Full text of DOI:10.1111/0002-9092.00076

How Decoupled Is U.S. Agricultural
Support for Major Crops?
C. Edwin Young and Paul C. Westcott
As we look towardthe future of WorldTrade
production were largely eliminated (Young
Organization (WTO) negotiations, it is useful andWestcott). Proudction flexibility con-
to look at the direct and indirect impacts of
U.S. farm programs on agricultural produc-
tion andtrade. U.S. agricultural support, as
reportedto the WTO, exceeded$58 billion
in 1997. Most of the support was reported
in the green box as “minimally trade distort-
ing,” while about $6.2 billion was reportedin
the amber box as potentially “trade distort-
ing.”1 However, amber box program expendi-
tures greatly expandedin 1998 and1999 due
in part to generally lower commodity prices.
With increasedfarm program expenditures,
the market impacts of major programs need
to be evaluatedrelative to potential produc-
tion and trade distortions as background for
WTO negotiations.
This paper examines the links from four
U.S. agricultural programs to agricultural pro-
duction and trade. Programs discussed are
production flexibility contracts, crop insur-
ance, marketing loans, andidsaster assis-
tance. To the extent that programs expand
domestic production, the impacts can be par-
tially transmittedto worldmarkets through
increasedexports andlower prices.
tract payments are not relatedto current
production and prices. Nonetheless, these
payments raise farmers’ wealth, which may
subsequently impact agricultural investment
decisions. The 1996 Act retained the non-
recourse loan program with marketing loan
provisions. When expectedmarket prices are
below marketing assistance loan rates, the
commodity loan program has the potential
to influence production decisions by guaran-
teeing a per-unit revenue floor to produc-
ers at the commodity’s loan rate. The 1994
Federal Crop Insurance Reform Act pro-
vided a major overhaul of the U.S. crop insur-
ance program by increasing subsidies and
by introducing low-cost catastrophic cover-
age. Crop andrevenue insurance are deliv-
eredthrough private insurance companies
with the USDA’s Risk Management Agency
subsidizing insurance premiums and adminis-
trative costs. These subsidies have increased
the use of crop insurance programs in the
UnitedStates andtherefore may have added
to annual croplandplantings by reudcing
financial risks associatedwith crop produc-
tion. Agricultural appropriations acts for fis-
cal years 1999 and2000 includedemergency
andmarket loss assistance to U.S. agriculture,
potentially creating long run market distor-
tions if perceivedby farmers as signaling a
continuation of such programs.
With the elimination of deficiency pay-
ments and the introduction of produc-
tion flexibility contract payments in the
1996 Farm Act, commodity-specific links
between income support payments and
C. Edwin Young and Paul C.Westcott are agricultural economists,
Economic Research Service, UnitedStates Department of Agri-
culture.
The views expressedin this article are those of the authors
anddo not necessarily reflect the views of the U.S. Department
of Agriculture.
How Programs Affect
Production and Trade
This analysis focuses on incentives created
by agricultural programs to alter produc-
tion decisions. Subsequent impacts on trade
from these programs derive from market
adjustments to production impacts. The 1994
Uruguay RoundAgreement on Agriculture
(URAA) placedlimits on odmestic agri-
cultural programs that require individual
1 Green box expenditures totaled $51.2 billion, with the major-
ity of the expenditures for domestic food aid ($36 billion), gen-
eral services ($6.8 billion), anddecoupledincome support ($6.3
billion). Amber box expenditures totaled $6.2 billion, of which
$5.8 billion was for dairy, sugar, and peanut market price sup-
port. Crop insurance is classifiedas a commodity nonspecific
amber box expenditure, but commodity nonspecific expenditures
are reportedas zero since, in total, they do not meet the demi-
nimus criteria. Expenditures of $0.8 billion were excluded from
the AMS basedon the deminimus criteria.
Amer. J. Agr. Econ. 82 (August 2000): 762–767
Copyright 2000 American Agricultural Economics Association

Young and Westcott
How Decoupled is U.S. Agricultural Support
763
countries to take into account the effects of
their domestic policies on the international
marketplace. The WTO is not concernedwith
the rationale or justification for different pro-
grams, but with the impact of the programs
on trade.
Production Flexibility Contract Payments
The 1996 U.S. Farm Act fundamentally
changedagriculture income support pro-
grams by replacing the target price/deficiency
payment program with a new program of
production flexibility contract (PFC) pay-
ments that are generally not relatedto cur-
rent farm-level production or market prices.
Total outlays for the payments were capped
at slightly over $36 billion for seven years,
1996–2002. To be eligible to receive payments,
farmers enteredinto proudction flexibility
contracts, requiring compliance with conser-
vation, wetland, and planting flexibility pro-
visions, as well as keeping the landin agri-
cultural uses (which includes idling). Land
eligible for PFCs included acreage enrolled
in annual farm programs for any year from
1991 through 1995. Payments under these
contracts are basedon enrolledacreage and
generally are not relatedto current plantings.
In its WTO notifications, the UnitedStates
reports PFC payments as decoupled green
box payments. Whether or not these pay-
ments are totally decoupled has been ques-
tioned(Tielu andRoberts, Hennessy). These
authors argue that PFC payments are at least
partially coupledsince they increase farm
operator wealth, which has several potential
effects on production. First, lenders are more
willing to make loans to farmers with higher
guaranteedincomes because of a lower risk
of default. This increase in loan availabil-
ity may facilitate additional agricultural pro-
duction. Second, with increased income from
PFC payments, farmers can more easily
invest in their farm operation. For example,
increased income may facilitate additional
agricultural investments by farmers who are
constrainedby debt or limitedliquidity. The
resulting increasedinvestment in farm oper-
ations contributes to higher agricultural pro-
duction in the long run. Another potential
effect of PFC payments on production is that
a guaranteedincome stream may make farm-
ers more willing to undertake riskier crops or
strategies which have the possibility of higher
returns. An increase in wealth resulting from
PFC payments can change farmers’ views of
the penalties associatedwith risk levels of
wealth changes. Farmers’ responses to risk,
therefore, might be very different at higher
levels of wealth. Such a change in risk atti-
tude can affect the mix of crops produced.
Initially, a decoupled payment affects farm-
ers in the same way as a lump sum payment.
Some farm programs mostly influence
aggregate landuse, with less effect on the
mix of crops planted. Commodity nonspe-
cific transfers, for example, can increase
the overall level of agricultural production
by increasing the wealth of farmers and
thereby increasing agricultural investments.
This increase in wealth shifts out the pro-
duction possibilities frontier for all produc-
tion, raising use of landandother inputs.
Greater wealth does not affect the rela-
tive marginal returns from producing alterna-
tive crops, allowing market signals in general
to allocate the additional acreage. Penalties
associatedwith risk may be perceiveddiffer-
ently by people who have different levels of
wealth and, therefore, farmers’ responses to
risk may vary with their wealth levels.
Other programs are linkedmore closely
to production of specific crops and thus dis-
tort the mix of crops planted in addition to
total landuse. Direct coupling occurs when
program benefits are linkedto the produc-
tion of specific crops, so that those bene-
fits increase the expectedmarginal returns
to the commodity. Production decisions are
basedon the level of government payments
in addition to expected returns from the mar-
ketplace. Cross-commodity effects can also
result, because changes in expectedreturns
for one commodity affect relative net returns
as well, thereby influencing decisions to pro-
duce other crops. For farmers with land con-
straints, a coupledsubsidy wouldlikely alter
the mix of crops planted, switching toward
the subsidized (or more highly subsidized)
crops. However, since some farmers wouldbe
able to expandlanduse, aggregate acreage
couldincrease with the mix of crops shifting
towardthose with higher relative subsidies.
In addition to the effects of current pay-
ments, agricultural production can also be
affectedby programs that influence expecta-
tions. Programs that reduce risk, for example,
modify the lower end of the distribution of
expectedreturns andcan leadto production
distortions (Hennessy). Expectations about
the nature of future programs may also affect
current production decisions. For example, if
farmers expect future payments to be based
on current plantings, they may keep current
plantings of those crops high.

764
August 2000
Amer. J. Agr. Econ.
Profits rise, but since per-unit production
costs do not change, output is not affected.
The potential for PFC payments to distort
the production decision depends largely on
the strength of the wealth effects. Chavas
andHolt edvelopedan acreage response
model for 1954 to 1985 that includes a
wealth measure, proprietor equity in agricul-
ture multipliedby the percentage of farm
acreage plantedto the crops. They derived
wealth effect elasticities of 0.087 for U.S.
corn acreage and0.270 for U.S. soybean
acreage. The corn wealth elasticity (0.087)
andcorresponidng acreage impact may be
too low in the current policy environment,
since corn supply response was constrained
by farm programs during part of the estima-
tion periodusedby Chavas andHolt. The
soybean wealth elasticity (0.270) wouldbe
less affectedby past programs andtherefore
may be representative of wealth effects more
generally under current law.
Assuming wealth effect elasticities for each
of the seven U.S. crops coveredby PFCs fall
within the range provided by these corn and
soybean estimates andassuming that farmers’
wealth increases by the full value of PFC pay-
ments implies a possible aggregate acreage
impact of 180,000 to 570,000 acres annually.
The increasedacreage will be allocatedacross
crops by market returns. Increasedinvest-
ment in other, nonlandfactors of proudc-
tion couldfurther raise production. However,
lower prices that result from the increased
production wouldleadto some moderation
of production impacts and potential trade dis-
tortions.
Act provided low-cost catastrophic coverage
andinstitutedprocedures to restrict enact-
ment of disaster assistance. Crop insurance
is provided through private insurance com-
panies with the USDA’s Risk Management
Agency providing direction to the insurance
companies, subsidizing premiums, partially
reimbursing the costs of selling andservic-
ing policies, andunderwriting risk protection.
The UnitedStates classifies crop insurance
as a commodity nonspecific amber box pro-
gram in its WTO notification, since the pro-
gram does not meet all of the WTO green
box criteria.
Insurance changes the distribution of
expectedrevenues by reducing the financial
risk associatedwith crop proudction vari-
ability. Government crop insurance subsidies
are likely to alter producer behavior for a
number of reasons (Schnepf andHeifner).
Subsidies to insurance premiums represent a
positive expectedbenefit andare likely to
encourage greater participation in insurance
programs. The increase in expectedreturns
per acre resulting from subsidized insurance
provides an incentive to marginally expand
area in crop production.
Second, subsidies are calculated as a per-
centage of the total premium, which varies
across crops andfarms to reflect idfferent
risks of loss associatedwith each crop and
each insurable acre. As a result, premium
subsidies are higher for production of riskier
crops and for production on riskier land.
This subsidy structure favors production on
acreage with higher yieldvariability andthus
may encourage production on land that might
not otherwise occur.
PFC payments may also distort crop
production decisions by requiring land to
“remain in agricultural uses.” While this
requirement permits cropland to be idled, the
PFC payments may be sufficient incentive
to prevent some landfrom being converted
to nonagricultural uses. Once the decision
not to convert is made, the farmer then may
decide to produce on that land if expected
returns exceedthe costs of production. Even
if the land is idled, it is available to return
to agricultural production if economic condi-
tions warrant.
Third, Federal subsidies to private insur-
ance companies for administration and
delivery expenses likely increase insurance
availability in remote areas.
Young et al. recently conducted a prelim-
inary assessment of the extent of market
distortions attributable to Federal crop insur-
ance subsidies. This study proposes a method-
ology in which county-level, crop-specific
subsidies on premiums, indemnities, and lia-
bilities are convertedto regional commodity-
specific price wedges andincorporatedinto
net returns measures usedfor proudction
decisions. In this framework, the availabil-
ity of subsidized crop insurance affects farm-
ers’ current crop production decisions by
Crop and Revenue Insurance
Crop insurance plays a prominent role as creating a direct incentive to expand pro-
part of the farm safety net in U.S. agricul- duction. Acreage and production impacts
tural policy. The 1994 Crop Insurance Reform of crop insurance subsidies were analyzed

Young and Westcott
How Decoupled is U.S. Agricultural Support
765
on a crop specific andregional basis using
distortions for U.S. exports occur for cot-
a multi-year, simulation model (POLYSYS- ton. Wheat, corn, andbarley exports increase
ERS) that accounts for intra- andinter-
regional acreage shifts andcross-commodity
moderately, while rice exports decline. Thus,
with the exception of cotton, the current crop
price effects. Their results represent average insurance program does not appear to distort
annual long-run effects with crop insurance U.S. exports significantly.
fully operational, allowing for market impacts
to stabilize.
Results suggest that crop insurance
subsidies generate small shifts in aggregate
plantings. For eight major fieldcrops, annual
federal crop insurance subsidies averaged
$1.4 billion during 1995–98. Assuming a con-
tinuation of this subsidy level, aggregate
acreage in the long run was estimatedto
increase annually by approximately 600,000
acres (0.2%).2 Interestingly, this estimated
crop insurance impact exceeds the PFC
impact estimatedin the previous section
even with a lower subsidy level, because crop
insurance program benefits are more directly
coupled to the production decision.
The insurance subsidies tend to move
acreage more towardriskier crops and
regions. Production is increased for six of
the eight major crops. Nationally, wheat and
cotton acreage show the largest percentage
gains due to crop insurance subsidies. Since
subsidies as a percentage of crop value for
rice andsoybeans are lower relative to those
for other crops, rice andsoybean acreage
decline slightly. Stronger effects emerge at
the regional level as plantedacreage shifts
away from the Southeast andFar West and
towardthe Plains States. The largest price
impacts occur for cotton (down 2%) and
wheat (down 1%). Price impacts for the other
commodities are 0.3% or less. An additional
important result is that cross-price andmulti-
year, price-feedback effects in the model sim-
ulations tendto dampen the initial own-price,
subsidy effect, giving acreage shifts that are
smaller than estimates which ignore those
effects.
Marketing Loans
Commodity loan programs in the United
States are one of the primary domestic
support programs andhave been in exis-
tence since the 1930s. Over the past fifteen
years, loan programs for major fieldcrops
have movedfrom price support programs
to marketing loan programs which provide
income support but no longer support mar-
ket prices. While costs of marketing loan
programs through 1997 were generally quite
small, lower commodity prices in the last
few years have resultedin significant pro-
gram costs. Total marketing loan benefits rose
from less than $200 million for 1997 crops
to more than $3.8 billion for 1998 andcould
exceed$7 billion for 1999 crops. For WTO
purposes, marketing loan costs are reported
as amber box.
Producers can benefit from the market-
ing loan program in two different ways. First
is through the loan program. Farmers place
their crop under the commodity loan pro-
gram by pledging and storing some of their
production as collateral for the loan, receiv-
ing a per-unit loan rate for the crop. Loans
may be repaidat the loan repayment rate,
which is basedon local, postedcounty prices
for wheat, feedgrains, andoilseeds andthe
prevailing worldmarket price for rice and
uplandcotton. When the loan repayment rate
is below the loan rate, therefore, marketing
loans allow repayment of the commodity loan
at less than the loan rate, with the difference
representing a cost to the government anda
program benefit to producers.
Consequently, crop insurance subsidies are
estimatedto have a small impact on U.S.
exports andtrade, reflecting market impacts
of production changes. The largest relative
Alternatively, farmers may choose to
receive marketing loan benefits through
direct loan deficiency payments (LDPs)
when loan repayment rates are lower than
commodity loan rates. In this case the farmer
forgoes the possibility of placing the crop
under loan. The LDP rate is the amount
by which the loan rate exceeds the posted
county price or the prevailing worldmarket
price and, thus, is equivalent to the marketing
loan gain that couldalternatively be obtained
for crops under loan.
2 Keeton, Skees, andLong argue that since many of the produc-
tion effects of crop insurance come from farm-level incentives,
aggregation of crop insurance impacts into a national model-
ing framework may underestimate production distortions. How-
ever, as one reviewer of this article points out, the Young et al.
methodology treats the insurance subsidy as a certainty equiv-
alent. Many farmers are reluctant to purchase subsidized crop
insurance because of the low probability of a payout, suggesting
that a certainty equivalent framework may overestimate produc-
tion effects.

766
August 2000
Amer. J. Agr. Econ.
Assuming that the sales price for the crop within a crop year and the reduction of down-
is equal to the postedcounty price, the mar-
keting loan program provides producers with
an effective per-unit revenue floor at the loan
rate. However, the marketing loan program
does not establish a floor for market prices.
A recent study (Westcott and Price) ana-
lyzedhow the marketing loan program can
lead to market distortions. The study uses
the USDA’s 1999 baseline andsimulations
of an econometric model for the U.S. agri-
cultural sector (FAPSIM), focusing on the
marketing loan program for soybeans. From
a scenario with soybean price expectations
40 to 50¢ below the soybean loan rate in
1999 and2000, soybean marketing loans were
estimatedto ad1.1 to 1.2 million acres
to soybean plantings in those years. Higher
net returns for soybeans drew some of this
increase in soybean plantings from competing
crops (cross commodity effects), particularly
corn, sorghum, anduplandcotton. However,
total plantedacreage increased100,000 to
200,000 acres. Importantly, acreage distor-
tions were largely confinedto those years
when prices were below the loan rate for
soybeans, years when marketing loan bene-
fits augmentedmarket returns anddistorted
production incentives. Only small effects on
plantings in subsequent years occurredwhen
prices rose above loan rates andmarketing
loan benefits were no longer present.
side income risk because the program trun-
cates the distribution of expected revenues
at the loan rate. Further, the magnitudes
of the estimatedimpacts were specific to
the situation that was analyzed, with results
dependent on the size of the marketing loan
benefits in the scenario. In higher price situa-
tions, impacts of marketing loans on markets
wouldbe smaller. Conversely, in lower price
situations, impacts wouldincrease.
Disaster Assistance
Adhoc disaster programs have hada promi-
nent role in U.S. agriculture. Crop insurance
reform legislation in 1994 included language
intendedto eliminate adhoc disaster assis-
tance, in part because such payments were
viewedas displacing some use of insurance
programs. However, more recently, emer-
gency spending legislation enacted in 1998
and1999 includeddisaster assistance for crop
losses as well as direct “market loss assis-
tance” payments to the sector.
Most disaster payments are typically made
ex post, occurring after production decisions
have been made. They thus can be argued to
not distort production and, as such, are per-
mittedunder the WTO to be reportedin the
green box, subject to specifiedcriteria. On the
other hand, if producers have expectations
of future assistance basedon past govern-
ment actions, then production decisions may
be influenced. In discussing financial aid to
the farm sector, Barry commentedthat “it is
hardnot to plan on government assistance
when it comes so easily.”
Trade distorting effects of soybean market-
ing loans in the simulations result from the
effects on plantedacreage. U.S. exports for
soybeans, soybean meal, andsoybean oil rose
by 1 to 2%. Cotton exports declined by 1 to
2%, while corn andsorghum exports declined
by less than 1%. As with the acreage effects,
impacts on U.S. exports were primarily in the
two years in the scenarios when marketing
loan benefits existed, with limited effects in
subsequent years.3
If disaster assistance is expected with some
probability when prices or production fall to
low levels, such expectations modify the bot-
tom of the revenue distribution. The result-
As notedin the stuyd, the full effects
of marketing loans are somewhat under- ing increase in expectedproudcer returns
representedin these results because of mod-
may leadto higher production than would
eling simplifications assumed. Other impacts otherwise occur. In so doing, these programs
would reflect additional marketing loan ben-
efits due to the seasonal movement of prices
encourage producers to keep riskier land in
production.
The more that disaster aidis viewedas
being effectively coupledto specific pro-
duction activities, the greater the degree
that expectedfuture benefits can influence
production choices. Disaster assistance that
addresses crop-specific production problems,
for example, can be viewedas similar to cou-
pledcrop insurance andis likely to leadto
3 This result differs from effects of a price-supporting loan pro-
gram as existed prior to the marketing loans being introduced.
For such a program, stock accumulation by the government
through loan defaults in lower-priced years leads to release of
government stocks at a later time, thus extending market impacts,
including exports effects, over a longer time period. However,
multi-year, cumulative impacts under a price-supporting loan
program are largely offsetting.

Young and Westcott
How Decoupled is U.S. Agricultural Support
767
expanded production of those crops. In con- past payments, these expectations can influ-
trast, less specific andless coupleddisaster
assistance payments wouldimpact aggregate
production more generally.
ence production.
In contrast, production flexibility contract
payments seem to be the least coupledof the
programs discussed. These payments create
a small incentive to increase aggregate pro-
duction, with the mix of crops planted based
on market signals. Among the four programs,
PFCs may have the smallest effect on mar-
kets per dollar because they are less directly
coupled to the production decision.
Implications
Each of the four U.S. domestic agricul-
tural support programs discussed in this arti-
cle increases U.S. production somewhat by
affecting planting decisions, in the aggregate
and/or for specific crops. As such, each pro-
gram has some market distorting effects on
U.S. exports and global trade. Production
distortions of these programs may overlap
somewhat, reflecting the potential for some
substitution between the programs, such as
expectations of disaster assistance displacing
use of crop insurance. Additionally, increased
production due to these programs will tend to
reduce prices, which with the planting flexibil-
ity provided by the 1996 Farm Act will result
in partly offsetting reductions in production
as producers respond to the lower prices.
Thus, while exports remain marginally higher
as a result of these programs, net impacts
appear to be small.
References
Barry, P. “Risk Management andSafety Nets for
Farmers.” Choices (ThirdQuarter 1999):1.
Chavas, J., andM.T. Holt. “Acreage Decisions
Under Risk:The Case of Corn and Soybeans.”
Amer. J. Agr. Econ. 72(August 1990):529–38.
Hennessy, D.A. “The Production Effects of Agri-
cultural Income Support Policies Under
Uncertainty.” Amer. J. Agr. Econ. 80(Febru-
ary 1998):46–57.
Keeton, K., J. Skees, andJ. Long. “The Poten-
tial Influence of Risk Management Programs
on Cropping Decisions at the Extensive Mar-
gin.” Selectedpaper, 1999 AAEA meetings,
Nashville TN, August 1999.
Crop insurance andmarketing loans create
direct incentives to expand production of
specific commodities by increasing expected
returns per unit of production. Crop insur-
ance changes the distribution of expected
income at low yields, with subsidies that
encourage production of riskier crops and
production in riskier regions. Marketing loans
truncate the distribution of expected per-
unit revenues, with program benefits creating
an underlying distortion to produce specific
crops when prices are near or below loan
rates.
Schnepf,R.,andR.Heifner.“Crop&RevenueInsur-
ance: Bargain Rates but Still a HardSell.” Agr.
OutlookAGO-263(August 1999):15–18.
Tielu, A., andI. Roberts. “Farm Income Support”.
ABARE Current Issues 98.4, August 1998.
Westcott, P.C., andJ.M. Price. “Impacts of the U.S.
Marketing Loan Program for Soybeans.” Oil
Crops Situation and Outlook Yearbook OCS-
1999(October 1999):15–21.
Young, E., R. Schnepf, J. Skees, andW. Lin.
“Production and Price Impacts of Agricul-
tural Crop Insurance: Preliminary Modeling
Results.” Proceedings of 1999 Federal Fore-
casters Conference, National Center for Edu-
cation Statistics, U.S. Department of Educa-
tion (1999):119–128.
If adhoc disaster assistance is not expected
by the recipients at planting time andoccurs
after production decisions are made, disas-
ter assistance may have little or no impact Young, C.E., andP.C. Westcott. The 1996 U.S. Farm
on current production. However, if produc-
ers of specific crops or in specific regions
expect periodic disaster assistance based on
Act Increases Market Orientation. Economic
Research Service, USDA, AIB 726, August
1996.