supply function also implies that, at the
margin, land rents (profits in their terminology) are equated across uses. This is a testable
hypothesis that warrants econometric investigation.
In Chapter 7, Sands and Kim calibrate AgLU to watershed data by permitting prices and
intrinsic yields to vary in order to satisfy the model’s equations when evaluated at
observed yields and land cover shares. So a natural way to validate the model would be to
work in the opposite direction, for a region where such data were available. Another
validation approach would be to estimate the log-Gumbel distribution directly and
compare the supply elasticity implied by these distribution parameters to those obtained
directly from estimation of land supply functions. In short, the appeal of the AgLU
approach is that it has strong, and observable implications, which may be tested against
In the case of any of the models based on land heterogeneity, authors typically “nest” the
supply functions such that producers first determine the allocation of land amongst crops,
then, based on the average return to crop land, an allocation is made between crops and
livestock or crops and forestland (e.g., Sands and Kim, Chapter 1023). In the case of
Eickhout et al., in Chapter 924, certain crops are singled out for special treatment in the
land supply nesting structure. As documented in Chapter 10 by Golub et al., this pattern
of nesting can have important implications for the long run supply of land to different
uses, as well as the path of land rents over time. Yet there are as many different patterns
of nesting as there are models, and little evidence in favor of one pattern over another.
This type of nesting, or separability, is amenable to econometric testing (e.g., Berndt and
Christensen, 1974). It amounts to a restriction on the cross-elasticities of supply between
land rents in one nest and land supply in another nest. Rigorous testing of these
22 GTAP Working Paper No. 45
23 GTAP Working Paper No. 48
24 GTAP Working Paper No. 47
separability hypotheses would greatly assist future authors in narrowing the range of
acceptable land supply nests.
One of the most difficult challenges faced by authors in this volume is that of
incorporating forestry into their analysis of climate change mitigation. In their chapter on
this topic (Chapter 1125), Sohngen et al. outline the problems and challenges of
adequately representing forestry in economic models of land use and land use change.
Unlike most other production processes in the economy, which can be adjusted within a
matter of a few years, it takes decades to grow a new forest. Furthermore, growth in the
forest stock, as well as sequestration potential, depends critically on the type of forest and
its vintage. There are very few global forestry models that handle all these aspects well in
partial equilibrium. So expecting proper treatment within general equilibrium in the near
future is probably asking too much. However, the authors in this volume have explored a
number of different approaches to tackling important (and different) pieces of the
problem. And we turn now to a discussion of these diverse approaches.
Perhaps the most obvious approach is to establish a soft link between an intertemporal,
partial equilibrium model of forestry and the CGE model. This is what is done in Golub
et al. (Chapter 1026). First the recursive, dynamic CGE model is run for 100 years to
establish the baseline path for the economy, and, in particular, the growth rate in
aggregate demand for forest products.27 Based on this path of global forest products
demand over the next century, Sohngen runs his Global Timber Model and produces a
price path for forest products. This price path embodies all of the forward-looking
behavior in the intertemporal, partial equilibrium forestry model and serves to inform the
CGE model about the change in value of forest products over time. The recursive
dynamic CGE model is then calibrated to follow this same price path over time. (The
calibration variable is the – as yet unobserved – rate of technological change in the forest
products using sector.) This approach is attractive due to its relative simplicity in
transmitting information about the intertemporal adjustment in the forestry sector over
time. However, the CGE results clearly depend on the method of calibration and more
work is required to explore alternative approaches and their implications.
Incorporation of the Forestry Sector in Global Models of Land Use
A similar, but somewhat more ambitious approach to linking the partial and general
equilibrium models with respect to forestry behavior is offered by Hertel et al. in Chapter
628. Those authors focus on deviations from baseline (in this case the baseline is simply
the current state of the economy, since the analysis is comparative static). In particular,
they consider the impact of alternative carbon prices on climate change mitigation. In the
case of forestry, this involves three major mechanisms for responding to the carbon price:
(a) averted deforestation, (b) afforestation, and (c) optimal changes in forest management
practices, including aging of the forest stock and use of more non-land inputs in more
25 GTAP Working Paper No. 49
26 GTAP Working Paper No. 48
27 Since the macro-economic baseline is largely independent of what is assumed about forestry, this is a
one-way flow on information (i.e. the authors do not iterate between the two models on this point).
28 GTAP Working Paper No. 44
intensive management practices. The authors begin by running the partial equilibrium
model many times in order to trace out the sequestration supply schedules for the forestry
sectors in each region of the global model. The sequestration is decomposed into the
response at the extensive margin (more land in forests), which has a direct impact on the
other land-using sectors, and the response at the intensive margin (more carbon in the
existing forest lands), which reflects optimal adjustments on existing forest land. The
CGE model is then calibrated so that it mimics both the intensive and extensive carbon
sequestration schedules. The authors find this to be a reasonably effective means of
incorporating the sequestration potential of forestry into a more comprehensive analysis
of climate change mitigation. Compared to other CGE treatments in this book, the most
important forestry element of the Hertel et al. approach seems to be the distinction
between the intensive and extensive margins. These have very different implications for
land use, and ignoring the intensive margin, as many integrated assessment models do,
tends to over-emphasize the competition for land in the wake of sequestration subsidies.
The biggest problem in the Hertel et al. analysis relates to the fact that the forest carbon
sequestration supply schedules shift over time. In Chapter 629, they focus on the 20 year
abatement schedules. However, if they had worked with 50 year schedules, the answer
would have been different (more sequestration at lower costs) due to the greater
opportunities for intertemporal adjustment.
In Chapter 730, Sands and Kim seek to incorporate key features of the forestry problem
directly into a recursive dynamic framework. These authors derive a steady-state
condition for the determination of optimal forest rotation and embed this in their CGE
model of the US economy. This has the virtue of capturing the impact of carbon
sequestration subsidies on the optimal timber rotation, thereby capturing an important
part of the intensive margin response to carbon prices. However, in so-doing, these
authors ignore the issues of vintages and adjustment paths. Clearly some combination of
these two approaches will be required on the path forward for improving upon the current
state of this literature.
A critical issue in modeling the long run supply of land to different activities in
agriculture and forestry is the availability of new lands that might be brought into
production. The simplest way to handle this problem is to construct a land supply
schedule in which rising land rents causes additional land to be brought under cultivation.
This is the approach adopted by Eickhout et al. in their specification of the LEITAP
model (Chapter 931). The appeal of their approach lies in the way they build up this
supply schedule. In particular, they capitalize on the detailed productivity information
available in the IMAGE data base. For each region, they first remove all the lands which
are: (a) non-productive, or (b) unavailable for conversion to agriculture (i.e., protected or
built-up). The remaining lands are thrown into a pool and subsequently arranged in order
of diminishing productivity. The authors then invoke the assumption that land rents are
29 GTAP Working Paper No. 44
30 GTAP Working Paper No. 45
31 GTAP Working Paper No. 47
Accessing New Lands
inversely related to yields, which gives them a land supply schedule where the total
amount of land in production is an increasing function of land rents. As equilibrium land
rents rise, the model brings in additional lands up to the point where the benefit of the last
hectare of land, as measured by its marginal value product, equals the “marginal cost” of
this land, i.e. the land rent that must be paid in the market place. This approach brings
detailed biophysical information in the integrated assessment model to bear on the issue
of land supply, with special areas, such a forest reserves and national parks omitted from
the commercial supply schedule. The problem is that there is no spatial component to the
supply decision – as might be the case if there were AEZs distinguished in the model.
Furthermore, the authors only consider the supply of land to agriculture and ignore the
use of land in forestry in their CGE model.
Indeed, much of the new land that might be brought into commercial production – either
agricultural or forestry – under future scenarios is currently covered with forest. Sohngen
and Mendelsohn (2007) term these “inaccessible forests”, i.e. forests that are not
economically accessible given current market conditions. Their modeling introduces a
supply function for these inaccessible lands and calibrates their model to bring in roughly
the same amount of these forest lands as has deforested over the past decade in key
regions of the world. This represents an important part of their baseline, and forestalling
this deforestation also becomes an important feature of the response to carbon prices in
In Chapter 1032 on the long run supply and demand for land, Golub et al., explore the
issue of inaccessible forests in considerable detail. They draw on the work of Gouel and
Hertel (2006), which formulates the access problem as an investment decision in which
the discounted stream of benefits of accessing an additional hectare is equated to the
marginal cost of access. By modeling the access costs explicitly (in this case as a function
of labor and capital), Golub et al. are able to close their general equilibrium model with
respect to these new lands. That is, they don’t just come in “out of the blue”. Rather, the
access of new lands requires real resources. These authors find that adding the
inaccessible lands makes a sizable difference in the long run scarcity of land in some
regions of the world.
In the literature review offered above, we emphasized the important role of biofuels in
long run climate change stabilization scenarios. Accordingly, this topic also receives
attention in the present volume. Chapter 833 by Reilly and Paltsev focuses specifically on
biofuels, but their analysis does not really delve into the implications for land use of
current biofuels programs. They focus primarily on bioenergy from cellulosic conversion
– something that remains a long way from commercial viability in current market
conditions. In Chapter 734, Sands and Kim, on the other hand, focus largely on biomass
for energy produced from crops. As such, they tie biofuels production explicitly to land
Biofuels and Land Use
32 GTAP Working Paper No. 48
33 GTAP Working Paper No. 46
34 GTAP Working Paper No. 45
use. They project a major role for biofuels for high carbon price scenarios (above
$100/TCE), but their analysis is focused only on the United States. In Chapter 935,
Eickhout et al. include biofuels in their global analysis, but the expansion of the biofuels
production is largely exogenous.
There appear to be two main obstacles to the incorporation of biofuels into global CGE
models. The first is simply the issue of data. In the case of biofuels, many of the
potentially important technologies (e.g., ethanol from cellulose) are not currently
commercially viable – so they don’t appear in our data bases at all! Introducing them into
the model requires coming up with an appropriate profile of costs, sales, and even trade
shares, to invoke when they would come into production. This is not a small task.
Secondly, there is the question of profitability – how high do competing energy prices
have to rise before these technologies enter commericial production? In their chapter36,
Reilly and Paltsev do a very nice job laying out the key assumptions in the case of “bio-
oil” and “bio-electric” technologies. However, they do so in a stylized example (e.g., the
feedstock in both cases is a “resource”). Taking this to a level of detail that bears directly
on land use will be more challenging; and, it is in the competition for land that the global
impact of biofuels production may be most significant. This fact is highlighted by the
work of van Vuuren et al. (2007), who project a massive increase in land devoted to
biofuels and hence continued rapid rates of deforestation in developing countries. Sorting
out how biofuels will compete with forestry and crops, particularly in the face of GHG
subsidies/taxes is a high priority for future work.
Agriculture is responsible for the vast majority of global methane and nitrous oxide
emissions and is considered to be important for managing the costs of climate change
mitigation policies. However, modeling of the detailed drivers of non-CO2 emissions
within agricultural production and internalization of the costs of non-CO2 emissions
mitigation is lacking in current analyses of climate change policy. This, in turn, frustrates
accurate evaluation of the mitigation potential of agricultural non-CO2 strategies. In
Chapter 537, Rose and Lee introduce an emissions data base linked to the GTAP global
economic data base that can support this kind of detailed analysis of non-CO2 emissions.
And the chapter by Hertel et al. (Chapter 638) utilizes these data to provide a more
realistic and comprehensive picture of non-CO2 emissions—with economic sectors
generating multiple emissions fluxes from various points of production and mitigation
possible through explicit substitution between more and less emissions intensive input
mixes (vs. output reduction).
Agricultural greenhouse gas emissions and mitigation
There are three key challenges to modeling agricultural emissions and mitigation. First,
agricultural soil carbon stock and flux modeling is noticeably absent from current
approaches; and, agricultural soils are thought to offer substantial carbon sequestration
35 GTAP Working Paper No. 47
36 GTAP Working Paper No. 46
37 GTAP Working Paper No. 43
38 GTAP Working Paper No. 44
potential (IPCC, 2007). This absence is, among other things, due to a lack of global
spatial soil carbon stock data and the difficulty of modeling changes in soil carbon, which
calls for an additional modeling tool as complex as a crop process model or as simple as a
spreadsheet carbon pool accounting tool. Second, tracking net greenhouse gas effects
from land conversion (e.g., forestry to agriculture) requires knowledge of previous and
future land use, and, for soil carbon stocks, historic land use. Tracking the evolution of
every parcel of global land is challenging given current technical capacities. However,
land transitions should not be ignored and simplifying strategies are needed to provide a
reasonable approximation of the net greenhouse gas implications of land conversion. A
notable example of such an approach is the LEITAP link to the IMAGE model (Eickhout
et al., Chapter 939), where regional agricultural land decisions are downscaled to grid
cells, with above and below ground soil carbon stocks adjusted and emissions based on
forest land conversion and agricultural land abandonment (see Leemans et al., 2002, for
details on this particular element of their modeling). Finally, technological change will
alter the emissions rates of agricultural production activities. Explicit consideration of
this interaction is important to avoid arbitrary emissions growth and explore emissions
uncertainties associated with technological uncertainty.
As noted previously, the chapters in this book studiously avoid dealing with demands for
land by the non-primary sectors, i.e., commercial, residential, recreational uses.40 Yet in
some parts of the world, these represent the main area of future growth in land use.
Furthermore, in some parts of the world, non-agriculture and non-forestry uses dictate
land values, and hence the opportunity cost of expanding these primary sectors. In the
U.S., nonagricultural uses have been shown to play a role in determining the value of
farmland in selected metropolitan areas (Lopez et al., 1988), but this has not proven to be
an important determinant of aggregate agricultural land values. However, in Japan, the
case is quite different. There, the proximity of farmland to major population centers is
much greater and arable/buildable land is extremely scarce. Thus the demand for
residential, recreational, and commercial land may be expected to exert considerable
pressure on land values in agriculture and forestry.
The Role of Non-primary Demands for Land
Of course, the degree to which land can move between primary sectors and other uses
depends on land use legislation. In Japan, landowners have historically been required to
obtain the permission of the prefecture or of the Ministry of Agriculture, Forestry, and
Fisheries in order to transfer farmland into other uses (ABARE, 1988, p. 75). Extremely
favorable property and inheritance taxation of farmland, coupled with high rates of
capital gains taxation, serve to further discourage movement of land into nonfarm uses.
As a result, the percentage of land devoted to agricultural uses in the three major
metropolitan areas in Japan (16 percent) exceeds the share of this land devoted to
residential, commercial, and industrial plant uses (11.5 percent). It also exceeds the share
of farmland in Japan's total land area (15 percent) (ABARE, 1988, p. 316). Despite these
39 GTAP Working Paper No. 47
40 The one small exception to this statement is the chapter by Eickhout et al., who net out such uses before
building their land supply curve.
distortions in the land market, there is evidence that nonfarm demands support
agricultural land values. For example, between 1979 and 1985 the price of rice relative to
the price of rice paddy land fell by about 20 percent (ABARE, 1988, p. 321), indicating
that those holding the land are likely focused on non-agricultural uses.
The first step in incorporating non-primary demands for land in CGE models involves
identifying its importance in the other sectors’ production functions. A natural place to
start would be the residential and commercial sectors, which are typically broken out in
CGE models and which are clearly land-intensive. In the long run the demand for land in
parks for recreation and the preservation of ecological diversity is likely to be very
important. But these sectors have not yet been well-developed in global CGE models.
Improving their specification, as well as estimating how the demand for their services is
likely to grow with higher incomes, will necessarily precede the incorporation of these
land demands into CGE models.
The role of global land use in climate change policy is extremely important. Yet, in
comparison to other areas, such as fossil fuel use, it has received relatively little attention
by climate economists – hence the need for this book. The chapters in this volume as a
group offer all the elements needed for a sound analysis of the role of agriculture and
forestry in greenhouse gas emissions and emissions reductions, as well as a structured
economic foundation for evaluating the net impact of climate change on these sectors.
The necessary data are now available in Part II of the book, and, Part III provides
applications that have been extended to study all parts of this issue. While none of the
individual frameworks presented here considers all aspects, in combination they do
achieve this goal. As such, they lay a solid foundation for a new generation of research in
Given the variety of different approaches to modeling global land use, there is need for
model testing and validation: Which of these approaches best fits the data? Which
approaches are based on maintained hypotheses that can be rejected, and therefore should
be abandoned? In this chapter we have discussed some approaches to model validation,
as well as key hypotheses to be tested. It is only through such systematic research that we
will be able to eliminate the least promising approaches and focus on those that are
worthy of further attention.
Throughout the book, we have largely ignored two “elephants in the global land use
room”. The first is that of spatially explicit analysis. Computable general equilibrium
models will not run any day soon on a spatial scale that satisfies natural scientists
modeling global change. The simplest approach involves proportional downscaling
(every unit within the aggregate inherits the same growth rate). But this ignores a great
deal of useful information and can result in absurd projections, e.g., where a developing
country ends up with GDP/capita twice as high as developed countries. Thus the current
compromise is to use a combination of optimization and statistical techniques to
downscale results of economic models to the grid. One of the most recent examples of
this work is the effort at IIASA which downscales global results in two steps (Grübler et
al., 2007). The first involves moving from aggregated regions to the country level based
on population projections and projected GDP growth rates. The economic growth rates
are based on the estimated “inverted U-shaped” relationship between GDP growth rates
and GDP per capita. The authors then solve a constrained optimization problem which
permits them to respect the regional constraints. At the second level, this approach to
down-scaling incorporates statistical relationships pertaining to rates of urbanization and
assumptions about the fundamentals under-pinning rural-urban income differentials, in
order to put some structure on the sub-national down-scaling of economic activity. These
authors are in the process of testing their down-scaling methodology against data-based
approaches, such as G-ECON developed by William Nordhaus (2006). Such comparisons
should help improve future downscaling methodologies and guide researchers to a
common standard, which may be used to downscale not only total economic activity, but
sectoral activity as well.
This book has also largely ignored water. Land is useless without water. Bio-energy and
food not only compete for land, but also for water. The models used in this book do not
omit water, but they treat it as a property of the land rather than as a scarce resource, that
can be traded and the value of which can be enhanced by investment. Rosegrant et al.
(2005) have developed a partial equilibrium model for analysis of global trade and water
issues based on 69 river basins. Berrittella et al. (2007) include water in a global
computable general equilibrium model – but their framework offers only a rudimentary
representation of land. Future research will need to integrate such analyses of land and
water into a single, global general equilibrium framework.
Omission of these two “elephants” notwithstanding, we think that this book represents a
useful step forward. It lays the practical foundations for economic policy analysis of land-
based greenhouse gas mitigation, and is therefore an important stepping stone to further
research in the economics of land use and climate change policy.
ABARE (Australian Bureau of Agricultural and Resource Economics), 1988, Japanese
Agricultural Policies: A Time of Change, Canberra: Australian Government Publishing
Adams, D.M., R.J. Alig, J.M. Callaway, B.A. McCarl, and S.M. Winnett (1996) “The
Forest and Agriculture Sector Optimization Model (FASOM): Model Structure and
Policy Applications.” USDA Forest Service, Pacific Northwest Research Station,
Portland, Oregon. Research Paper PNW-RP-495. 60 p.
Alcamo, J., E. Kreileman, M. Krol, R. Leemans, J. Bollen, J. van Minnen, M. Schaeffer,
S. Toet, and H.J.M. de Vries (1998), Global modeling of environmental change: an
overview of IMAGE 2.1, in J. Alcamo, R. Leemans, and E. Kreileman, Global change
scenarios of the 21st century. Results from the IMAGE 2.1 model. Eleseviers Science,
Asadoorian, M.O. (2005), Simulating the Spatial Distribution of Population and
Emissions to 2100, Report 123, MIT Joint Program on the Science and Policy of Global
Berndes, G., M. Hoogwijk, and R. van den Broek, 2003: The contribution of biomass in
the future global energy system : a review of 17 studies. Biomass and Bioenergy, 25(1),
Berndt, E.R. and L.R. Christensen (1974), ‘Testing for the Existence of a Consistent
Aggregate Index of Labor Inputs’, American Economic Review, 64 (3), 391-404.
Berrittella, M., A.Y. Hoekstra, K. Rehdanz, R. Roson and R.S.J. Tol (2007), ‘The
Economic Impact of Restricted Water Supply: A Computable General Equilibrium
Analysis’, Water Research, 42, 1799-1813.
Brakman, S., H.Garretsen, and C.van Marrewijk (2001), An Introduction to Geographical
Economics - Trade, Location and Growth Cambridge University Press, Cambridge, UK.
Brakman, S., H.Garretsen, and M.Schramm (2006), 'Putting New Economic Geography
to the Test: Free-ness of Trade and Agglomeration in the EU Regions', Regional Science
and Urban Economics, 36, 613-635.
Brovkin, V., M. Claussen, E. Driesschaert, T. Fichefet, D. Kicklighter, M.F. Loutre, H.
D. Matthews, N. Ramankutty, M. Schaeffer and A. Sokolov, 2006. Biogeophysical
effects of historical land cover changes simulated by six Earth system models of
intermediate complexity. Climate Dynamics, 26:587-600. DOI: 10.1007/S00382-005-
Brovkin, V., A. Ganopolski, M. Claussen, C. Kubatzki, and V. Petoukhov, 1999:
Modelling climate response to historical land cover change. Global Ecology and
Biogeography, 8(6), pp. 509-517.
Bruinsma, J. (ed.) (2003), World agriculture: towards 2015/2030 - An FAO Perspective
Earthscan Publications Ltd, London.
Burniaux, J.-M. (2002), Incorporating Carbon Sequestration into CGE Models: A
Prototype GTAP Model with Land Use, Center for Global Trade Analysis Project, West
Cassman, K.G., A. Dobermann, D.T. Walters, and H. Yang, 2003: Meeting cereal
demand while protecting natural resources and improving environmental quality. Annual
Review of Environmental Resources, 28, pp. 10.1-10.44.
Clarke, L., J. Edmonds, H. Jacoby, H. Pitcher, J. Reilly, R. Richels, 2007. Scenarios of
Greenhouse Gas Emissions and Atmospheric Concentrations. Sub-report 2.1A of
Synthesis and Assessment Product 2.1 by the U.S. Climate Change Science Program and
the Subcommittee on Global Change Research. Department of Energy, Office of
Biological & Environmental Research, Washington, DC., USA, 154 pp.
Clarke, J.F. and J.A. Edmonds (1993), Moldelling energy technologies in a competitive
market, Energy Economics 15(2), 123-129.
Cramer, W., A. Bondeau, S. Schaphoff, W. Lucht, B.E. Smith, and S. Sitch, 2004:
Tropical forests and the global carbon cycle : impacts of atmospheric carbon dioxide
climate change and rate of deforestations. Philosophical Transactions of the Royal
Society of London Series B, 359(1443), pp. 331-343.
Christaller, W. (1933), Central Places in Southern Germany, Fischer, Jena.
Darwin, R.F. (1999), 'A FARMer's View of the Ricardian Approach to Measuring
Agricultural Effects of Climatic Change', Climatic Change, 41, (3-4), 371-411.
Darwin, R., Tsigas, M., Lewandrowski, J., Raneses, A. (1996), ‘Land use and cover in
ecological economics’, Ecological Economics, 17, 157–181.
Darwin, R., M. Tsigas, J. Lewandrowski and A. Raneses (1995). World Agriculture and
Climate Change: Economic Adaptations, Agricultural Economic Report #703,
Washington, D.C.: USDA.
Darwin, R.F. and R.S.J.Tol (2001), 'Estimates of the Economic Effects of Sea Level
Rise', Environmental and Resource Economics, 19, 113-129.
Dixit, A.K. and J.E.Stiglitz (1977), 'Monopolistic Competition and Optimum Product
Diversity', American Economic Review, 67, (3), 297-308.
Duranton, G. (2006), 'Some Foundations for Zipf's Law: Product Proliferation and Local
Spillovers', Regional Science and Urban Economics, 36, 542-563.
Eichner, T. and R. Pethig (2006), 'Economic Land Use, Ecosystem Services, and
Microfounded Species Dynamics', Journal of Environmental Economics and
Management, 52, 707-720.
Finnoff, D. and J.Tschirhart (2003), 'Protecting an Endangered Species while Harvesting
its Prey in a General Equilibrium Ecosystem Model', Land Economics.
Fujita, M., P.R.Krugman, and A.J.Venables (1999), The Spatial Economy - Cities,
Regions, and International Trade, First paperback edition, 2001 edn, MIT Press,
Cambridge, USA/London, UK.
Gitz, V., and P. Ciais, 2004: Future expansion of agriculture and pasture acts to amplify
atmospheric CO2 in response to fossil fuel and land-use change emissions. Climate
Change, 67(2-3), pp. 161-184.
Grübler, A. (1994), ‘Technology’, in: W. Meyer and B. Turner II (eds.), Changes in Land
Use and Land Cover: A Global Perspective, Cambridge University Press, Cambridge, pp.
Grübler, A., B. O’Neill, K. Riahi, V. Chirkov, A. Goujon, P. Kolp, I. Prommer, S.
Scherbov and E. Slentoe (2007), 'Regional, National and Spatially Explicit Scenarios of
Demographic and Economic Change Based on SRES’, Technological Forecasting and
Social Change, 74, 980-1029.
Gouel, C. and T.W. Hertel (2006), Introducing Forest Access Cost Functions into a
General Equilibrium Model, GTAP Research Memorandum 8, Global Trade Analysis
Project, West Lafayette,
Henderson, J.V. (1974), ‘The Sizes and Types of Cities’, American Economic Review, 64
(4), 640-656.Hertel, T.W., 1997, Global Trade Analysis: Modeling and Applications
(New York: Cambridge University Press).
Hertel, T.W., ed. 1997. Global Trade Analysis Modeling and Applications. Cambridge
Hertel, T.W. and M.E. Tsigas, 1988, Tax Policy and U.S. Agriculture: A General
Equilibrium Approach, American Journal of Agricultural Economics 70(2):289-302.
Heistermann, M., C.Mueller, and K.Ronneberger (2006), 'Land in Sight? Achievements,
Deficits and Potentials of Continental to Global Scale Land-Use Modeling', Agriculture,
Ecosystems and Environment, 114, 141-158.
Houghton, R.A. 2003. Revised estimates of the annual net flux of carbon to the
atmosphere from changes in land use and land management 1850–2000. Tellus. 55b: 378-
House, J.I., I.C. Prentice, and C. Le Quere, 2002: Maximum impacts of future
reforestation or deforestation on atmospheric CO2. Global Change Biology, 8(11), pp.
Hubacek, K. and J.C.M.C.van der Bergh (2006), 'Changing Concepts of 'Land' in
Economic Theory: From Single to Multi-Disciplinary Approaches', Ecological
Economics, 56, 5-27.
IPCC (2007), Climate Change 2007 - Impacts, Adaptation and Vulnerability,
Contribution of Working Group II to the Fourth Assessment Report of the United Nations
Intergovernmental Panel on Climate Change.
Isard, W. (1954), Location and Space-Economy, MIT Press, Cambridge.
Jakeman, G. and B.S.Fisher (2006), 'Benefits of Multi-Gas Mitigation: An Application of
the Global Trade and Environment Model (GTEM)', Energy Journal (Multi-Greenhouse
Gas Mitigation and Climate Policy Special Issue), 323-342.
Krugman, P.R. (1991), 'Increasing Returns and Economic Geography', Journal of
Political Economy, 99, (3), 483-499.
Krugman, P.R. (1998a), Development, Geography, and Economic Theory MIT Press,
Krugman, P.R. (1998b), 'What's New about the New Economic Geography?', Oxford
Review of Economic Policy, 14, (2), 7-17.
Kuhn, A. (2003), From World Market to Trade Flow Modelling—The Re-Designed
WATSIM Model, Final report, Institute of Agricultural Policy, Market Research and
Kurosawa, A. (2006), 'Multigas Mitigation: An Economic Analysis Using GRAPE
Model', Energy Journal (Multi-Greenhouse Gas Mitigation and Climate Policy Special
Leemans, R., B. Eickhout, B. Strengers, L. Bouwman and M. Schaeffer, 2002. The
consequences of uncertainties in landuse, climate and vegetation responses on the
terrestrial carbon. Science in China, 45: 126-141
Levy, P.E., M.G.R.Cannell, and A.D.Friend (2004), 'Modelling the impact of future
changes in climate, CO2 concentration and land use on natural ecosystems and the
terrestrial carbon sink', Global Environmental Change, 14, 21-30.
Loesch, A. (1940), The Economics of Location, Fischer, Jena.
López, R.A., A.O. Adelajo, and M.S. Andrews, 1988, The Effects of Suburbanization on
Agriculture, American Journal of Agricultural Economics 70(2):346-358.
Ludena, C.E., T.W. Hertel, P.V. Preckel, K. Foster and A. Nin, 2007. “Productivity
Growth and Convergence in Crop, Ruminant and Non-Ruminant Production:
Measurement and Forecasts”, Agricultural Economics 37, (1), 1-17.
Martin, R. (1999), 'The New 'Geographical Turn' in Economics: Some Critical
Reflections', Cambridge Journal of Economics, 1999, (23), 65-91.
Matthews, H.D., A.J. Weaver, M. Eby and K.J. Meissner, 2003. Radiative forcing of
climate by historical land cover change. Geophysical Research Letters, 30: 1055. DOI:
Mendelsohn, R.O., W.D.Nordhaus, and D.Shaw (1994), 'The Impact of Climate on
Agriculture: A Ricardian Analysis', American Economic Review, 84, (4), 753-771.
Millenium Ecosystem Assessment, 2005: Ecosystems and human well-being: Scenarios.
Findings of the Scenarios Working Group. Island Press, Washington DC.
Neary, J.P. (2001), 'Of Hype and Hyperbolas: Introducing the New Economic
Geography', Journal of Economic Literature, 39, (2), 536-561.
Nordhaus, W.D. (2006), 'Geography and Macroeconomics: New Data and New Findings',
Proceedings of the National Academy of Science
Rao, S. and K.Riahi (2006), 'The Role of Non-CO2 Greenhouse Gases in Climate Change
Mitigation: Long-Term Scenarios for the 21st Century', Energy Journal (Multi-
Greenhouse Gas Mitigation and Climate Policy Special Issue), 177-200.
Riahi, K., A.Gruebler, and N.Nakicenovic (2007), 'Scenarios of Long-Term Socio-
Economic and Environmental Development under Climate Stabilization', Technological
Forecasting and Social Change, 74, 887-935.
Ricardo, D. (1817), Principles of Political Economy and Taxation, John Murray, London.
Rose, S., H. Ahammad, B. Eickhout, B. Fisher, A. Kurosawa, S. Rao, K. Riahi, D.P. van
Vuuren (2007) Land in Climate Stabilization Modeling: Initial Observations, Energy
Modeling Forum, Stanford University, www.stanford.edu/group/EMF/
Rosegrant, M.W., C. Ringler, S. Msangi, S.A. Cline and T.B. Sulser (2005), International
Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT-WATER):
Model Description, International Food Policy Research Institute, Washington, D.C.,
Rossi-Hansberg, E. (2005), 'A Spatial Theory of Trade', American Economic Review, 95,
Rounsevell, M.D.A., J.E. Annetts, E. Audsley, T. Mayr and I. Reginster (2003),
‘Modelling the spatial distribution of agricultural land use at the regional scale’,
Agricultural, Ecosystems, Environment, 95, 465–479
Samuelson, P.A. (1954), ‘The transfer problem and transport costs’, Economic Journal
Sanchirico, J.N. and J.E. Wilen (1999), 'Bioeconomics of Spatial Exploitation in a Patchy
Environment', Journal of Environmental Economics and Management, 37, 129-150.
Sands, R.D. and J.A.Edmonds (2005), 'Climate Change Impacts for the Conterminous
USA: An Integrated Assessment', Climatic Change, 69, 127-150.
Sands, R.D. and M.Leimbach (2003), 'Modeling agriculture and land use in an integrated
assessment framework', Climatic Change, 56, 185-210.
Sathaye, J., W.Makundi, L.Dale, P.Chan, and K.Andrasko (2006), 'GHG Mitigation
Potential, Costs and Benefits in Global Forests: A Dynamic Partial Equilibrium
Approach', Energy Journal (Multi-Greenhouse Gas Mitigation and Climate Policy
Special Issue), 127-162.
Schlenker, W., W.M. Hanemann, and A.C. Fisher (2005), 'Will US Agriculture Really
Benefit from Global Warming? Accounting for Irrigation in the Hedonic Approach',
American Economic Review, 95, (1), 395-406.
Smith, M.D. and J.E. Wilen (2003), 'Economic impacts of marine reserves: the
importance of spatial behavior', Journal of Environmental Economics and Management,
Sohngen, B. and R. Mendelsohn (2007), “A Sensitivity Analysis of Carbon
Sequestration." In Human-Induced Climate Change: An Interdisciplinary Assessment.
Edited by M. Schlesinger. Cambridge University Press.
Sohngen, B., and R. Mendelsohn (2003), An optimal control model of forest carbon
sequestration. American Journal of Agricultural Economics, 85(2), pp. 448-457.
Sohngen, B., R. Mendelsohn, and R. Sedjo (1999), “Forest Management, Conservation,
and Global Timber Markets.” American Journal of Agricultural Economics. 81: 1-13.
Sohngen, B. and R.Sedjo (2006), 'Carbon Sequestration in Global Forests under Different
Carbon Price Regimes', Energy Journal (Multi-Greenhouse Gas and Climate Policy
Special Issue), 109-126.
Stephenne, N. and E.F. Lambin (2001), ‘A dynamic simulation model of land-use
changes in Sudano-sahelian countries of Africa (SALU)’, Agriculture, Ecosystems and
Environment, 85, 145–161.
Stephenne, N. and E.F. Lambin (2004), ‘Scenarios of land-use change in Sudano-sahelian
countries of Africa to better understand driving forces’, GeoJournal, 61, 365–379.
Steuart, J. (1767), Inquiry into the Principles of Political Economy, Printed for A. Millar
and T. Cadell in the Strand, London.
Stelder, D. (2005), 'Where Do Cities Form? A Geographical Agglomeration Model for
Europe', Journal of Regional Science, 45, (4), 657-679.
Strengers, B., R. Leemans, B. Eickhout, B. de Vries, and L. Bouwman, 2004: The land-
use projections and resulting emissions in the IPCC SRES scenarios as simulated by the
IMAGE 2.2 model. GeoJournal, 61(4), pp. 381-393.
Timmins, C. (2006), 'Endogenous Land Use and the Ricardian Valuation of Land Use',
Environmental and Resource Economics, 33, 119-142.
Tschirhart, J. (2000), 'General Equilibrium of an Ecosystem', Journal of Theoretical
Biology, 203, 13-32.
Tubiello, F.N. and G. Fischer (2007), 'Reducing Climate Change Impacts on Agriculture:
Global and Regional Effects of Mitigation: 2000-2008’, Technological Forecasting and
Social Change, 74, 1030-1056.
Turgot, A.-R.-J. (1793), Reflections on the Formation and Distribution of Wealth, Printed
by E. Spragg for J. Good, Bookseller no 159, New Bond Street, John Anderson, no 62,
Holborn Hill, and W. Richardson, Royal Exchange, London.
USEPA, 2006a. Global Emissions of Non-CO2 Greenhouse Gases: 1990-2020. Office of
Air and Radiation, US Environmental Protection Agency (US-EPA), Washington, D.C.
USEPA, 2006b. Global Mitigation of Non-CO2 Greenhouse Gases. United States
Environmental Protection Agency, Washington, DC, 430-R-06-005,
USEPA (2005), Greenhouse Gas Mitigation Potential in U.S. Forestry and Agriculture,
EPA-R-05-006. Washington, D.C: U.S. Environmental Protection Agency, Office of
Van Vuuren, D.P., M.G.J. den Elzen, P.L. Lucas, B. Eickhout, B.J. Strengers, B. van
Ruijven, S. Wonink and R. van Houdt (2007), ‘Stabilizing Greenhouse Gas
Concentrations at Low Levels: An Assessment of Reduction Strategies and Costs’,
Climatic Change, 81 (2), 119-159.
van Vuuren, D.P., B.Eickhout, P.L.Lucas, and M.J.G.den Elzen (2006), 'Long-Term
Multi-Gas Scenarios to Stabilise Radiative Forcing -- Exploring Costs and Benefits
within an Integrated Assessment Framework', Energy Journal (Multi-Greenhouse Gas
Mitigation and Climate Policy Special Issue), 201-233.
Veldkamp, A. and L.O. Fresco (1996), ‘CLUE-CR: an integrated multi-scale model to
simulate land use change scenarios in Costa Rica’, Ecological Modelling, 91, 231–248.
Verburg, P.H., A. Veldkamp and L.O. Fresco (1999), ‘Simulation of changes in the
spatial pattern of land use in China’, Applied Geography, 19, 211–233.
Vitousek, P.M., H.A. Mooney, J. Lubchenco, and J.M. Melillo, 1997: Human domination
of Earth's ecosystems. Science, 277(5325), pp. 494-499.
von Thuenen, J. (1826), The Isolated State, Pergamon Press, London.
Wassenaar, T., P. Gerber, P.H. Verburg, M. Rosales, M. Ibrahim and H. Steinfeld (2007),
‘Projecting Land Use Changes in the Neotropics: The Geography of Pasture Expansion
into Forest’, Global Environmental Change, 17, 86-104.
Weber, A. (1909), Theory of the Location of Industries, University of Chicago Press,
Zipf, G.K. (1935), The Psychobiology of Language, Houghton-Mifflin, New York.
1990 20002010 2020 2030204020502060 2070 20802090 2100
GRAPE-EMF21 IMAGE-EMF21IMAGE-MA-AMIMAGE-MA-GO IMAGE-MA-OS
1990 20002010 20202030 2040 2050 2060 20702080
19902000 20102020 203020402050 2060 2070208020902100
GRAPE-EMF21IMAGE-EMF21IMAGE-MA-AM IMAGE-MA-GO IMAGE-MA-OS
IMAGE-MA-TG AgLU-0.0% AgLU-0.5%AgLU-1.0%
Notes: IMAGE-EMF21 = van Vuuren et al. (2006a) scenario from EMF-21 Study; IMAGE-MA-xx = Millennium Ecosystem
Assessment (2005) scenarios from the IMAGE model for four storylines (GO = Global Orchestration, OS = Order from
Strength, AM = Adaptive Mosaic, TG = TechnoGarden); AgLU-x.x% = Sands and Leimbach (2003) scenarios with x.x%
annual growth in crop yield; GTM-2003 = Sohngen and Mendelsohn (2003) global forest scenario; GTM-EMF21 = Sohngen
and Sedjo (2006) global forest scenario from EMF-21 Study; GCOMAP-EMF21 = Sathaye et al. (2006) global forest scenario
from EMF-21 Study; GRAPE-EMF21 = Kurosawa (2006) scenario from EMF-21 Study
Figure 1. Global cropland (a), forest land (b) and grassland (c) projections (2010 = 1;
shaded areas indicate SRES scenario ranges, post-SRES scenarios denoted with solid
2000 20102020 20302040 20502060 2070 20802090 2100
Notes: MESSAGE-EMF21 = Rao and Riahi (2006) scenario from EMF-21 Study; GTEM-EMF21 = Jakeman and Fisher (2006)
scenario from EMF-21 Study; MESSAGE-A2r = Riahi et al. (2007) scenario with revised SRES-A2 baseline; IMAGE 2.3 = van
Vuuren et al. (2007) scenario. The IMAGE 2.3 LUCF baseline scenario also emits non-CO2 emissions (CH4 and N2O) of 0.26,
0.30, 0.16 GtCO2eq in 2030, 2050, and 2100 respectively.
Figure 2. Baseline land-use change and forestry carbon net emissions.
32 Download full-text
IMAGE 2.3 MESSAGEIMAGE 2.3 IMAGE2.3IMAGE-
2.9 W/m2 ~3.0 W/m2 3.7 W/m2 4.5 W/m2
Source: Rose et al. (2007)
Figure 3. Cummulative cost-effective agricultural, forestry, and biomass abatement 2000-
2100 from various 2100 stabilisation scenarios.