Food Versus Fuel: How Biofuels Make Food More Costly and Gasoline Cheaper
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
V. 12 no. 1 • Sep/Oct 2008
Food Versus Fuel:
How Biofuels Make Food More Costly and Gasoline Cheaper
Steven Sexton, Deepak Rajagopal, David Zilberman, and Gal Hochman
B
iofuel production in the United The rapid increases in food prices
States has been ridiculed in recent that began in 2007 have resulted in
This paper describes forces behind months, following the release of deadly food riots, increased robberies
rising food prices and presents a
reports that suggest ethanol and biod- of food-aid caravans, export restric-
model to characterize the magnitude of
iesel not only increase greenhouse gas tions in grain-producing countries,
biofuel impacts on food and gasoline
emissions relative to fossil fuels, but and pleas for supplemental funding
prices. The results of this model are
compared to other estimates. We also raise food prices and lower food for food-aid programs. A 140 percent
argue that a renewed commitment production. The impact of biofuels on increase in food prices from 2002 to
to agricultural productivity growth is food markets came under particular 2008 led humanitarian organizations to
needed to overcome current food and scrutiny this year as the world entered predict human suffering and starvation
fuel challenges. its first food crisis in more than 30 years. not seen in more than a generation. The
It is certainly true that biofuels have Food and Agriculture Organization of
increased the price of agricultural com- the United Nations (FAO) reports food
modities, but the magnitude of biofuel prices increased 53 percent in just one
impacts on food markets is unsettled. year from March 2007 to March 2008.
High food prices have been accompanied Vegetable oils rose 97 percent, followed
Also in this issue by record high oil prices, and, while by grains, which rose 87 percent. Food
biofuels have been blamed for exacer- price increases in the past year constitute
bating the former, they have not been the most rapid increase over a 12-month
credited with mitigating the latter. But period in more than 30 years. The 55
How Can We Avoid Another just as surely as they have contributed to percent increase in food prices in the
Food Crisis in Niger? raising food prices, biofuels have helped past 12 months is exceeded only by
Jenny C. Aker...................................7 reduce oil prices relative to prices that their doubling from 1973 to 1974. The
would prevail absent biofuel produc- poor will suffer most from high prices
tion. This article presents a model to because they devote large shares of their
Faculty Profile: demonstrate the effects of biofuels on household budgets to food purchases.
Pierre Mérel...................................11 corn, soybean, and gasoline prices, and Even in countries where the rural
to derive the distribution of benefits poor benefit on average from higher
from U.S. biofuel production. We con- prices for their agricultural output,
In the next issue... clude that biofuels have a nontrivial the poorest of the poor will suffer.
impact on food security. We argue that
Special Issue: Causes and underinvestment in research and over- Forces Driving Food Price Inflation
Consequences of the Food regulation of agricultural biotechnology Biofuels are not solely responsible, nor
Price Crisis led to a decline in productivity growth necessarily principally responsible, for
that is also responsible for higher prices changes in the food security climate. A
and must be reversed if global food variety of supply-side and demand-side
and energy security are to improve. forces are at play. Among these is thefor commodities planted to genetically
Figure 1. Fuel Costs as Share of Total Operating Costs
modified (GM) seed and commodities
40 not produced from agricultural bio-
35
technology. Figure 2 shows persistent
gains in productivity for corn, soybeans,
Percent of Operating Costs
30 and cotton—crops for which GM tech-
25
nology has been employed. Yields for
staple crops like wheat and sorghum,
20 however, are shown to have stagnated
15
since 1990, as gains from the Green
Revolution are exhausted. With slow
10 productivity growth, food prices will be
5
propelled higher by a rising world popu-
lation and a shift toward meat-intensive
0 diets induced by rising incomes.
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Biofuels also increase demand for
Cotton
agricultural production. Biofuels have
Corn Soybean X Wheat
been supported by governments despite
increase in oil prices in recent years. those that are energy intensive, like their inflationary effect on food prices
Oil prices have nearly doubled since fertilizers (70-90 percent of fertilizer because they are perceived to reduce
2006 from an average $66/barrel in costs are embodied in energy). As costs greenhouse gas emissions relative to
2006 to a forecasted average $116/barrel of production increase, the supply fossil fuels, improve energy indepen-
in 2008. Prices rose 60 percent in just curve for agricultural commodities dence, and spur rural development. As
the past year. These oil price increases shifts up, raising market prices. a consequence of favorable government
raise the cost of agricultural production On the demand side, population intervention and energy prices that
and transportation. Direct costs of fuel growth and income growth generate make biofuels cost-competitive with
purchases have averaged about seven sustained upward pressure on prices. gasoline, global biofuel production has
percent of farm operating costs since By 2050, the world population is grown markedly in recent years, reach-
1992, but have begun to rise as a share expected to grow by half. Historically, ing 6,500 billion gallons in 2007—four
of costs in recent years. Figure 1 depicts such a rate of population growth was times the production of 2000.
the increasing fuel share of operating exceeded by agricultural productiv- Existing biofuels are produced from
costs for five field crops (cotton, ity growth. From 1950 to 2000, for agricultural crops traditionally used for
corn, soybeans, wheat, and rice) from instance, per capita food production food or feed. Ethanol, which dominates
2000 to 2009 (data for 2007–2009 are increased even as the world popula- biofuel production in the United States
forecasts). High fuel prices also raise tion doubled. This improvement in and Brazil (and India to some extent),
the price of farm inputs, particularly per capita food production occurred is presently produced from corn and
despite a shrinking agricultural land sugarcane. Biodiesel, produced mainly
Table 1. Elasticity Assumptions base because adoption of farming tech- in the E.U., is made from soybeans and
for Simulations nologies like mechanization, irrigation, rapeseed. Biofuel, therefore, increases
Scenarios and chemicals promoted significant demand for these staple crops so long
Own price High Mid Low productivity gains. There is, however, as crop prices are not too high to make
supply elasticities little capacity for additional gains from biofuels unprofitable. In 2007, 20 per-
these technologies, particularly in the cent of the U.S. corn harvest was used in
Corn 0.5 0.4 0.3
developed world. New sources of pro- ethanol production and farmers planted
Soy 0.5 0.4 0.3
ductivity growth are needed to reverse the largest crop in 63 years—93 mil-
Gas 0.3 0.4 0.5
a trend of stagnating yield gains. Agri- lion acres—nearly a 20 percent increase
Own price
cultural biotechnology is one source from 2006. A second generation of bio-
demand elasticities
of yield improvements, though it is fuels will make use of energy-specific
Corn -0.5 -0.4 -0.3 underutilized and overregulated. Still, cellulosic crops that can be grown on
Soy -0.5 -0.4 -0.3 the effect of biotechnology cannot be marginal land. Cellulosic biofuels would
Gas -0.3 -0.4 -0.5 missed when comparing yield trends reduce the diversion of food crops for
2 Giannini Foundation of Agricultural Economics • University of CaliforniaFigure 2. Production, Area Harvested, Inventories, and Yield for Staple Crops, 1990–2006
35 12 100 3.5
Corn Soybean
90
30 10 3.0
80
25
Yield (MT/Ha)
70 2.5
Yield (MT/Ha)
8
20 60
2.0
6 50
15 1.5
40
4
Area (1000s Ha.), Stocks (1000s MT)
Area (1000s Ha.), Stocks (1000s MT)
10 30 1.0
2 20
5 0.5
10
0 0 0 0
1990
1992
1994
1996
1998
2000
2002
2004
2006
1990
1992
1994
1996
1998
2000
2002
2004
2006
30 1.2 80 3.5
Wheat
Cotton
70
25 1.0 3.0
60
2.5
Yield (MT/Ha)
Yield (MT/Ha)
20 0.8 50
2.0
15 0.6 40
1.5
30
10 0.4
20 1.0
5 0.2 0.5
10
0 0 0 0
1990
1992
1994
1996
1998
2000
2002
2004
2006
1990
1992
1994
1996
1998
2000
2002
2004
2006
25 5
Sorghum
Area (1000s Ha.), Stocks (1000s MT)
20 4
Yield (MT/Ha) Production
15 3
Yield
10 2
Area
5 1
Ending Stocks
0 0
1990
1992
1994
1996
1998
2000
2002
2004
2006
corn-ethanol and soybean-biodiesel, but feedstocks displace food crops and low, food prices are susceptible to any
they are yet to be commercially viable. recruit idled land back into production. deviation in production from long-
Biofuels also raise the costs of agri- In doing so, it reduces the supplies of term trends, which may explain why
cultural commodities not directly food and environmental preservation. wheat and rice prices increased more
used in energy production. By raising The price boom since 2006 is also than corn prices in the past year.
demand for inputs in farm production, the result of tight markets with grain Amid such tight markets and consid-
from tractors and fertilizer to water inventories at historic lows. A number erable uncertainty, rice-producing
and land, biofuels raise production of negative supply shocks to wheat and countries imposed export controls to
costs throughout agriculture. Perhaps rice in recent years led to production protect domestic prices. Had produc-
nowhere is the pressure exerted by bio- shortfalls and caused a drawdown in tion not exceeded expectations in 2008,
fuels felt more acutely than in livestock crop inventories. Inventories act as a the constrained international markets
production, a sector that faces rising buffer to random shocks to markets, could have collapsed as more and more
costs for its primary input—feed. As such as floods, droughts, or unusual countries sought to insulate themselves
biofuel raises the rental rate of land, pest pressure. When inventories are
Giannini Foundation of Agricultural Economics • University of California 3Figure 3. U.S. Corn Market
(PB > PA ) and reduces the quantity of
corn for other uses (QFB < QA ). We
assume the U.S. market clears and deter-
$
mines the price for traded corn; the U.S.
S supplies 70 percent of traded corn. The
soy, ethanol, and gasoline markets are
modeled as described in Rajagopal et al.
(See references on page 6.)
P0 We provide results for three sce-
narios, which we call high, mid, and
PB low, depending on the change in net
PA consumer (surplus) due to biofuels.
The high scenario is characterized by an
elastic (price responsive) food market
and inelastic (unresponsive to price)
DB gasoline market. It is in this scenario
that biofuel supply has the largest
positive impact on gasoline consumers
DA
and smallest negative impact on food
QFB QA QTB Q
consumers. In the opposite scenario,
from higher and higher prices with Specifically, we assume demand for characterized by a highly inelastic food
greater export restrictions. corn in the United States is composed of market and a highly elastic gasoline
domestic ethanol demand, domestic market, food consumers suffer the most
Quantifying the Effects demand for other uses (such as food and and gasoline consumers benefit the
To estimate the impact of U.S. biofuel feed), and world excess demand. Figure least. The mid scenario lies in between.
production on food and fuel markets 3 depicts the U.S. market for corn where The elasticities we use in the three
and to determine the magnitude of demand is given by DA with no demand scenarios are reported in Table 1.
welfare effects, we developed a global for ethanol and DB with demand for Research suggests our “high” sce-
multi-market partial equilibrium biofuel. DA represents domestic nario may not be too optimistic and
model constituted by markets demand for corn for uses other than bio- may, in fact, be conservative: elastici-
for corn, soybeans, biofuel and fuel and world excess demand. DB also ties for gasoline, soy, and corn tend to
gasoline. We consider two regions, includes demand for ethanol, which is be less than 0.25 in the short run.
the United States and the Rest of the assumed to be zero above P0. We Although we include the impact of
World (ROW), but assume that the assume linear supply and demand for biodiesel on the soy market, we do not
responsiveness to prices of quantity simplicity. Without biofuel demand, estimate the impact of biodiesel produc-
supplied and demanded (elasticities) price is PA and production is QA. The tion on diesel prices, which also serves
does not vary across regions. Using quantity of corn for food and feed is QA. to make our estimate of the fuel market
observed prices and quantities from With biofuel demand, the price is PB, impact of biofuels a conservative one.
2007—when there was demand production is QTB and production for
for biofuels—and assumptions on food and feed is QFB. The quantity of Results and Discussion
supply and demand elasticities, corn for ethanol is QTB– QFB. As can be Using data from 2007, in which 18.3
we constructed the counterfactual seen by comparing equilibria with and percent of U.S. corn production was
prices and quantities that would have without biofuel demand, biofuel used for ethanol, we find that ethanol
prevailed absent biofuel demand. demand increases the price of corn raised corn prices at least 18 percent
Table 2. Price Changes from U.S. Biofuel Production
and perhaps as much as 39 percent,
depending on elasticity assumptions.
Corn Price Changes Soybeans Price Changes Gasoline Price Changes These results are summarized in Table
Percent $/ Bushel Percent $/ Bushel Percent $/Gallon 2, along with dollar savings per bushel,
High -15% -0.72 -10% -1.00 2.4% 0.07 based on an average price of corn in
Mid -20% -0.92 -13% -1.34 1.8% 0.05 the United States of $4.72/bushel in
Low -28% -1.31 -20% -2.02 1.4% 0.04 2007. Under reasonable estimates, we
4 Giannini Foundation of Agricultural Economics • University of Californiafind that U.S. ethanol production in
Table 3. Welfare Effects of U.S. Biofuel Production (in billions of dollars)
2007 (4.4 billion gallons on an energy-
equivalent basis) reduced gasoline High Mid Low
prices at least 1.4 percent and as much Welfare Change
as 2.4 percent, or $0.04 to $0.07 per World consumers* -2.5713 -25.2875 -54.8611
gallon. These results are also reported U.S. consumers* -3.8153 -10.2098 -19.1565
in Table 2. The Energy Information U.S. consumers net of tax* -7.5303 -13.9248 -22.8715
Administration has estimated savings ROW consumers* 4.959 -11.3627 -31.9896
as high as $0.11 per gallon. World gas consumers 41.695 31.2709 25.0166
Based on this analysis, we find that U.S. gas consumers 9.6271 7.2203 5.7762
under the most optimistic assump- ROW gas consumers 32.0678 24.0506 19.2404
tions on biofuel impacts (our “high” World food consumers -40.5513 -52.8434 -76.1627
scenario), gasoline consumers around
U.S. food consumers -13.4424 -17.4301 -24.9326
the world benefited from lower gasoline
ROW food consumers -27.1089 -35.4133 -51.23
prices by $41.7 billion. Consumers of
World corn and soy producers 20.7519 27.016 38.8874
soybeans and corn, however, lost $40.6
U.S. corn and soy producers 8.4425 10.9703 15.7465
billion from higher food prices. On net,
consumer welfare declined by $2.6 bil- ROW corn and soy producers 12.3095 16.0457 23.1408
lion from U.S. ethanol production in Total U.S. Welfare Change 0.9122 -2.9545 -7.125
the best-case scenario (after deducting
Total World Welfare Change 18.1806 1.7285 -15.9737
the taxpayer cost of ethanol subsidies
from the net consumer benefits). In * Total consumer welfare change is the net effect of food and gas price effects. By region, the change
in welfare for food consumers is added to the change in welfare for gas consumers.
the worst-case scenario, world con-
sumers lost $54.8 billion in surplus. Council of Economic Advisors (CEA), in 2008. Interestingly, inventories for
These results are summarized in Figure the USDA, and the Farm Foundation. non-biofuel crops have fallen as much
4, while Table 3 summarizes welfare These other studies generate a range in as or more than the stocks of biofuel
gains to U.S., foreign, and all produc- the share of food price increases attrib- crops (corn and soybeans), as seen
ers and consumers under the three sets utable to biofuels of 23–61 percent. in Figure 2. This suggests that slow
of elasticity assumptions. The total A report by the World Bank identi- productivity growth is an important
welfare in the United States (net U.S. fied much larger impacts from biofuels factor in the decline of food security.
consumer benefit + net U.S. producer on food markets. It found that biofuels
benefit) improves by $0.9 billion with were responsible for three-fourths of a Biotechnology and the
biofuel production under our optimis- 140 percent increase in food prices from Food-Fuel Trade-off
tic (“high”) scenario. But it declines 2002 to 2008, or roughly a 50 percent Investment in agricultural research
under the other two scenarios. Like- increase in the past year. This estimate has declined in recent years, perhaps
wise, total global welfare is improved is considerably higher than an estimate the result of complacency during
under the “high” and “mid” scenarios by the CEA that biofuels raised food a period of stable food prices. A
by $1.7–18.2 billion, but falls under prices 1.5 percent from 2007 to 2008. lack of commitment to research
the “low” scenarios by $16 billion. The World Bank analysis included indi- and development can be blamed
Our analysis demonstrates that bio- rect effects and long-term trends, while for declining rates of yield growth.
fuels reduce the price of gasoline to others did not. This may account for Productivity growth since 1990
the benefit of gasoline consumers and the magnitude of the World Bank esti- has been half as fast as it was from
confirms other reports that biofuels hurt mate. The fall in stockpiles since 1990 1970–1990. It is expected to continue
food consumers. Ours is the only analy- can be seen in Figure 2 for five crops. declining over the next ten years,
sis to consider distributional concerns, Depletion of stocks may be particu- according to the U.S. Department of
which suggest a trade-off between fuel larly responsible for remarkable price Agriculture. From 1990–2007, the
for the rich and food for the poor. Our spikes since 2007. Biofuel production world population grew at a rate of 1.4
analysis suggests biofuels are responsi- increased most dramatically in 2006 percent per year. Yields in grains and
ble for between 25–60 percent of recent and 2007, and has not grown consider- oilseeds grew at only 1.1 percent per
corn price increases, which is consis- ably in 2008. Yet food prices have risen year. Until 2017, yields are expected
tent with reports from the President’s most quickly and become most volatile to grow at 0.8 percent, 0.3 percentage
Giannini Foundation of Agricultural Economics • University of California 5Figure 4. Net Benefits to Gasoline and Food Consumers from Ethanol Supply in 2006
considered is a food fund that could be
tapped into to buy food for the poor in
60
crisis situations. Another option may
40 be to tie government support for bio-
Net Benefit in Billions of 2006 USD
fuel to food market outcomes. When
20
food production falls too low, subsidies
0 and mandates should be scaled back to
High Mid Low protect against hunger. Policymakers
-20
may also find it necessary to reconsider
-40 existing policies in light of the current
food situation. They and researchers
-60 must recognize that the management
-80
All Gasoline Corn and Soy of agriculture is increasingly becom-
Consumers Consumers Consumers
ing a balancing act between energy,
-100 environment, and food objectives.
points less than forecasted population existing technologies, but also slowed
growth. These trends are ominous, development of next generation geneti-
given expectations for income growth cally modified seed that is expected to Steven Sexton is a Ph.D. student in the ARE
department at UC Berkeley. He can be reached
and expansion of biofuel production. introduce drought-tolerant plants and at ssexton@are.berkeley.edu. David Zilberman
Continued biofuel production and staple crops infused with additional is a professor in the same department. Deepak
ongoing growth in food demand could nutrients like beta-carotene. Without a Rajagopal is a Ph.D. candidate in the Energy
cause a persistence of food security chal- market for their innovations, however, Resources Group at UC Berkeley. Gal Hochman
is a visiting scholar at UC Berkeley from the
lenges. The food insecurity observed there is little incentive for firms to invest Technion - Israel Institute of Technology. He
today is not a matter of inevitability, in agricultural biotechnology R&D. can be reached at ghochman@are.berkeley.edu.
however. It can likely be overcome The authors wish to thank Colin Carter, Gordon
today and avoided in the future without Conclusion Rausser, and Aaron Smith for helpful comments.
This research is funded by the Energy Biosciences
abandonment of biofuels if the world World agriculture is facing great Institute and Farm Foundation.
regains its commitment to agricultural challenges as growing demand for food
productivity growth and harnesses the and fuel creates scarcity and induces
potential of agricultural biotechnology. hunger. Even when considering just
In the past 15 years, genetically modi- U.S. biofuel production, it is clear
fied (GM) crops have increased yields biofuels have significantly reduced
of cotton, rice, and corn 30–50 percent. gasoline prices, but at the expense
GM crops are infused with genes to of contributing to food shortages.
kill certain pests or provide immunity The challenge for agriculture to meet For more information, the authors
to common herbicides. They reduce growing food demand and growing recommend the following:
the share of crops that is damaged and, energy demand requires all the tools
Rajagopal, Deepak, Steven E. Sexton,
thereby, improve productivity. With the available to improve productivity, David Roland-Holst, and David
demand pressures facing agriculture, including agricultural biotechnology. It Zilberman. “Challenge of Biofuel:
biotechnology is an valuable mecha- also creates urgency for the development Filling the Tank without Empty-
nism for resolving an untenable food of commercially viable cellulosic biofuel ing the Stomach.” Environmental
and energy situation. GM crops lessen technologies that reduce demand for Research Letters, December 2007.
the land constraint and permit greater staple crops and make more efficient use Mitchell, Donald. “A Note On
production of food and energy crops of resources than current technologies. Rising Food Prices.” The World
Bank Policy Research Work-
on the existing agricultural land base. The development of second genera- ing Paper 4682. The World Bank:
Though GM crops have been adopted tion biofuels may take some time and Washington (DC). July 2008.
around the world at an astonishing productivity gains from biotechnol- Abott, Phillip C., Christopher Hurt,
rate, regulation, particularly in Europe, ogy may be gradual. Therefore, it will and Wallace E. Tyner. “What’s
has reduced the market for agricultural be necessary to develop mechanisms Driving Food Prices?” Farm Foun-
biotechnology in recent years. This to fight food shortages in the short- dation Issue Report. Farm Founda-
has not only slowed yield growth from run. One mechanism that should be tion: Oak Brook (IL). July 2008.
6 Giannini Foundation of Agricultural Economics • University of CaliforniaYou can also read
