Download Rising to the Challenge: Changing Course to Feed

RISING TO THE CHALLENGE:
Changing Course to Feed the World in 2050
ActionAid USA
Report
October 2013
COVER PHOTO: Leaders of the Tuzamure Agaseke cooperative in Karongi, Rwanda
perform a traditional dance. This cooperative has received support and training from
ActionAid in organic farming, human rights, co-op management, and community
participation in decision making. Co-op members, most of whom are genocide
widows, grow corn and supply maize flour to local markets.
CONTENTS
Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
FAO Modeling: Strengths, Weaknesses, and Misinterpretations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
The Challenge and Utility of Long-Term Economic Modeling — Box 1 . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Common Themes, Questionable Assumptions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Foresight Modeling: Types, Uses, and Limitations — Box 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Scenario Modeling for Policy Formulation: Gaps and Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Biofuels Expansion.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Climate Change. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
The Agriculture Modeling Improvement Project (AgMIP) — Box 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Modeling Broader Questions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Biophysical Modeling: Land Use — Box 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Biophysical Modeling: Water Management — Box 5.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Future Research Needs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Endnotes.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3
RISING TO THE CHALLENGE:
Changing Course to Feed the World in 2050
ActionAid USA
Report
October 2013
EXECUTIVE SUMMARY
by 2050 are based on outdated or flawed economic
forecasting and misleading characterizations of this
research. More reliable estimates of current supply,
productivity, and demand trends — assuming businessas-usual policies — suggest both the need and the
capacity to increase agricultural production by 60% over
2005-7 levels by 2050. This is a far cry from doubling
food production. In fact, the failure to distinguish food
production from agricultural production obscures the
largest single contributor to recent food price spikes:
the massive expansion of agricultural biofuel production.
Since the 2007-8 food price crisis, alarms have sounded
regarding our ability to feed a growing population in
2050. Some warn that we need to double food production;
others estimate that food production must increase by
60-70%. All feed the alarmist notion that global hunger
is the result of flagging food production amid looming
resource constraints. The misguided policy prescriptions
that follow typically call for the expansion of industrialscale agricultural development, ignoring the true threats
to our global food supply: biofuels expansion, inadequate
investment in climate-resilient agriculture, lagging support
for small-scale and women food producers, and the
massive loss of food to spoilage and waste.
Rather than fueling alarmist agricultural productivism,
the utility of food security forecasts should be to help
decision-makers identify policies that are contributing
to high and volatile prices, food insecurity, and looming
resource constraints on agricultural production as well
as changes that could alleviate these impacts. Most
economic forecasting fails to adequately incorporate
several key variables:2050
Most of the recent warnings derive from a group of
economic modeling studies that were recently reviewed
by researchers at Tufts University’s Global Development
and Environment Institute. In their assessment, the Tufts
researchers found that many of the public pronouncements calling for a doubling of global food production
EXECUTIVE SUMMARY
4
n Biofuels – Biofuels expansion is a relatively recent
phenomenon that has been poorly captured by most economic modeling to date. Few models
adequately account for current trends, with some
underestimating business-as-usual expansion by
100%. With national mandates and targets significantly driving biofuels expansion, updated forecasts
are urgently needed to help policymakers assess
the food security implications of current policies.
Those policies are incontrovertibly resulting in rising
and more volatile food prices, with up to 40% of
recent price increases in agricultural commodities
attributable to biofuels expansion. Those policies
are projected to divert as much 13% of cereal
production from needed food production by 2030.
n Climate change – We are only just beginning to
understand the implications of climate change
for agriculture and food security. These impacts,
plagued by multiple layers of uncertainty, are poorly
incorporated into most economic forecasts. With
the outcome of international climate negotiations
uncertain, urgent attention is needed to mitigate
industrial agriculture’s tremendous contribution to
global warming and help developing country food
producers to adapt to a changing climate.
In all of these areas, policymakers need forecasts to
help them interrogate established policies and practices
that need to change, such as consumption patterns,
energy policies, unfair distribution and access, land use,
and investment priorities.
n Inadequate and poorly targeted agricultural
investment – Agricultural investment is key to
increasing food production. Whereas many
projections stress the importance of agricultural
productivity growth, few models assess different
priorities for agricultural research and investment.
A growing consensus supports increased investment in climate-resilient food production, focusing
on small-scale producers in food-insecure parts of
the world. Yet most research, private and public,
focuses on large-scale, input-intensive agricultural
development. So too does most investment, driven
by private sector-led projects, such as the “New
Alliance for Food Security and Nutrition” initiated
by the G8.
Meanwhile, a growing body of experience at the local
and regional levels, demonstrates the lasting value of
investments in smallholder farming and sustainable
agricultural methods. Strategic policy changes and
investments in this area can scale-up successful
approaches and expand them to regions where they
are most appropriate and most needed, especially in
regions where food security is tenuous despite high
agricultural potential.
This report reviews the economic forecasting on which
most of the alarmist 2050 pronouncements are based,
presents alternative modeling that can add useful
insights, and identifies areas in which further research
can guide policymakers to change failing business-asusual policies. This much is clear: hunger, now and in
the future, is less a matter of inadequate production
than inequitable access to food and food-producing
resources; and a singular focus on increasing production
is misguided as we simultaneously waste one-third
of the food that is produced and pursue a course to
devote another 13% of cereals to feeding our cars
instead of our people.
n Food waste and spoilage – One-third of global
food production fails to nourish anyone. In industrialized countries, wasteful consumption patterns result
in tremendous losses, while in developing countries
poor infrastructure means high rates of spoilage
before food makes it to market. Most current
forecasts ignore the possibility that measures
could be taken to address this problem, assuming
continued waste of food at current rates. This
assumption alone puts alarmist calls for increased
food production into question.
5
INTRODUCTION
communities, increase production by 60 per cent by
2050, drastically reduce huge losses and waste of food
and manage our natural resources sustainably, so that it
flourishes for future generations.”6
“Model outputs should not be misinterpreted as
forecasts with well-defined confidence intervals.
Rather they are meant to provide quantified insights about the complex interactions in a highly
interdependent system and the potential general
size order of effects, which cannot be obtained
by qualitative and theoretical reasoning alone.”1
This report assesses the evidence for these and other
claims about feeding the world in 2050. A cottage
industry has developed around this question, with
studies diverging widely in their assessments and
prescriptions. These estimates matter precisely because
they drive both public discourse and public policy. Are
we facing a Malthusian future of food scarcity, or a
“limits to growth” scenario in which the carrying capacity
of the planet reaches exhaustion? More importantly,
how well do various estimates of global demand and
supply incorporate the uncertainties occasioned by
changing economic and environmental trends, from
climate change to biofuels expansion, from slowing
growth in agricultural productivity to rising meat demand
from a growing global middle class?
Alarm bells sounded in 2008 as global agricultural
commodity prices doubled and pundits forecasted a
world unable to feed itself. With the global population
expected to surpass nine billion in 2050, the heads of
the UN’s Food and Agriculture Organization (FAO) and
World Food Program (WFP) called for a doubling of
global food production by 2050 to meet rising demand
from a growing population expected to consume more
meat as well as from the rapidly growing demand for
bioenergy crops.2
Although future-oriented “foresight” modeling often
relies on “business-as-usual” scenarios that treat current
practices as inevitable, the best use of such modeling
is to guide policymakers in changing precisely these
practices. How well does current 2050 modeling help
policymakers assess and address the drivers of unsustainably high agricultural prices through public policies,
including the sustainable use of resources? We hope
this report will serve as a “user’s guide” to much of the
existing modeling. (For a more detailed assessment, see
the recent Tufts University paper on the subject.i) We
begin by tracing the origins of the FAO’s recent assertions, explaining the reliability and limitations of their
modeling efforts. Then, we take a step back to examine
the difficulties inherent in such long-range modeling,
identify some useful modeling typologies, and present
some of the more interesting alternative approaches. In
the following section, we look at scenario modeling that
has been done on key drivers of 2050 forecasts, namely
biofuels expansion and climate change. We conclude
with some observations on the strengths and limitations
of 2050 modeling to date and suggest areas in which
this modeling can make the most useful contributions
to decision-making.
“With almost 80 million more people to feed each year,
agriculture can’t keep up with the escalating food
demand,” warned Frank Rijsberman, head of the
Consultative Group on International Agricultural
Research (CGIAR). “FAO estimates that we have to
double food production by 2050 to feed the expected
9 billion people, knowing that one billion people are
already going to bed hungry every day.”3
Will world food needs finally outstrip productive capacity
as Thomas Malthus warned in his famous 1798 treatise,
An essay on the principle of population?4 His predictions,
from what amounts to one of the first global modeling
studies on the world’s ability to feed its growing population, have been widely discredited. But current resource
constraints, exacerbated by uncertainties over climate
change, have revived questions about the ability of
society and the planet to feed our growing population.
In actuality, the FAO’s expert team of agricultural modelers
hadn’t really called for a doubling of food production by
2050. Rather, the agency’s models had indicated the
need to increase overall agricultural production —
including food production — by 70% from 2005/07
to 2050.5 A 2012 update of these estimates brought
the figure down to 60% for the same period. As the
modeling team noted before the 2012 Rio +20 summit,
“We need to improve people’s access to food in their
i
6
“Can We Feed the World in 2050? A Scoping Paper to Assess the
Evidence,” by Timothy A. Wise, GDAE Working Paper No. 13-04,
September 2013. http://www.ase.tufts.edu/gdae/policy_research/
FeedWorld2050.html
We proceed on a cautionary note. Estimates of our
ability to “feed the world” rely primarily on global
estimates of supply and demand, yet ecosystems,
agricultural production, and hunger occur at local and
regional levels. As a result, global estimates of “our”
ability to feed “the world” quickly break downii, begging
the more important questions of how food systems
develop across widely differing landscapes, societies,
and levels of economic development, and how equitably
the food is distributed. In the end, “the world” is not fed
in aggregate, and there is no collective “we” doing the
feeding.
1. Much of the data has been updated to more recent
base years (2005-7) and the modelers have
incorporated recent and improved estimates of
food demand, as well as land and water resources.
The shift from 70% to 60% reflects less a change in
estimated demand than it does an updated figure, reflecting actual production in the 2005-7 base year
period.
2. They are based on widely accepted (though still
uncertain) population projections (9.15 billion
people by 2050iii), well-grounded estimates of
economic growth (average global GDP growth of
1.36%/year), and an expected growth in demand
which incorporates the expected shift in developing
countries to more meat-based diets (all agricultural
commodities, all uses, 1.1% annual growth).
FAO MODELING: STRENGTHS,
WEAKNESSES, AND MISINTERPRETATIONS
3. Estimates of agricultural yield growth are moderate
but consistent with historical trends (1.1% per year),
and not based on unrealistic assumptions regarding
productivity improvements.
The most widely quoted figures on agricultural supply
and demand in 2050 come from the FAO’s efforts to
gauge future food demand. FAO estimates following the
2007-8 food price spikes, which suggested the need
to double food production by 2050, were the basis for
international alarms about our ability to feed the world.
While the 100% figure is still cited by some policymakers,
the FAO’s later estimate that a 70% increase in
agricultural production is needed seems to have taken
hold as the most commonly cited figure. The FAO’s
most recent update in June 2012, which lowered this
figure to 60%, is generally recognized as the best official
estimate, though the 70% figure remains widely cited in
government circles and in the media.
4. The FAO does not assume an implausible conversion
of land to agricultural uses, a problem in some
modeling studies. Instead they assume that 70 million
hectares are converted by 2050, a 9% increase.
5. The FAO validates its projections against data for
more recent years and against the FAO-OECD
ten-year projections to 2020.7
Thus, the latest FAO estimates provide substantial
reassurance that, with the right policies, global
agriculture is capable of meeting projected demand
for both food and non-food uses in 2050.
How reliable are these estimates? A number of misconceptions about the nature of such modeling have implications for any effort to assess likely outcomes under
different environmental or policy scenarios. (See Box 1)
First, the FAO is very clear that it is not answering the
question, “How can we feed the world in 2050?” Rather,
it is answering a more straightforward question: “Will
world production increase enough to meet projected
demand?”
Agricultural economist Thomas Hertel generally
agrees with this overall assessment.8 He notes that
income growth and changing demand are relatively
The FAO study assigns particular importance to this point: “
... examining the issue of food insecurity by means of global
variables (e.g. can the world produce all the food needed for
everyone to be well-fed?) is largely devoid of meaning” and
“In conclusion, the issue whether food insecurity will be eliminated
by the end of the century is clouded in uncertainty, no matter
that from the standpoint of global production potential there
should be no insurmountable constraints” (Alexandratos, Nikos
and Jelle Bruinsma (2012). World agriculture towards 2030/2050:
the 2012 revision. ESA Working paper. Rome, Food and
Agriculture Organization: 20-21)
ii
The FAO’s answer to this question is yes. And their
estimates should offer some reassurance — with very
important caveats — to those who would sound alarms
over our ability to produce enough food to feed the
global population in 2050. Why are these findings
reassuring?
iii
7
Population growth rates are anticipated to vary widely depending
on the country. Alexandratos et al (ibid) notes that the majority
of countries whose population growth is expected to be fast in the
future are those showing inadequate food consumption and high
levels of undernourishment, mostly in sub-Saharan Africa.
BOX 1 — The challenge and utility of long-term economic modeling
The studies and analysis in the FAO’s collection22 are a good representation of the economic modeling that has generated some
of the most widely cited estimates of food needs in 2050. Alexandratos’s observations highlight the challenges inherent in such
long-term modeling and the sensitivity of the results to assumptions about variables of great uncertainty, be they economic,
environmental, or policy-related. Despite their significance, such assumptions are often made explicit only in technical annexes,
and sometimes not at all. The results, on the other hand, are generally presented with a high degree of certainty, after which
they are repeated as definitive by policymakers and the media. Such has been the case with the studies estimating agricultural
production and demand to 2050.
Indeed, Michael Reilly of the UK’s Government Office for Science and Dirk Willenbockel of the Institute of Development Studies
at the University of Sussex, in their excellent overview of food system modeling, warn of this precise problem. “Model outputs
should not be misinterpreted as forecasts with well-defined confidence intervals. Rather they are meant to provide quantified
insights about the complex interactions in a highly interdependent system and the potential general size order of effects, which
cannot be obtained by qualitative and theoretical reasoning alone.”23
Reilly and Willenbockel point out that food system modeling requires the analyst to coordinate the mapping of one uncertain
system — agricultural production and consumption — with another — ecosystems. Both systems are characterized by gaps
in data and knowledge, limited confidence in predicting the future from past trends, and likely instability of future systems
behavior. One set of uncertainties compounds the other, leaving, logically, a virtually unlimited range of possible outcomes. This
is true of the “known unknowns,” such as the extent to which rising CO2 levels will have some positive effects on agricultural
production (CO2 fertilization) or the rate of agricultural productivity growth. Add in “unknown unknowns,” such as extreme but
low-probability climate events, and it becomes clear that the forecasting potential for long-range modeling is extremely limited.
Still, even though long-range modeling shouldn’t be used as a “crystal ball,” such efforts have a great deal to offer those
struggling to identify the best way forward. At best, they can challenge the “mental maps” of policymakers by drawing out the
plausible real-world implications of business-as-usual policies and alternative approaches. They can also assess the relative
importance of various drivers of change.
well-understood and can be reliably modeled. He
concurs that yield growth is likely to keep up with
demand growth, which will slow. He points out that
there is greater potential for productivity gains on rainfed land, where yields are often below 50% of their potential, than on irrigated land.9 He cites studies showing
that bringing such currently cultivated lands up to their
potential production, using existing technology, would
generate increases in 2050 of 60% for wheat, 50% for
maize, 40% for rice, and 20% for soybeans.10 Most such
lands are in less developed regions of developing countries, presenting both an obstacle (resources) and an
opportunity (reducing hunger and poverty).
n Biofuels expansion – To arrive at its 60% estimate,
the FAO assumes sufficient biofuel expansion to
meet existing mandates through 2020, then no
further expansion beyond that. This is both
unrealistic and, from the perspective of policymakers,
unhelpful. Current estimates project first generation
biofuel demand in 2030 at double the FAO’s
assumed levels.11
n Climate change – The FAO modelers openly
acknowledge their inability to incorporate the
impacts of climate change on agricultural production.
Even with perfect mitigation today we would see
measurable climate change by 2050. In light of
this fact, FAO projections are clearly in need of
significant adjustment. As the authors themselves
acknowledge, “In principle, a scenario that assumes
no climate change has no place in the array of
scenarios to be examined.”12
Hertel and the FAO also agree on two very important
wild cards, neither of which is adequately reflected
in most modeling to date: bioenergy production and
climate change.
8
Obviously, these are not trivial shortcomings. Both
suggest that FAO projections err on the side of
overestimating food availability in 2050, as both
trends imply negative impacts on global supply.
extremely small upward adjustments to the assumed
productivity growth rate.
Growth assumptions vary widely among the different
models and projections. Some researchers estimate that
we will need an annual average yield growth of 1.25%
to meet global food needs in 2050.14 The latest FAO
projections, on the other hand, estimate that the world
average cereals yield would need to grow at just 0.7%
per year to meet 2050 demand.15
Common Themes,
Questionable Assumptions
The impact of biofuels and climate change were
among the issues addressed at a 2009 expert meeting
convened by the FAO to assess the implications of the
food price crisis for the world’s ability to meet future
food needs. The resulting papers were later published as
a book, Looking Ahead in World Food and
Agriculture: Perspectives to 2050.iv
The World Bank’s model, by contrast, is exceedingly
optimistic, projecting 2.1% annual growth in agricultural
productivity as a result of technological innovation. This
is well beyond historical trends, and nearly twice the
productivity growth rate assumed by the FAO (and most
others). Such optimistic assumptions can mask a range
of dangers. Most significantly, they assume that resource
constraints such as climate change, land use, and water
availability, will be addressed through innovation. This
may turn out to be true in some cases, but incorporating
such implicit assumptions into an agricultural model can
skew the results dramatically. Not surprisingly, World
Bank projections suggest that we will face few
constraints on feeding the world in 2050.
Interestingly, a common finding indicated that future food
prices will not be as high as recent experience suggests.
Price is the key indicator in most economic models, as
it gauges the balance between supply and demand —
higher prices suggest food shortages while lower prices
suggest agricultural surpluses. The models discussed at
the 2009 FAO meeting generally projected price levels
lower than the “post-surge” prices of recent years, and in
line with “pre-surge” (2003-5) price levels through 2030,
then rising by 2050 to about 30% above pre-surge prices (but still well below current post-surge price levels).
These findings are in stark contrast to prevailing characterizations of food prices as permanently
high and rising.13
Modeling results are also very sensitive to economic
growth assumptions, with faster economic growth
increasing demand at a faster rate but also increasing
incomes and reducing poverty. World Bank modeling
tends to be optimistic in this area as well, assuming an
average annual growth rate to 2050 of 1.6% for highincome countries and 5.2% for developing countries.16
FAO researcher Nicos Alexandratos identifies some
of the factors that drive agricultural modeling results,
demonstrating that these models are extremely sensitive
to a few key assumptions. One is population growth
— the greater the projected population, the greater the
projected demand for agricultural products. FAO uses
a 2050 estimate of 9.15 billion people, which is based
on the middle path among three United Nations
population scenarios, from the 2008 revision of UN
population estimates.v
Professor Evan Hillebrand, who served as an economic
analyst and modeler for over 30 years at the CIA, offers
a more sober assessment, contrasting this “market first”
high-growth scenario, with a scenario that assumes
developing countries each grow at the rates of the
The volume includes some valuable updates on the 2009 papers
and added material and analysis, including a comparison of the
main models and an assessment of their differences and relative
strengths and limitations. As such, it represents one of the more
comprehensive efforts to gather and assess the results from a
range of researchers and models. Conforti, Piero (2011).
Looking Ahead in World Food and Agriculture: Perspectives to
2050. Rome, Italy, Food and Agriculture Organization.
iv
Modeling results are also extremely sensitive to
assumptions about agricultural productivity growth.
Over a 40-year time horizon (2010-2050) each 0.1%
change in the assumed growth rate produces a 4%
change in total output in 2050; an additional 1% per
year adds 40%. Within an economic model therefore,
one can dramatically affect the outcome through
Unfortunately, population assumptions are not uniform across
models, or even within the same modeling exercise, making
results very difficult to compare. One IPCC scenario, for example,
assumes a 2050 population of 11.3 billion, dramatically increasing
projected food demands.
v
9
previous 25 years. This “trend growth” scenario for
developing countries estimates that extreme poverty
worldwide in 2050 will be 20% rather than the 2.6%
projected by the “market first” model. Contrasting
these scenarios highlights just how variable regional
performance is likely to be. Under the “trend growth”
scenario, 53% of Sub-Saharan Africa is projected to live
in extreme poverty by 2050 with 78% below the $2.50/
day poverty line (a higher percentage than in 2005),
versus 12% under the more optimistic scenario.17
of the existing problems with the current global food
system”.18 She warns of prior ideological commitments
to a framing of the food security issue that defines food
security as an issue of production rather than access
and utilization.
Tomlinson also notes that the measure of food security
used in many studies is based on per capita food
consumption in calories, derived from estimates of
availability. Such food availability projections allow for
broad trend estimates but they neglect demand-side
issues such as food waste, not to mention unequal
access and distribution.19 This methodology feeds
inaccurate perceptions that economic growth and
increases in agricultural production will by themselves
lead to reductions in food insecurity.20 Analyst Michiel
Van Dijk of the Dutch Agricultural Economics Institute
(LEI) agrees, pointing out that most modeling fails to give
Professor Isobel Tomlinson, of the University of London’s
Department of Development Studies, offers a useful
critique of the “new productivism” fueled by economic
models that foresee looming shortfalls in global production. As Tomlinson points out, “Increasing production on
such a scale was never intended as a normative goal of
policy and, secondly, to do so would exacerbate many
BOX 2 — Foresight Modeling: Types, uses, and limitations
Reilly and Willenbockel provide a useful typology of foresight, or future oriented modeling, distinguishing between projections,
exploratory scenarios, and normative scenarios.
Projections include efforts like the FAO’s and most of the other studies reviewed thus far. They include baseline models, which
attempt to lay out what will happen if we stay on our current path, as well as “what if” scenario modeling in which one or two
factors (such as climate change impacts or biofuels expansion) are altered to assess their significance.
Exploratory scenarios introduce of a more complex and related set of changes that represent various possible pathways for
societies. These pathways include what the authors call “strategic” scenarios, such as the Millennium Ecosystem Assessment,
an in-depth evaluation of the consequences of ecosystem change for human well-being. As these evocative scenario names
suggest ­— Global Orchestration, TechnoGarden, Order from Strength, Adapting Mosaic — exploratory scenarios attempt to
model the implications of a complex group of interrelated factors, rather than one isolated variable.
Normative scenarios take this approach a step further by starting with a range of possible paths and modeling the implications
of each of them for key parameters, from water and land use to climate mitigation and food production. Within this category,
“transformative” modeling is the most ambitious, as the researchers define a desired future and model what it would take to
get there. One of the more comprehensive such efforts is the Agrimonde project, which is discussed below.
Each of these modeling approaches offers valuable insights. Normative modeling may be the most provocative as it tends to
highlight the distance between our present path and an optimal one in light of a particular goal. Similarly, “strategic” modeling
charts different paths, futures, and implications, offering sometimes stark contrasts between distinct trajectories. This information
is invaluable as societies struggle to confront complex challenges and respect the carrying capacity of the planet. However in
both approaches, the complexity of the changes being modeled makes it difficult to determine the impact of any one discreet
factor. For better or worse, the fact is that policymakers are usually only able to address particular factors in isolation. Reilly
and Willenbockel cite a small body of literature suggesting very limited policy impacts as a result of such modeling.27
10
Biofuels expansion
adequate attention to consumption data and patterns.
Van Dijk highlights promising new efforts within the EU’s
Foodsecure Project that incorporate detailed household survey data on consumption into such modeling.21
Tomlinson, in turn, recommends stepping outside such
economic models to learn from efforts that take into account concerns for health, equity, and the environment
and model alternatives to “productivist” solutions.
It is surprising that there has not been more careful
modeling of the impacts of biofuels expansion on 2050
food production despite widespread agreement that it is
an important factor. Unlike the case of climate change,
the key variables are well-defined. The impacts in the
case of first-generation biofuels, in terms of crop diversion and land use, are direct. Policies promoting biofuels
production and use, such as consumption mandates,
are clear and in place. And the trajectory, at least over
the period of the established mandates (10-12 years),
is well specified. Variables ripe for modeling include:
changes to biofuels policies; energy prices; and the
timing, scale, and impact of second generation biofuels
as they are developed and come onto the market.
SCENARIO MODELING FOR POLICY FORMULATION:
GAPS AND CHALLENGES
Modeling that begins with a narrowly defined question
(such as “what if” scenario modeling in which just one
factor is altered to gauge the impact), can be more
directly useful to policymakers than larger, more
complex models (see text box 2). If the baseline
established by the model is transparent, realistic, and
considers the uncertainties inherent in such long-range
efforts, the introduction of a limited set of plausible
policy or parameter changes can identify a range of
achievable impacts. These, in turn, can inform policymakers as they consider discreet alternatives.
One reason many of the 2050 models take inadequate
account of biofuel expansion is the difficulty of incorporating such a recent phenomenon into scenario modeling that necessarily relies on base years that predate
the biofuels boom. Many 2050 models have a base year
of 2000, and even the relatively updated models, with
2005 base years, do not capture much of the recent
surge in biofuel production. To the extent such models
fail to account for biofuel expansion in their scenario exercises, their results will be of limited use to policymakers concerned with biofuels’ impact on food security.
Some of the factors that drive estimates of 2050 food
supply and demand have been modeled better than
others. While uncertainties plague attempts to estimate
agricultural productivity growth, historical trends do offer
reliable baselines for assessing future growth, and public
policies can demonstrably increase growth rates. A wide
range of scenario studies illustrate the importance and
effectiveness of policies that foster productivity growth,
and many of these are potentially useful to policymakers.vi
In any case, the results of 2050 modeling will be driven
to a significant extent by the assumptions modelers
make about biofuels expansion. For example, the FAO
in its 2012 update incorporated more recent estimates,
but assumed expansion only to 2019, with no further
expansion to 2050. Simultaneously, they acknowledged
that this would likely underestimate biofuel demand,
which they correctly note will be driven significantly by oil
prices.
Two factors in particular — biofuels expansion and
climate change — have thus far been poorly incorporated
into most foresight modeling on meeting future food
needs. Here we examine some of the reasons for this
shortcoming and review some of the more significant
efforts to project the impacts of various biofuel and
climate scenarios. Where climate change presents
modelers with a daunting range of uncertainties, it is
easier to project the impact of biofuels expansion on
food production and availability.
Although it is flawed and quite out of date, the detailed
effort by Günther Fischer of the International Institute for
Applied Systems Analysis (IIASA), presents an example
of the kind of “what if” scenario modeling that policymakers require to evaluate biofuel expansion. Using a
For a more extensive discussion of agricultural productivity
scenario modeling, see the Tufts University study. “Can We Feed
the World in 2050? A Scoping Paper to Assess the Evidence,”
by Timothy A. Wise, GDAE Working Paper No. 13-04, September
2013. http://www.ase.tufts.edu/gdae/policy_research/FeedWorld2050.html
vi
11
modeling framework based on the FAO/IIASA Agro-Ecological Zone (AEZ) model and the IIASA World Food
System model, Fischer models a series of biofuels
expansion scenarios in addition to several climate change
scenarios.
The results rely on base year data from 2000 and
incorporate additional data only up to 2008, but still
speak to the consequences of today’s policy and
pathway choices. The table below summarizes the key
scenarios and results.28 (See Table 1)
current path, suggesting a 13% increase in food
prices attributable to biofuels expansion by 2050.
This modeling is not perfect, of course. Because the
base year is outdated, it makes more sense to use the
results to gauge the price impacts of different scenarios
in relation to one another, rather than in relation to the
base year.vii
Still, it is easy to see how this sort of “what if” scenario
modeling can offer policymakers a clearer picture of
the impacts of their policies. As Fischer’s work indicates, in order to feed the world affordably by 2050 we
must speed up the development and deployment of
advanced biofuels and/or slow demand for first generation biofuels by reducing mandates. Otherwise we are
putting undue pressure on low-income consumers (via
price hikes) and on the environment (via land and other
resource demands).
The results are instructive, providing estimates of food
price increases (an economic measure of supply in
relation to demand), hunger, and agricultural land use
change. To summarize:
n The so-called WEO-V1 scenario, which uses 2008
International Energy Agency projections for demand, shows that even moderate additional biofuel
demand over 2008 levels raises food commodity
prices 7%, increases those at risk of hunger by 21
million people,
and requires a 21% increase in cultivated land, even
with early and gradual deployment of second
generation biofuels.
It is surprising that more up-to-date modeling has not
been carried out on an issue that is so important to
future food security. Table 2 below shows just how
outdated most biofuels modeling assumptions are in
comparison to the best current estimates, namely the
World Energy Outlook (WEO)’s 2012 projections for
first generation biofuels use. The FAO estimate, which
relies on FAO-OECD projections to 2020, but then holds
biofuels use constant through 2050, is less than half of
WEO’s current projections for 2030. The International
Food Policy Research Institute (IFPRI)’s IMPACT model
uses old base year assumptions and assumes minimal
growth. World Bank projections do not explicitly account
for growing biofuel use. The ongoing Agricultural Model
n If advanced biofuels are not available until 2030
(WEO-V2), prices increase 11%, hunger risks rise to
42 million, and there is a 29% increase in cultivated
land. In other words, delays in the deployment of
advanced biofuels have serious implications.
n Scenario TAR-V1 uses what we now know are more
realistic estimates of demand, in light of current
mandates and targets. This raises projected biofuel
demand by 100%, with dramatic results — 20%
food price increases, 136 million at risk of hunger,
48% increase in cultivated land.
The FAO’s Alexandratos points out that the model overestimates
consumption levels overall, which raises demand in relation to
supply. Land is modeled inadequately, with additional demand
translating too readily into additional cultivated land, and with little
accounting for the land needs of second generation biofuels. He
also suggests that the model does not adequately account for
supply responses by farmers to higher agricultural prices, nor the
impact of oil prices on biofuel demand. Notably, only rain-fed land
is included in this modeling run. Alexandratos, Nikos (2011).
Critical evaluation of selected projections. Looking Ahead in
World Food and Agriculture: Perspectives to 2050. P. Conforti.
Rome, Italy, Food and Agriculture Organization: 465-508.
vii
n Only the rapid deployment of advanced biofuels
(TAR-V3) mitigates these impacts, reducing price
impacts to only 9%, with 74 million additional
people at risk of hunger, and a 29% increase in
cultivated land.
Unfortunately, thus far AgMIP has provided new modeling only
to estimate the implications of faster deployment of secondgeneration biofuels.Lotze-Campen, Hermann, Martin von Lampe,
Page Kyle, Shinichiro Fujimori, Petr Havlik, Hans van Meijl,
Tomoko Hasegawa, A. Popp, Christoph Schmitz, Andrzej Tabeau,
Hugo Valin, Dirk Willenbockel and M. Wise (2013). “Impacts of
increased bioenergy demand on global food markets: an AgMIP
economic model intercomparison.” Agricultural Economics
forthcoming.
viii
n The four sensitivity analyses (SNS V1-V4) show the
increasing impact on agricultural prices of different
expansion scenarios for first-generation biofuels,
from 4% in the low case to 35% in the high case.
Note that based on current estimates, the “medium”
scenario (SNS-V2) may well be the closest to our
12
Intercomparison and Improvement Project (AgMIP) however (see text box below) assumes trend growth through
2030, which results in estimates comparable to WEO’s.viii
advanced biofuels, but as explained above Fischer’s
modeling still needs improvement to be useful. Updating
Fischer’s base year and modeling parameters, addressing the shortcomings noted by Alexandratos in footnote
9 above, and adjusting scenarios to reflect current policy
considerations would yield results that could enable
policymakers to better manage biofuel expansion to
avoid negative impacts on food security.
As the table shows, of the IIASA scenarios modeled
by Fischer, his worst-case scenario (TAR-V1) seems
the closest to business-as-usual projections, at least
through 2030. Fisher’s TAR-V1 scenario assumes the
fulfillment of mandates and the slow deployment of
TABLE 1 — Modeling Biofuel Scenarios
CHANGE IN 2050 ADDED PEOPLE AT RISK
LAND USE
OF HUNGER IN 2030 CHANGE 2050(%)²
SCENARIO NAME
DESCRIPTION
PRICE INDEX (%)
FAO-REF-01
Freezes expansion at 2008 levels¹
WEO-V1
IEA 2008 projections; 2nd generation from 2015
7
21 million
21
WEO-V2
WEO-V1, no 2nd gen. until 2030: all demand to 2030 from 1st gen.
11
42 million
29
TAR-V1
WEO-V1, but mandates fulfilled by 2020: doubles demand for 1st gen.
20
136 million
48
TAR-V3
TAR-V1, but quick 2nd gen: 33% global demand in 2020, 50% in 2030
9
74 million
29
SNS-V1
Scenarios based on share 1st gen. in transport fuels 2020, 2030, 2050:
Low: 2%, 2.5% and 3% 4
-
-
SNS-V2
Medium: 4%, 5% and 6%.
13
-
-
SNS-V3
High: 6%, 7.5% and 9%
23
-
-
SNS-V4
Very high: 8%, 10% and 12%
35
-
-
¹ Ref scenario: Price Index (1990 = 100): 115; Risk of hunger: 458 million; Cultivated land: 1.7 billion ha
² Relative to reference scenario of no agricultural crops used for biofuel production
Source: Fischer (2011)
TABLE 2 — Projected Cereal Use for 1st Generation Biofuels (millions tons)
MODEL
SCENARIO
WEO 20121
Current Projections
2010
Alexandratos (2012)
IMPACT2
Msangi (2011) Baseline
20352050
148
239
421
na
2008
2020
2030
2050
65 (2005)
182
182
182
20
75
110
110
REVIEWED MODELS
FAO
2020
World Bank van der Mensbrugghe (2011)
n/a
n/a
n/a
n/a
AgMIP4
baseline, 1st generation
148
273
397
397
IIASA
FAO-REF-01 (2008 levels)
83
83
83
WEO-V1
83
181
206246
WEO-V2
83
192
258376
TAR-V1
83
327
437446
TAR-V3
83
238
272262
3
1
2
3
4
83
2010 actuals from FAO-OECD (2012); WEO 2012 projected growth rate to 2020, 2035
Drawn from Alexandratos (2011), pg. 48
Numbers not currently available; Author assumes fulfillment of mandates as of 2009
2010 actuals from FAO-OECD (2012); AgMIP growth rate for other years (Lotze-Campen 2013)
13
Climate Change
Africa which assumes that adaptation measures restore
agricultural productivity levels in this region to those
in his non-climate-change reference scenario. Price
increases in this scenario are significantly moderated,
even with the assumed failure of mitigation or adaptation
efforts in other parts of the world. He concludes from
this that investment in regional climate adaptation would
have significant benefits for agricultural production and
food security.31
Climate change is one of the most difficult variables to
model due to the layered uncertainties associated with
climate change and its impacts on agriculture:
n There is uncertainty about the extent and timing of
climate change likely to result from current levels of
emissions.
n Projections to 2050 and beyond must account
for uncertainties related to the extent and pace of
mitigation going forward, including whether global
policy changes will reduce emissions and future
impacts.
Günther Fischer incorporates into the IIASA World Food
Systems model two different climate modelsx for running
scenarios with and without CO2 fertilization. Fischer
runs his scenarios out to 2080, which is useful given the
likelihood of rising climate change impacts over time,
but this model is limited in that it examines the impact
of climate change on rain-fed, but not irrigated, land.32
Fischer draws three conclusions from his scenario runs:
n The impact of temperature changes on the earth’s
ecosystems is only imperfectly understood.
n There is great uncertainty about the impacts of
temperature and ecosystem changes on agricultural
production.
1. The regional impact of climate change is significant,
posing threats to future food production.
n It is difficult to predict with any certainty what
adaptation measures will be taken and how effective
they will be.
2. There could be some improvements in rain-fed
production if there is a positive CO2 fertilization
effect and if farmers are able to adapt to changing
climates.
n Impacts are expected to worsen over time, making
the 2050 time horizon too short to gauge the full
extent of climate risks.
3. The post-2050 period is particularly worrisome, as
modeling projects increasingly negative and rapid
impacts on production in most regions.
Below we highlight scenario modeling that exemplifies
the state of the art while demonstrating the profound
difficulties involved in dealing with so many uncertainties.ix
Fischer also combines both biofuels and climate change
impacts in the modeling scenarios outlined in the
biofuels section above. If his most pessimistic scenario
(TAR-V1) is indeed the closest to the path we’re now
following, and if there is limited CO2 fertilization to
moderate the impact of climate change, the prospects
are grim. In this scenario, Fischer’s modeling shows
price impacts on cereals growing from 49% in 2020,
to 53% in 2050 and expanding to 87% in 2080, though
it is worth noting that his baseline for comparison
assumes no climate change and little biofuel expansion.
As with scenario modeling for other key variables, some
research is intended less to forecast possible outcomes
than to demonstrate the importance of the issue through
quantitative methods. Dirk Willenbockel relies on an
impressive range of external estimates gauging the
productivity impacts of climate change to 2030, by
region and crop.29 His assumptions are pessimistic,
based on high temperature changes and sensitivities
of crops to warming and low CO2 fertilization effects.
In comparison to his reference scenario, price increases
are significantly higher in 2030 for the main grains (about
140% versus roughly 90%), with corn experiencing the
largest increases. Notably, the impacts are especially
severe in Sub-Saharan Africa.30
See Wright, Julia (2010). Feeding Nine Billion in a Low Emissions
Economy: Challenging, but Possible: A Review of the Literature
for the Overseas Development Institute. Oxford, UK, Oxfam
International, Overseas Development Institute. Wright provides
an extensive review of the literature on climate mitigation and
the potential impacts of climate change on our ability to feed the
world in 2050. She finds few studies that address the intersection
of these two questions, but a good deal on each individually.
ix
Interestingly, Willenbockel includes a scenario of
successful climate change adaptation in Sub-Saharan
HadCM3 and CSIRO models using IPCC A2 emissions pathways
in the modeling.
x
14
IFPRI researchers ran a series of climate scenarios
using their IMPACT model. As with Fischer, the baseline
is perfect mitigation, i.e., no change in the climate from
2010 forward. These simulations allow researchers
to estimate the economic and social impacts of four
climate change scenarios of increasing intensity, defined
as increasingly hotter and wetter. Using their “middle-ofthe-road” assumptions of GDP and population growth
from 2010 to 2050, IFPRI modelers find that prices for
the three main grain staples would rise between 20
and 32% on average in comparison to the baseline
case of no climate change. As one would expect, the
most intense climate change scenario generates price
increases 25-30% higher than that. Using their “middle
of the road” assumptions, the researchers go on to
estimate that within their four climate change scenarios,
the number of malnourished children in 2050 increases
by 8.5%-10.3%.33
for Climate Systems Research, is directly confronting
the wide variability in climate-related modeling simulations. (See text box 3) This important work aspires to
harmonize models by introducing common parameters
and assumptions and imposing common scenarios.
The initial set of comparisons shows negative impacts
on food security across nearly all the included models.
This is a noteworthy change from previous estimates
of climate impacts on agriculture, some of which
suggested production gains in many regions.
The authors of a forthcoming article on the AgMIP
climate modeling in Agricultural Economics, emphasize
that “assumptions that are typically buried in technical
reports can have significant effects on highly visible
output measures.” Due to the scale of these often
invisible impacts, they propose that the effects of
underlying modeling assumptions should be a high
priority for further research and call for more reliable
agricultural data so that crop models can be
fine-tuned.35
The authors also use the IFPRI model to simulate an
extended drought in South Asia from 2030 through
2035 — a frequently forecasted impact of climate
change. This simulation provides an opportunity to
examine aspects of climate change that are difficult
for long-run modeling to predict — rising temperature
variability and extreme weather events. The results
include sharp price increases for all three major grains
during the drought years and a dramatic increase in
the number of malnourished children.
Researchers affiliated with Tufts University and the
Stockholm Environment Institute warn that many of the
existing 2050 scenarios fail to reflect the most recent
climate science, which suggests greater disruptions to
agriculture due to climate change.36 Of course, 2050
scenarios fail by definition to reflect climate change
impacts in the second half of the century, when the
effects of a changing climate are projected to worsen
significantly. Those who have extended their projections
to 2080 and beyond warn of escalating impacts on
agriculture after 2050.37
Other researchers used IIASA’s Global Biosphere
Management Model (GLOBIOM) to assess the food
security impacts of yield uncertainty due to climate
change. They conclude that high levels of uncertainty
about yields increase the need for decision-makers to
plan for overproduction, a feasible but potentially costly
practice. Their modeling also highlights the importance
of reducing trade barriers to allow agricultural products
to flow from surplus to deficit regions and expanding
irrigation to help stabilize yields and increase production.
Finally, they note the value of expanding global storage
capacity for basic grains in order to reduce vulnerability to the short-term yield variations expected to come
with climate change. This is one of the few mentions we
found in the 2050 literature of the importance of food
reserves to global food security.34
MODELING BROADER QUESTIONS
Of course the set of questions policymakers face as
they contemplate the long-range future is too broad
to be answered solely through examination of policy
choices related to agricultural productivity, biofuels
expansion, and climate change. As a matter of
necessity, approaches to answering these questions
go well beyond economic modeling to include
biophysical modeling of key natural resources.
(See Box 4 on land use and Box 5 on water use)
The Agricultural Modeling Improvement Project (AgMIP),
housed at Columbia University’s Earth Institute Center
Key concerns within this broader sphere include the
15
BOX 3 — The Agriculture Modeling Improvement Project (AgMIP)
A forthcoming special issue of Agricultural Economics makes an important contribution to long-range global agricultural
modeling. The issue presents analysis from the Agriculture Modeling Improvement Project (AgMIP), a collaborative effort to
improve modeling, particularly as it relates to key uncertainties such as climate change and bioenergy production. The issue
compares results from ten different models by harmonizing some of their key assumptions regarding population growth,
GDP growth, agricultural productivity growth, energy prices, and base years before introducing alternative socio-economic,
climate change, and bioenergy scenarios.xi The primary goal is to improve modeling by identifying important differences in
assumptions, but the comparison offers a rich set of results across a range of economic models.
While the simulations are not necessarily grounded in the most realistic scenarios, they permit a number of important
conclusions about climate change and economic modeling. These conclusions and findings, summarized in the overview
chapter that was generously made available for this review, include:
1. The underlying assumptions of each model in the AgMIP comparison, account for significant differences in modeling
results. Controlling for these differences narrowed the range of results considerably, reinforcing the importance of
consistency and transparency in modeling assumptions.
2. All of the controlled variables significantly impact results well out into the future and all present important uncertainties
that cannot be resolved conclusively. A scenario changing assumptions from “middle of the road” to a higher population,
lower GDP growth scenario demonstrates just how bad the outcome could potentially be for production, consumption, and
prices. If population in developing countries grows 11% more by 2050 and GDP growth is over 30% lower due to slowed
annual growth rates over time, global per capita calorie consumption will be 6-10% lower on average, with even worse
impacts in poorer regions.
3. In contrast to many earlier estimates, climate change scenarios show clear negative impacts on yields at the global level.
Estimated price impacts ranged from +2% to +79%. Per capita calorie availability declined across the globe.
The authors conclude that scenario modeling is critical to good policy and investment decisions and call for improved
exchange and dialogue between decision makers and modelers.
extent to which developing countries adopt Western
diets high in animal protein, the sustainability of continued reliance on large quantities of fossil-fuel-based
inputs on large monoculture farms, the potential positive
impact of reduced inequality in access to food, and the
impacts of high price volatility on agricultural production
and distribution. Such questions do not lend themselves
easily to global quantitative modeling, precisely because
the range of variables is so high. Still, quantitative
projections remain important and modelers have utilized
so-called “exploratory scenarios” in an attempt to
quantify alternative paths.
develop a coherent set of scenarios — Shared Socioeconomic Pathways (SSPs) — that can be used in a
consistent way by researchers to generate comparable
assessments while pursuing multi-faceted exploratory
modeling.xii
Models participating in this comparison include: Asia-Pacific
Integrated Model (AIM) from the Japanese National Institute for
Environmental Studies; ENVISAGE, based on the World Bank’s
LINKAGE model, now housed at the FAO; Emissions Prediction
and Policy Analysis (EPPA) from MIT; Global Trade and Environment Model (GTEM) from the Australian institute ABARES; Future
Agricultural Resources Model (FARM) from the USDA; Modular
Applied GeNeral Equilibrium Tool (MAGNET) from Wageningen
University; Global Change Assessment Model (GCAM) from
the Pacific Northwest National Laboratory; Global Biosphere
Optimization Model (GLOBIOM) from IIASA; IFPRI’s International
Model for Policy Analysis of Agricultural Commodities and Trade
(IMPACT); Model for Agricultural Production and its Impact on
the Environment (MAgPIE) from the Potsdam Institute for Climate
Impact Research.
xi
These exploratory scenarios chart broad societal
pathways, quantifying some of the implications, and
highlighting the key drivers of change, and they can be
useful in shifting or informing our “mental maps” of a
particular issue. The applicability of future efforts will
benefit from a project currently underway through the
Intergovernmental Panel on Climate Change (IPCC) to
van Dijk, Michiel (2012). A review of global scenario exercises
for food security analysis: Assumptions and results. Food
Secure Working paper. Hague, Food Secure. See, too, a detailed
literature review on the evidence base for climate mitigation from
organic agriculture: Azeez, G. (2009). Soil Carbon and Organic
Farming. Bristol, The Soil Association.
xii
16
BOX 4 — Biophysical Modeling: Land Use
In most economic models, available agricultural land is identified through biophysical modeling. New land is projected to
be converted to agricultural uses to achieve equilibrium between supply and demand. In the case of the FAO model, 70 million
hectares of additional agricultural land is required by 2050, an increase of 9%. Assumptions about land availability are key
to the outcomes for other variables as well. A modeling constraint on the expansion of agricultural land will produce greater
supply and demand imbalances and higher agricultural prices, for example.
Scientists from the Rockefeller University’s Program for the Human Environment employ a land-use model to project the
pressures on uncultivated lands from the demands of a growing population and changing diets. These researchers optimistically
suggest that we may be nearing the point of “peak farmland,” meaning that so-called “land sparing” — the extent to which
uncultivated land remains in forest or other states — could actually increase by 2050. According to their calculations, existing
resources could be adequate to fulfill future food needs. However as the researchers themselves note, this finding could
potentially be invalidated by the “wild card” of global biofuels production.
Faculty from Chalmers University of Technology in Sweden incorporate FAO economic modeling into biophysical scenarios to
project the impacts on land use. These modelers start with the FAO baseline (assuming that by 2030 global agricultural area will
expand from the current 5.1 billion ha to 5.4 billion ha) and proceed to model the land use impacts of three ”what if” scenarios.
Faster growth in livestock production efficiency (basically feed efficiency) is projected to reduce land demands from 5.4 billion
to 4.8 billion ha. A 20% shift from beef to less land-intensive pork and poultry reduces demand further, to 4.4 billion ha. Finally,
a broader shift to diets relying less on meat and reductions in food waste in the world’s wealthier regions reduce land use by an
additional 15%.
Other researchers have used global food demand estimates to model the land use implications of intensification versus
extensification scenarios. Using this approach, modelers from the University of Minnesota’s Department of Ecology, Evolution,
and Behavior find that meeting global food demand by bringing more land into production (extensification) would mean one
billion more hectares of land cleared by 2050, with high levels of greenhouse gas emissions and nitrogen use. By contrast,
moderate intensification of production in low-productivity regions would reduce land demands by 80%, cut emissions by
two-thirds, and slightly reduce global nitrogen use.
Here we highlight a few such efforts.
3. “Adapting Mosaic” – includes greater local and
regional diversity of development paths, as WTO
negotiations and global climate talks fail. The results
vary widely from region to region, and food production globally is much lower than in the previous two
scenarios.
The United Nations’ Millennium Ecosystem Assessment
(MA) examines four scenarios, summarized below:
1. “Global Orchestration” – is characterized by global
trade liberalization and cooperation as agriculture
moves toward large-scale industrial production,
with limited environmental management. The result
is high global per capita calorie availability, with a
40% decrease in child hunger. But environmental
damage is significant, including the projected loss of
50% of Sub-Saharan Africa’s forested land.
4. “Order from Strength” – features high trade barriers, limited global cooperation, and little attention
to ecosystem management. Weak agricultural
investment drives extensive cropland expansion,
and climate change contributes to hunger and mass
migration in Africa. Food output grows very slowly to
2050, as child malnutrition increases.
2. “TechnoGarden” – involves high technological
development, low trade and investment barriers,
and better environmental management in the global
North. Private investment transforms developing
country agriculture through intensification, dramatically increasing production in Sub-Saharan Africa.
Food production and hunger levels are similar to
scenario one.
For the UN researchers, these modeling results highlight
the dangers of “reactive” versus “proactive” environmental management and of fractured versus coordinated
global approaches to trade and economic cooperation.
In a similarly ambitious exercise, researchers at the
French organization CIRAD (Agricultural Research for
17
BOX 5 — Biophysical Modeling: Water Management
The Comprehensive Assessment of Water Management in Agriculture (CAWMA) is a recent attempt to integrate biophysical
and economic information for policy-relevant scenario assessment. This model relies on the International Food Policy Research
Institute’s IMPACT model to estimate food demand and supply and the International Water Management Institute’s WATERSIM
model to simulate the same for water. The modelers start with estimated crop demand to 2050 and estimate that water use for
such production will increase by 70%-90% under business-as-usual policies. They proceed to model five “what if” scenarios,
to assess the impact of different approaches to water resource investment. This detailed set of controlled simulations allows
policymakers to evaluate the likely impacts of different approaches to a narrowly defined policy challenge.
Notably, climate change is not included in this simulation, making the results less relevant as forecasts but more relevant to
policymakers, as the results of different approaches are more easily compared. The five scenarios are:
1. “Rain-fed Optimistic” – This scenario assumes no increase to 2050 in irrigated crop production, the development focus
instead being improved water management by poor, rural smallholders. The scenario is optimistic because it assumes
that 80% of yield gaps — the gap between current productivity and potential productivity — are closed. As a result, water
availability is adequate to meet global agricultural demand.
2. “Rain-fed Pessimistic” – Only 20% of rain-fed yield gaps are closed, necessitating a 53% increase in rain-fed cropland to
meet food needs. This has high environmental costs, and many countries have to increase their food imports significantly.
Food insecurity increases, to the highest level of any of the scenarios considered.
3. “Irrigation Expansion” – This scenario assumes large investments in irrigation, particularly in Sub-Saharan Africa and
Asia, to reduce food import dependence. The cost is high — US$400 billion — and the 33% increase in irrigated area
supplies less than one-quarter of the expected rise in food demand. Food security improves for many, but this approach
increases pressure on freshwater resources, more than doubling the number of people suffering water scarcity to 2.6
billion in 2050.
4. “Irrigation Yield Improvement” – This simulation prioritizes improved efficiency in the use of land and water. The results
include a 75%-80% reduction in yield gaps and a 9% global expansion of irrigated cropland, meeting half the additional
demand for food by 2050. It is not inexpensive — US$300 billion — and 32% of freshwater resources are diverted to
agriculture.
5. “Trade” – A final scenario examines the outcome when much of the rising demand for food is met by trade, with relatively
water-abundant grain producers — USA, Canada, Argentina, etc. — exporting greater volumes to water-deficient countries.
This scenario reduces water stresses, but many obstacles make it unlikely, including the high price of food imports for
less developed countries, the energy use required by international trade, and government wariness of excessive import
dependence.
The study concludes with a normative scenario designed to optimize water-resource investment, responding to the limitations
highlighted by each of the five simulations discussed above. Recognizing that different strategies will work best in different
regions, the authors model a portfolio approach. In Sub-Saharan Africa, for example, a dual approach is followed, including
significant expansion of irrigation for cash crops alongside efforts to achieve water-efficiency gains for smallholders on
rain-fed lands. Overall, the simulation projects yield increases of 58% for cereals on rain-fed lands, while irrigated agricultural
yields rise 55% based on 31% and 38% increases in water efficiency respectively. Strong regulations limit environmental
impacts, intensification limits demand for additional agricultural lands, and there is only a 13% increase in freshwater use for
agricultural production in 2050.
18
Development) used its Agribiom model in their Agrimonde foresight study, comparing a business-as-usual
scenario (high-growth industrial agriculture derived from
the MA’s Global Orchestration scenario) to one based on
equity and agro-ecological intensification of agricultural
production. In both, it is assumed that the world produces enough food, the questions modeled are “how” and
“at what cost?”
productivity, and land use. Climate change and bioenergy scenarios are modeled, as are the impacts of four
different diets, from “western high meat” to relatively low
meat-based diets. Livestock models range from intensive to humane to organic. Land use for crops is varied
from the FAO’s baseline of a 9% increase up to19%. The
result is an impressive but somewhat dizzying array of
72 distinct scenarios.
For the Agrimonde project, diets of 3,000 kilocalories
are assumed for all people in all countries. The researchers find that the world can meet future food needs even
as yield growth slows due to a shift away from industrial agriculture and toward smaller scale farms and
agro-ecological practices. This exercise clarifies some
of the perhaps-undesirable impacts of such a path. To
achieve this hypothetical future, there would need to
be a 39% increase in cropland to make up for low yield
growth, though as researchers point out, this can come
largely from pasture as the world relies less on meatbased diets. They also project a stunning 740% increase
in global food trade, as surplus regions supply deficit
regions. (There is no accounting for how food-deficit regions would afford such a dramatic increase in imports.)
The Klagenfurt researchers synthesize their results to
assess a “wholly organic” future, as well as an intermediate scenario. They conclude that the present path
(high meat, industrial production) is feasible only with
increased conversion of new crop and pasture land
(20% instead of the FAO’s baseline of 9%) alongside
intensification of production. The low-meat, organic scenario was also found to be feasible, with a similar level of
farmland expansion (for grains for human consumption
in place of grazing for meat production). The impacts
of climate change would depend a great deal on the
unknown effects of CO2 fertilization.
The International Food Policy Research Institute in turn
has modeled the impact of a 50% reduction in meat
demand to 2050, in developed countries as well as
in China and Brazil. Not surprisingly, the results show
significant reductions in agricultural prices (higher supply
relative to demand) and in food insecurity, with the
changes in China and Brazil having the greatest impact.
A modeling effort led by the Institute of Social Ecology
at Vienna’s Klagenfurt University presents an extensive
attempt to model alternative scenarios. Like Agribiom, this model is not economic, based instead on the
supply and demand for available biomass from crop and
grazing land. Researchers use FAO estimates for their
baselines on population, economic growth, agricultural
19
CONCLUSIONS
uncertainties inherent in all such projections alongside
known factors — particularly biofuels expansion and
climate change — that have not yet been adequately
considered and addressed.
While this review does not represent an exhaustive
assessment of the literature, it should at least provide
a basic understanding of the primary sources for many
of the widely quoted estimates of global food needs in
2050, alongside some possible alternatives. Our review
suggests that there is much to be gained from careful
modeling based on reliable data that attempts to
integrate biophysical and economic trends and
relationships.
The forthcoming overview of the AgMIP project in
Agricultural Economics, concludes with a strong call for
modeling that utilizes a range of assumptions and presents a range of results, as a way to maximize its usefulness to policymakers. “Exploring the outcomes from a
range of plausible drivers is essential, not least
as these drivers in part depend on decisions on public
policies and private investments.”47 In other words, the
most useful drivers to model are those that are most
susceptible to policy intervention. Beyond the limited
predictive value of economic modeling, a key purpose
of such research should be to illuminate the consequences of policy and behavioral change. Biofuels
scenario modeling, for example, has not received the
attention it deserves given the potential impacts on food
production and resource use, as well as the prominent
role of public policies in promoting (or potentially discouraging) such use.
At the same time, a careful review of current modeling
literature cannot fail to highlight the high level of
uncertainty in biophysical and economic data and
the sensitivity of long-range results to small variations
in basic assumptions that together make any such
modeling projections highly speculative. Harkening back
to Reilly and Willenbockel’s words from the introduction,
modeling outcomes should not be taken as forecasts.
But unfortunately, that is often how they are presented,
usually through no fault of the modelers themselves.
This much is clear — the path we are currently on does
not lead to a Malthusian future, and. there is no need
for the alarmist “productivism” often occasioned by
decontextualized and overly simplified presentations
of modeling outcomes. Warnings of future scarcity
often fuel misguided campaigns for the expansion of
large-scale, resource intensive, corporate-controlled
agriculture, as exemplified by the G8’s New Alliance
for Food Security and Nutrition. They drive large-scale
land acquisitions and smallholder dispossession in the
name of looming food scarcity and the expansion of
input-intensive industrial farming, despite the resource
constraints that prompted 400 scientific and social
science experts from around the world to call that
development path into question through the
International Assessment of Agricultural Science
and Technology for Development (IAASTD).
Perhaps even more striking is the absence of the
potential impacts of reductions in food loss and waste,
which currently prevent an estimated one-third of
food from nourishing anyone.48 The FAO is actively
campaigning on this issue, focusing on simple and
achievable solutions, such as improved storage and
infrastructure in developing countries and standards
and public education to reduce retail and consumer
waste in developed countries. Surely those who warn
of looming food shortages can develop clear scenario
modeling to estimate the potential impact of, for
example, a ten percentage point reduction in food loss
by 2050. Under such a scenario, the FAO’s baseline
estimate of 2050 agricultural needs might drop again,
from 60% to 50%. Some researchers have even
estimated that reducing loss and waste to currently
achievable levels could cut losses in half and provide
enough additional food for one billion people.49
The FAO’s 2012 call for a 60% increase in agricultural
supply should be viewed as a starting point for
discussion, not a looming threat. This significantly
reduced projection should allay some of the worst
fears that population growth and changing diets will
necessarily overwhelm our ability to produce food.
At the same time consideration of the FAO’s findings
should lead us to consider and accommodate the
While modeling can be invaluable in evaluating
specific policy options, such as biofuels mandates
or steps to reduce food waste, its utility is less certain
when it comes to broader systemic issues or development paths, such as whether to pursue large-scale,
high-input agriculture or foster smaller scale, low-input
20
systems. Current modeling on the question of “feeding
the world” tends to be biased by assumptions about
basic aspects of our food and agricultural systems that
urgently need to change: consumption patterns, unfair
distribution and access, land use, trade structures, and
investment priorities, for example. In this context,
modeling a shift to agro-ecological systems can yield
global scale results driven by other underlying assumptions — dramatic increases in land use because of assumed lower yields from organic agriculture for example,
or vast expansion of trade in agricultural commodities to
compensate for assumed slower productivity growth.
A growing body of experience and evidence at local
and regional levels demonstrates the long-term value
of investments in smallholder farming and sustainable
agricultural methods. Policymakers would do well to
examine these on-the-ground experiences as a basis
for policy changes and investments that can scale-up
successful approaches in regions where they are most
appropriate and necessary, regions in which food
security is tenuous despite high agricultural potential.xiii
As a general rule, productive engagement with
affected communities and policymakers in developing
countries should guide future-oriented “foresight”
research, grounding such modeling exercises in
real-world experiences and developing-country priorities.
In a broad critique of predominant social science
research on food systems Thompson and Scoones
of the UK-based Institute for Development Studies
emphasize this point, calling for multi-disciplinary
approaches that look beyond economic growth, focus
on sustainability, and rely on participatory research.50
Only modeling that genuinely reflects and responds to
the lived experience of the world’s most food-insecure
communities will enable policymakers, advocates and
the communities themselves to work toward a future
where all have enough.
See Jules Pretty’s work on the sustainable intensification of
African agriculture, for example, which synthesizes a wide range
of field experience involving thousands of farmers and millions of
hectares of land across a wide range of countries: Pretty, Jules ,
Camilla Toulmin and Stella Williams (2011). “Sustainable Intensification in African Agriculture.” International Journal of Agricultural
Sustainability 9(1).
xiii
21
RECOMMENDATIONS
more slowly than is now likely. AgMIP’s attention to
the impact of advanced biofuels after 2030 is helpful, but it begs the more important question
facing policymakers: Can we feed the world if we
are simultaneously feeding our vehicles from the
same crops and lands?
The review above leads us to the following recommendations regarding future modeling to help policymakers
address the challenges that lie ahead:
n Global estimates are useful, but national and
regional figures are far more instructive, as are
local and regional strategies. As noted earlier,
there is no “we” who feed “the world.” Global
adequacy of projected food supplies can hide a
plethora of shortages and injustices at the regional,
national, and local levels. This dynamic will worsen
as climate change affects agricultural systems, as
many of the world’s poorest regions are expected
to be hit the hardest. While trade will address a
significant portion of local shortfalls, it would be
a mistake not to assess regional food production
capacities and focus on closing yield gaps, as many
researchers have suggested.
n Other contributors to supply and demand
imbalances deserve attention. This need not
come from economic modelers, but it could.
Reductions in food waste — farm-to-market as well
as at the retail and consumer levels — can have a
direct impact on food accessibility and supply. An
estimated one-third of all food is lost or wasted.51
Obviously, reducing such losses would make a great
deal more food available for consumption while
reducing resource use. It is striking that the principal
economic models reviewed here do not model
scenarios of food-loss reduction, which in developing countries could come from improvements
in storage and other infrastructure to reduce
post-harvest and processing losses. This is, in itself,
a valuable development goal.
n Climate change poses a particular challenge
for modelers. The AgMIP effort is an important
attempt to improve modeling and increase its value
for assessing climate change. Uncertainties will
inevitably remain of course, and it is incumbent on
researchers to present their results transparently,
especially the sources and levels of uncertainty.
Moreover, basic modeling parameters must be
established so that FAO and other widely quoted
studies on feeding the world in 2050 account for
climate change.
n Investment in developing country agricultural
productivity is a top priority. We have seen how
sensitive future food supplies are to increases and
decreases in agricultural productivity. While it would
be helpful if modelers could improve the quality and
consistency of the data they use, we do not need
to wait for their projections to invest in developing
country agriculture. Supporting the world’s smallholder farmers — who produce most of the food in
developing countries while simultaneously constituting most of the world’s hungry people — is the
single biggest opportunity to reduce hunger and
increase food security worldwide. Public investment
is particularly critical in reaching this community as
services for smallholders are less likely to be viewed
as profitable by the private sector.
n Biofuels expansion deserves more focused
attention from modelers. First generation
biofuels compete directly and indirectly for crops
and resources, exerting upward pressure on food
prices. Second generation biofuels should exert
less pressure, but uncertainties remain. Meanwhile,
most economic models to date have failed to take
into account the likely and possible pathways for
first generation biofuel expansion. To the extent that
biofuels markets are driven by consumption mandates and other public policies, it is incumbent on
modelers to represent the costs of such policies in
their long-range projections. To date, they haven’t.
Fischer’s excellent scenario work with the IIASA
model is out of date, and subsequent efforts rely on
unfounded assumptions that have biofuels expanding
n The priorities that should guide those
investments are less a modeling question than
a contextual public policy question. As the
CAWMA project on water investments suggests
(Box 5), a portfolio approach tailored to the specific
needs of a given region is the best way to maximize
productivity within given resource constraints. The
same will be true for other agricultural investments.
22
And failing to raise productivity and food security in
the regions that have historically been left behind
will have dire consequences for the world’s poor, as
indicated by the normative modeling used by the
Agrimonde and Millennium Ecosystem Assessment
projects.
On the other hand, energy markets may ultimately
prove decisive in determining the expansion
trajectories for biofuels. In this case, we may
need to consider a different set of national and
international policies to regulate the impact of
energy prices on food security.
n Scaling up proven strategies is the best path
to food security. We have seen how alarmist
projections can lead to reflexive investments in
large-scale industrial agriculture, the G8’s New
Alliance initiative being the most prominent example
of late. Yet long-term resource constraints will affect
many of the fossil-fuel-based inputs on which such
systems depend. Global long-range modeling
provides less guidance in such matters than do
case studies detailing proven strategies for the
sustainable intensification of food production.
n Better integration of energy scenarios into
agricultural modeling: The long-term energy
outlook has implications far beyond biofuel
expansion. It is directly relevant to the economic
viability of high-input agriculture, reliant as it is on
fossil-fuel-based inputs. If these inputs are expected
to rise in cost (or decline in availability), developing
countries should be wary of agricultural development paths that increase their reliance on them.
Is IAASTD right that business-as-usual is not an
option, and if so what are the alternatives?seareeds
n 2050 is too short a time horizon to assess
long-term sustainability. As noted earlier, climate
change and related resource constraints will have
increasing impacts over the course of the century.
Most models project far greater disruption to
agricultural production after 2050 than before.
Policymakers need to recognize that 2050
estimates are not the end of the story and may
be overly optimistic.
n Climate change and its impacts on agriculture:
Building on AgMIP’s valuable initiative and the
scientific community’s ongoing efforts to improve
our understanding, additional research on regional
and crop-specific impacts is needed, particularly
in the developing world. Adaptation needs and
strategies are local, and must be guided by a sound
local agricultural and economic analysis. As a matter
of necessity, this research would focus less on
global food provision and more on local strategies
for climate-resilient agriculture.Future Research eds
n Long-range economic modeling generally
minimizes the impacts of volatility, as supply
and demand resolve within the models. Yet
we anticipate a future of increased volatility in the
weather, affecting agricultural systems, and in
agricultural markets, thanks to thin reserves and
increasing financial speculation. More work is
needed to incorporate volatility into global
long-range economic modeling.52
n Synthesizing and scaling up successful local
efforts to promote increased food production
through climate-resilient sustainable methods.
Far too little research draws on the vast wealth of
field experience, much of which has been led by
farmers working closely with researchers and
extension agents. Scaling up successful strategies
can close yield gaps efficiently, improving food
security for the most vulnerable in a sustainable way.
Future Research Needs
n Bringing stakeholders to the table: There must be
a process to bring researchers into direct dialogue
with farmers and developing country governments,
so that research priorities and methodologies can
be informed by the communities that will (ideally)
benefit from the results.
n Biofuels scenarios, including public policy
analysis: Research needs to catch up to the latest
developments in global biofuels markets, including
careful modeling of biofuels expansion scenarios
and, to the extent possible, analysis of the impacts
of consumption mandates and other public policies.
Such research is important because, on the one
hand, such policies may undermine food security.
23
END NOTES
1
Reilly, Michael and Dirk Willenbockel (2010). “Managing uncertainty: a review of food system scenario analysis and modelling.”
Philosophical Transactions of the Royal Society 365: 3049-3063. Page 2053.
2
World Food Program. (2009, 1/26/2009). “World must double food production by 2050: FAO chief.” Retrieved 10/18, 2012, from
http://www.wfp.org/content/world-must-double-food-production-2050-fao-chief.
3
Rijsberman, Frank. (2012, 9/19/2012). “CGIAR: a global research partnership for a food secure future.” Retrieved 2/5/13, from
http://www.cgiar.org/consortium-news/cgiar-global-research-partnership-for-a-food-secure-future/.
4
Malthus, Thomas R. (1798). An essay on the principle of population. London, J. Johnson.
5
FAO (2009). How to Feed the World in 2050. Rome, Italy, Food and Agriculture Organization.
6
Nwanze, Kanayo F., José Graziano da Silva, Ertharin Cousin and Emile Frison. (2012, 6/20/12). “Opinion: Food Security Must Be On The
Table At Rio+20.” Retrieved 2/20/13, from http://www.wfp.org/news/news-release/food-security-must-be-table-rio20.
7
OECD-FAO (2011). OECD-FAO Agricultural Outlook 2011-2020. Paris, OECD-FAO.
8
Hertel, Thomas W. (2011). “The Global Supply and Demand for Agricultural Land in 2050: A Perfect Storm in the Making?” American
Journal of Agricultural Economics 93(2): 259-275.
9
Lobell, D.B., K.G. Cassman and C.B. Field (2009). “Crop Yield Gaps: Their Importance, Magnitudes, and Causes.” Annual Review of
Environment and Resources 34(1): 179–204.
10
Monfreda, C., N. Ramankutty and J.A. Foley (2008). “Farming the planet: 2. Geographic distribution of crop areas, yields, physiological
types, and net primary production in the year 2000.” Global Biogeochemical Cycles 22(1): 1-19, Licker, R., M. Johnston, C. Barford, J.A.
Foley, C.J. Kucharik, C. Monfreda and N. Ramankutty (2010). “Mind the Gap: How do climate and agricultural management explain the
“yield gap” of croplands around the world?” Global Ecology and Biogeography 19(6): 769-782.
11
IEA (2012). World Energy Outlook Paris, International Energy Agency (IEA).
12
Alexandratos, Nikos and Jelle Bruinsma (2012). World agriculture towards 2030/2050: the 2012 revision. ESA Working paper. Rome, Food
and Agriculture Organization: 93.
13
Alexandratos, Nikos (2011). Critical evaluation of selected projections. Looking Ahead in World Food and Agriculture: Perspectives to
2050. P. Conforti. Rome, Italy, Food and Agriculture Organization: 465-508.
14
Fischer, Tony, Derek Byerlee and Greg Edmeades (2012). Crop yields and food security: will yield increases continue to feed the world?
Australian Society of Agronomy. Armidale, Australia, Australian Society of Agronomy.
15
Aexandratos, Nikos and Jelle Bruinsma (2012). World agriculture towards 2030/2050: the 2012 revision. ESA Working paper. Rome, Food
and Agriculture Organization: Table 4.13.
16
van der Mensbrugghe, Dominique, Israel Osorio-Rodarte, Andrew Burns and John Baffes (2011). Macroeconomic Environment and
Commodity Markets: A Longer-Term Outlook. Looking Ahead in World Food and Agriculture: Perspectives to 2050. P. Conforti. Rome,
Food and Agriculture Organization: 191-229.
17
Hillebrand, Evan Ibid.Poverty, growth and inequality over the next 50 years: 159-190.
18
Tomlinson, Isobel (2011). “Doubling food production to feed the 9 billion: A critical perspective on a key discourse of food security in the
UK.” Journal of Rural Studies: 1.
19
Barrett, C. (2010). “Measuring Food Insecurity.” Science 327(5967): 825-828.
20
Lappe, Frances Moore, Jennifer Clapp and Timothy A. Wise (2013). Framing Hunger: A Response to ‘The State of Food Insecurity in the
World 2012’. Cambridge, Small Planet Institute.
24
END NOTES
21
van Dijk, Michiel (2012). A review of global scenario exercises for food security analysis: Assumptions and results. Food Secure Working
paper. Hague, Food Secure.
22
Conforti, Piero (2011). Looking Ahead in World Food and Agriculture: Perspectives to 2050. Rome, Italy, Food and Agriculture
Organization.
23
Reilly, Michael and Dirk Willenbockel (2010). “Managing uncertainty: a review of food system scenario analysis and modelling.”
Philosophical Transactions of the Royal Society 365: 3049-3063. (p 2053)
24
Reilly, Michael and Dirk Willenbockel (2010). “Managing uncertainty: a review of food system scenario analysis and modelling.”
Philosophical Transactions of the Royal Society 365: 3049-3063.
25
Carpenter, S.R., P.L. Pingali, E.M. Bennett and M.B. Zurek (2005). Ecosystems and human well-being: findings of the Scenarios Working
Group of the Millennium Ecosystem Assessment. Millennium Ecosystem Assessment Series. Washington, DC, Island Press.
26
Paillard, Sandrine, Bruno Dorin, Tristan Le Cotty, Tevecia Ronzon and Sebastien Treyer (2011). Food Security by 2050: Insights from the
Agrimonde Project. European Foresight Series EFP Brief, CIRAD, INRA, Paillard, Sandrine, Sebastien Treyer and Bruno Dorin, Eds. (2011).
Agrimonde: Scenarios and Challenges for Feeding the World in 2050. Versailles Cedex, Editions Quae.
27
Reilly, Michael and Dirk Willenbockel (2010). “Managing uncertainty: a review of food system scenario analysis and modelling.”
Philosophical Transactions of the Royal Society 365: 3060.
28
Fischer, G. (2011). How can climate change and the development of bioenergy alter the long-term outlook for food and agriculture?
Looking Ahead in World Food and Agriculture: Perspectives to 2050. P. Conforti. Rome, Italy, Food and Agriculture Organization: 95-158.
29
Willenbockel, Dirk (2011). Exploring Food Price Scenarios Towards 2030 with a Global Multi-Region Model. Oxfam Research Reports.
Oxford, Oxfam International: 26.
30
Willenbockel, Dirk (2011). Exploring Food Price Scenarios Towards 2030 with a Global Multi-Region Model. Oxfam Research Reports.
Oxford, Oxfam International.
31
Ibid.
32
Fischer, G. (2011). How can climate change and the development of bioenergy alter the long-term outlook for food and agriculture?
Looking Ahead in World Food and Agriculture: Perspectives to 2050. P. Conforti. Rome, Italy, Food and Agriculture Organization: 95-158.
33
Nelson, Gerald C., Mark W. Rosegrant, Amanda Palazzo, Ian Gray, Christina Ingersoll, Richard Robertson, Simla Tokgoz, Tingju Zhu,
Timothy B. Sulser, Claudia Ringler, Siwa Msangi and Liangzhi You (2010). Food Security, Farming, and Climate Change to 2050:
Scenarios, Results, Policy Options. Washington, DC, International Food Policy Research Institute.
34
Fuss, Sabine, Petr Havlík, Jana Szolgayová, Erwin Schmid and Michael Obersteiner (2011). Large-Scale Modelling of Global Food
Security and Adaptation under Yield Uncertainty. EAAE 2011 Congress. Zurich, European Association of Agricultural Economists (EAAE).
35
Nelson, Gerald D., Dominique van der Mensbrugghe, Tomoko Hasegawa, Kiyoshi Takahashi, Ronald D. Sands, Page Kyle and Katherine
Calvin (2013). “Agriculture and Climate Change in Global Scenarios: Why Don’t the Models Agree?” Agricultural Economics forthcoming.
36
Ackerman, Frank and Elizabeth A. Stanton (2013). Climate Impacts on Agriculture: A Challenge to Complacency? Working Paper.
Medford, MA, Global Development and Environment Institute (GDAE).
37
Fischer, G. (2011). How can climate change and the development of bioenergy alter the long-term outlook for food and agriculture?
Looking Ahead in World Food and Agriculture: Perspectives to 2050. P. Conforti. Rome, Italy, Food and Agriculture Organization: 95-158.
38
von Lampe, Martin, Dirk Willenbockel, Elodie Blanc, Yongxia Cai, Katherine Calvin, Shinichiro Fujimori, Tomoko Hasegawa, Petr Havlik,
Page Kyle, Hermann Lotze-Campen, Daniel Mason d/Croz, Gerald D. Nelson, Ronald D. Sands, Christoph Schmitz, Andrzej Tabeau,
Hugo Valin, Dominique van der Mensbrugghe and Hans van Meijl (2013). “Why Do Global Long-term Scenarios for Agriculture Differ?
An overview of the AgMIP Global Economic Model Intercomparison.” Agricultural Economics forthcoming.
25
END NOTES
39
Ausubel, Jesse H., Iddo K. Wernick and Paul E. Waggoner (2012). “Peak Farmland and the Prospect for Land Sparing.” Population
and Development Review 38(Supplement): 217-238.
40
Wirsenius, Stefan, Christian Azar and Göran Berndes (2010). “How much land is needed for global food production under scenarios
of dietary changes and livestock productivity increases in 2030?” Agricultural Systems 193(9): 621–638.
41
Tilman, D., C. Balzer, J. Hill and B.L. Befort (2011). “Global food demand and the sustainable intensification of agriculture.”
Proceedings of the National Academy of Sciences of the United States of America 108(50): 20260–20264.
42
de Fraiture, Charlotte, Dennis Wichelns, Johan Rockström, Eric Kemp-Benedict, Nishadi Eriyagama, Line J. Gordon, Munir A. Hanjra,
Jippe Hoogeveen, Annette Huber-Lee and Louise Karlberg (2007). Looking ahead to 2050: scenarios of alternative investment
approaches. Water for food, water for life: a comprehensive assessment of water management in agriculture. D. Molden. London, UK,
Earthscan: 91–145.
43
Carpenter, S.R., P.L. Pingali, E.M. Bennett and M.B. Zurek (2005). Ecosystems and human well-being: findings of the Scenarios Working
Group of the Millennium Ecosystem Assessment. Millennium Ecosystem Assessment Series. Washington, DC, Island Press.
44
Paillard, Sandrine, Sebastien Treyer and Bruno Dorin, Eds. (2011). Agrimonde: Scenarios and Challenges for Feeding the World in 2050.
Versailles Cedex, Editions Quae.
45
Erb, Karl-Heinz, Helmut Haberl, Fridolin Krausmann, Christian Lauk, Christoph Plutzar, Julia K. Steinberger, Christoph Müller, Alberte
Bondeau, Katharina Waha and Gudrun Pollack (2009). Eating the Planet: Feeding and fuelling the world sustainably, fairly and
humanely – a scoping study. Social Ecology Working Paper. Vienna, Institute of Social Ecology, Faculty for Interdisciplinary Studies,
Klagenfurt University
46
Rosegrant, Mark W., Simla Tokgoz, Prapti Bhandary and Siwa Msangi (2013). Looking Ahead: Scenarios for the Future of Food. 2012
Global Food Policy Report. IFPRI. Washington, International Food Policy Research Institute (IFPRI).
47
von Lampe, Martin, Dirk Willenbockel, Elodie Blanc, Yongxia Cai, Katherine Calvin, Shinichiro Fujimori, Tomoko Hasegawa, Petr Havlik,
Page Kyle, Hermann Lotze-Campen, Daniel Mason d/Croz, Gerald D. Nelson, Ronald D. Sands, Christoph Schmitz, Andrzej Tabeau,
Hugo Valin, Dominique van der Mensbrugghe and Hans van Meijl (2013). “Why Do Global Long-term Scenarios for Agriculture Differ?
An overview of the AgMIP Global Economic Model Intercomparison.” Agricultural Economics forthcoming.
48
FAO (2011). Global Food Losses and Food Waste: Extent, Causes, and Prevention. Rome, UN Food and Agriculture Organization.
49
Kummu, M., H. de Moel, M. Porkka, S. Siebert, O. Varis and P.J. Ward (2012). “Lost food, wasted resources: Global food supply chain
losses and their impacts on freshwater, cropland, and fertiliser use.” Science of The Total Environment 438: 477-489.
50
Thompson, John and Ian Scoones (2009). “Addressing the dynamics of agri-food systems: an emerging agenda for social science
research.” Environmental Science & Policy 12(4): 386-397.
51
FAO (2011). Global Food Losses and Food Waste: Extent, Causes, and Prevention. Rome, UN Food and Agriculture Organization.
52
Munier, Bertrand R., Ed. (2012). Global Uncertainty and the Volatility of Agricultural Commodities Prices. Amsterdam, IOS Press.
26
ACKNOWLEDGMENTS
This report was written by Timothy A. Wise and Kristin Sundell, with
support from Marie Brill.
It is adapted with permission from the recent Tufts University working
paper, “Can We Feed the World in 2050? A Scoping Paper to Assess
the Evidence,” by Timothy A. Wise:
http://www.ase.tufts.edu/gdae/policy_research/FeedWorld2050.html.
Timothy A. Wise is Director of the Research and Policy Program at
Tufts University’s Global Development and Environment Institute.
He is grateful to Elise Garvey for invaluable research assistance and
to unnamed reviewers, whose comments and suggestions vastly
improved the content and presentation of this paper.
ActionAid works with people living in poverty and those acting in
solidarity to end poverty and injustice. We fight hunger, hold
companies and governments accountable, seek justice and
education for women, and help people to cope with emergencies
in over 40 countries around the world.
ActionAid USA
1420 K Street, NW, Suite 900
Washington, DC 20005
www.actionaidusa.org
(202) 835-1240
[email protected]