Cotton Program: Costly and Complex Government Program Needs to Be
Reassessed (Chapter Report, 06/20/95, GAO/RCED-95-107).
Pursuant to a congressional request, GAO reviewed the Department of
Agriculture's (USDA) cotton program, focusing on the program's: (1) cost
and complexity; (2) distribution of payments; (3) effects on producers'
costs and returns; and (4) effectiveness in enhancing U.S. cotton
exports.
GAO found that: (1) over the past 60 years, the cotton program has
evolved into a costly, complex maze of domestic and international price
supports that benefit cotton producers at the cost of the government and
society; (2) from 1986 through 1993, the cotton program's costs totalled
$12 billion and averaged $1.5 billion a year; (3) in 1993, 295 producers
received payments of more than $250,000, which is the legislated amount
that a producer can receive under certain farm programs; (4) revenues
from domestic markets were more than sufficient to cover the producers'
short-run production costs, and most of the average producers' long-run
costs from 1986 through 1993; (5) USDA efforts to enhance cotton exports
have not been effective despite federal efforts to assist exporters and
producers through marketing loans and step 2 payments; (6) U.S. cotton
exports are not beneficial because production costs and government
payments, taken together, are consistently higher than the adjusted
world price; and (7) it may be more difficult in the long-run for the
United States to subsidize cotton exports while also supporting cotton
prices under the loan program.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: RCED-95-107
TITLE: Cotton Program: Costly and Complex Government Program Needs
to Be Reassessed
DATE: 06/20/95
SUBJECT: Agricultural programs
Exporting
Farm credit
Farm income stabilization programs
Subsidies
Foreign trade policies
Cost control
Price supports
Export regulation
Commodity marketing
IDENTIFIER: USDA Cotton Program
USDA Conservation Reserve Program
USDA Export Promotion Program
USDA Disaster Payments and Crop Insurance Program
USDA Pest Management Program
USDA Acreage Reduction Program
Dept. of the Interior Water Subsidies Program
USDA Market Promotion Program
USDA GSM-102 Program
USDA GSM-103 Program
USDA 50/92 Program
Canada
Mexico
North American Free Trade Agreement
NAFTA
**************************************************************************
* This file contains an ASCII representation of the text of a GAO *
* report. Delineations within the text indicating chapter titles, *
* headings, and bullets are preserved. Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are *
* identified by double and single lines. The numbers on the right end *
* of these lines indicate the position of each of the subsections in the *
* document outline. These numbers do NOT correspond with the page *
* numbers of the printed product. *
* *
* No attempt has been made to display graphic images, although figure *
* captions are reproduced. Tables are included, but may not resemble *
* those in the printed version. *
* *
* A printed copy of this report may be obtained from the GAO Document *
* Distribution Facility by calling (202) 512-6000, by faxing your *
* request to (301) 258-4066, or by writing to P.O. Box 6015, *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time. *
**************************************************************************
Cover
================================================================ COVER
Report to the Honorable
Richard K. Armey, House of
Representatives
June 1995
COTTON PROGRAM - COSTLY AND
COMPLEX GOVERNMENT PROGRAM NEEDS
TO BE REASSESSED
GAO/RCED-95-107
Cotton Program
Abbreviations
=============================================================== ABBREV
ACR - acreage conservation reserve
AMS - Agricultural Marketing Service
ARP - Acreage Reduction Program
ASCS - Agricultural Stabilization and Conservation Service
AWP - adjusted world price
CCA - coarse count adjustment
CCC - Commodity Credit Corporation
CIF - cost, insurance, and freight
CRP - Conservation Reserve Program
ERS - Economic Research Service
GATT - General Agreement on Tariffs and Trade
LDP - loan deficiency payment
MPA - maximum payment acres
NAFTA - North American Free Trade Agreement
NE - Northern European
NFA - normal flex acres
OFA - optional flex acres
PIP - producer incentive price
SCS - Soil Conservation Service
USDA - U.S. Department of Agriculture
USNE - United States Northern European
Letter
=============================================================== LETTER
B-260593
June 20, 1995
The Honorable Richard K. Armey
House of Representatives
Dear Mr. Armey:
This report responds to your request that we review the U.S.
Department of Agriculture's cotton program. The report describes the
program's cost and complexity, distribution of payments, effects on
producers' costs and returns, and effectiveness in enhancing U.S.
cotton exports. The report suggests alternatives that the Congress
may wish to consider to streamline the program and reduce its cost.
As arranged with your office, unless you publicly announce its
contents earlier, we plan no further distribution of this report
until 30 days after the date of this letter. At that time, we will
send copies to appropriate House and Senate committees and
subcommittees; interested Members of Congress; the Secretary of
Agriculture; and the Director, Office of Management and Budget.
Copies will also be made available to others upon request.
This work was performed under the direction of John W. Harman,
Director, Food and Agriculture Issues, who may be reached at (202)
512-5138 if you or your staff have any questions. Other major
contributors to this report are listed in appendix V.
Sincerely yours,
Keith O. Fultz
Assistant Comptroller General
EXECUTIVE SUMMARY
============================================================ Chapter 0
PURPOSE
---------------------------------------------------------- Chapter 0:1
The cotton program, a U.S. Department of Agriculture (USDA) program
to support cotton farmers and cotton exports, cost an average of $1.5
billion annually for crop years 1986 through 1993. Concerned about
the cost, Representative Richard K. Armey asked GAO to conduct a
comprehensive evaluation of the cotton program. Specifically, GAO
was asked to evaluate the program's cost and complexity, distribution
of payments, effects on producers' costs and returns, and
effectiveness in enhancing U.S. cotton exports.
BACKGROUND
---------------------------------------------------------- Chapter 0:2
Since the turn of the century, U.S. cotton producers have frequently
experienced excess production capacity, high stocks, and low product
prices. These problems were intensified during the Great Depression,
when U.S. farm families saw their income decline significantly as
farm commodity prices dropped by 50 percent. In response, the
Congress established a comprehensive program in 1933 aimed at
controlling the production of designated basic commodities, including
cotton, and supporting farm income and prices. This program was
intended to be temporary.
Since then, the cotton program has been modified many times. The
most recent reforms in 1985 and 1990 were designed to reduce
government costs and to enhance exports by making U.S. cotton more
competitively priced in world markets.
While the cotton program is voluntary, about 90 percent of all
acreage devoted to cotton is enrolled. Producers must agree to abide
by any USDA-imposed limits, under the acreage reduction program, on
the number of acres farmed for cotton. Under other programs that
deal with the cotton supply, producers may voluntarily idle
additional acreage. Early in the calendar year, participating
producers must register (sign up) how much of their cotton acreage
they intend to plant.
The program has two basic components to support producers' income.
The first component--known as the deficiency payment--guarantees
producers a given level of revenue by paying them the difference
between a set target price, established by the Congress, and either
the domestic market price or the USDA-determined loan rate, whichever
is higher. When producers sign up, they may receive a portion of
their projected deficiency payment to provide funds for planting.
The second component--the nonrecourse loan--provides government money
to producers, using their cotton as collateral, and allows them to
repay the loan or forfeit the cotton as full payment of the loan.
This loan provides funds to producers at harvest, when prices are
low, and producers may want to store cotton until prices rise. A
concept introduced in 1985--the marketing loan--changed the
nonrecourse loan repayment process. This loan permits producers to
repay their loan at the USDA-determined loan rate or at the
USDA-adjusted world price, whichever is lower. If the loan is repaid
at the adjusted world price, producers keep the difference and thus
receive a gain referred to as a marketing loan gain. This loan is
attractive to producers during periods of low prices. It allows
producers to wait for prices to rise but lessen their market risk
because the government pays producers' storage costs for up to 10
months.
Producers can decide not to put their cotton under government loan
and instead receive a so-called loan deficiency payment. This
payment, available when the adjusted world price is less than the
loan rate, is calculated to represent the difference between the loan
rate and the adjusted world price. During any crop year, producers
can place a portion of their cotton under nonrecourse loan and
receive a loan deficiency payment on the remainder.
The price of U.S. cotton from 1986 through 1993 was higher than the
world price for cotton. The world price for cotton is determined by
a pricing mechanism anchored in northern Europe. This higher
domestic price tends to make U.S. cotton noncompetitive on the world
market. U.S. cotton is priced higher because the cotton program's
import restrictions effectively prevent any cotton imports. To
address this problem, the cotton program provides for
subsidies--called step 2 payments--to exporters and domestic mills to
help keep domestically grown cotton competitively priced in world
markets. Competitive U.S. prices are important because about 40
percent of U.S. cotton is exported.
RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3
The cotton program has evolved over the past 60 years into a costly,
complex maze of domestic and international price supports that
benefit producers at great cost to the government and society. From
1986 through 1993, the cotton program's costs totaled $12 billion, an
average of $1.5 billion a year. Moreover, the program is very
complex, with dozens of key factors that interact and counteract to
determine price, acreage, and payments and to restrict imports.
The severe economic conditions and many of the motivations that led
to the cotton program in the 1930s no longer exist. Cotton farming
has become a concentrated business, with only 20 percent of the
producers growing most of the cotton. Consequently, most of the
program's payments also go to those producers. For most producers,
domestic prices together with government payments provide revenues
above the amount needed to cover their total production costs,
including a return on assets.
USDA's marketing loan and step 2 payments are intended to maintain
cotton exports by bringing the price of U.S. cotton in line with the
lower world price. These provisions have not succeeded despite
government payments of about $3 billion from 1986 through 1993: In
1988, 1991, and 1992, U.S. cotton exports and market share actually
declined despite heavy government expenditures. In addition, because
of production costs and government payments, exports of U.S. cotton
occur at a loss to the nation. In 1993, the adjusted world price was
56 cents per pound; however, from a national standpoint, U.S. cotton
cost 90 cents per pound--66 cents to produce and 24 cents in federal
payments. Despite this loss, the government continues to fund
programs that promote cotton exports. Furthermore, certain
objectives of the North American Free Trade Agreement (NAFTA) and the
General Agreement on Tariffs and Trade (GATT)--to encourage less
government support for agriculture and a more open world market--may
make it more costly in the long term to maintain price supports and
other aspects of the current U.S. cotton program.
PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4
PROGRAM IS COSTLY AND
COMPLEX
-------------------------------------------------------- Chapter 0:4.1
The cotton program's costs are incurred primarily through a variety
of government payments to producers. In crop year 1993, the
government paid producers $1 billion in deficiency payments, $238
million in marketing loan gains, and $305 million in loan deficiency
payments.
Reforms over the years have made the program more complex. Because
of this, a labyrinth of factors must interact for the program to
function. The interactions are frequently at cross-purposes to one
another, such as when the program supports the price the producer
gets, then subsidizes buyers to reduce their costs.
RELATIVELY FEW PRODUCERS
RECEIVE MOST BENEFITS, OFTEN
IN EXCESS OF PRODUCTION
COSTS
-------------------------------------------------------- Chapter 0:4.2
In 1993, 20 percent of cotton farmers produced an estimated 79
percent of the cotton under the program and received a like
percentage of program payments. Legislation limits the amount of
certain farm programs' payments a person can receive, generally to
$250,000. However, as part of the concentration of cotton
production, producers have in some cases organized their operations
into entities, such as joint ventures or partnerships composed of
multiple "persons," each eligible to receive payments up to the
applicable limit. In 1993, for example, 295 producers each received
more than $250,000--4 of these received more than $1 million--in
payments from USDA for their cotton operations.
Furthermore, revenues from domestic market prices were sufficient to
more than cover producers' short-run production costs and most of the
average producer's total (long-run) costs of production over the
8-year period 1986-93. When government payments were added,
producers' revenue averaged 17 percent higher than the total costs of
production. If the cotton program were not in place and the 1986-93
average costs of production and market prices were in effect, some
producers would not be able to cover their production costs and would
have to either reduce costs or go out of business.
While the cotton program has worked well to ensure producers' income,
it has not been economically sound. On the basis of its economic
model, GAO estimates that the program's means of supporting prices
resulted in losses to society of $738 million, on average, from crop
years 1986 to 1993 (August 1 to July 30). GAO's economic model is
designed to evaluate the impact of a commodity program on the
economic welfare of society as a whole by measuring the
inefficiencies resulting from aspects of that program that keep land
from being used productively. With land taken out of production,
society is prevented from benefiting economically from additional
crops. At the same time, reduced supplies raise prices. In
addition, through the program's benefits, the government pays
producers not to produce on the idled acres.
FEDERAL EFFORTS TO ENHANCE
COTTON EXPORTS ARE COSTLY
-------------------------------------------------------- Chapter 0:4.3
USDA's efforts to maintain and enhance exports when world prices are
low have not been effective despite marketing loan and step 2
payments of $3 billion since 1986 to producers, exporters, and
domestic mills. Cotton exports fell significantly in 1988, 1991, and
1992--by 222 million pounds, or 7 percent, in 1988; by 500 million
pounds, or 14 percent, in 1991; and by 700 million pounds, or 23
percent, in 1992.
The marketing loans and step 2 payments were designed to bring the
U.S. price more in line with the lower adjusted world price in order
to maintain exports. However, these provisions did not prevent a
continuing divergence of the adjusted world price and the U.S. price
or a significant decline in exports and market share. This situation
occurred because the marketing loans and step 2 payments cannot
counteract the effect of the cotton program's other provisions, such
as import restrictions, that tend to keep the U.S. price of cotton
higher than the adjusted world price, thereby limiting exports.
Furthermore, because USDA's loan program enables producers to keep
cotton in storage without charge for up to 10 months, producers can
keep their cotton off the market at no cost and speculate on
receiving higher prices.
Although exports are usually considered to be beneficial, this is not
true for cotton. While the cotton industry benefits from exports,
during the period covered by our review, production costs and
government payments, taken together, were consistently higher than
the adjusted world price. As a result, the United States sold cotton
on the world market for less than its cost. Despite these losses,
the United States spent $428 million from fiscal year 1986 through
fiscal year 1994 to promote and market cotton exports. Certain
objectives of NAFTA and GATT--encouraging less government support for
agriculture and a more open world market for raw cotton and cotton
textiles--may make it increasingly difficult in the long term for the
United States to subsidize cotton exports while also supporting
prices under the loan program.
MATTERS FOR CONGRESSIONAL
CONSIDERATION
---------------------------------------------------------- Chapter 0:5
The cotton program has become very costly and complicated as it has
tried to accomplish conflicting objectives, such as supporting farm
prices through direct subsidies and import restrictions while
subsidizing domestic mills and exporters to purchase higher-priced
U.S. cotton. While the cotton program does fulfill its objectives
of protecting farmers' income and managing the cotton supply, a small
percentage of producers receive most of the program's benefits.
Furthermore, the program's provisions to enhance exports have not
prevented declines in exports when world prices are low.
The Congress may wish to consider whether benefits from the cotton
program are worth its costs and whether the program should be
continued. However, any reductions or changes should be made
cautiously. We recognize that if government support is reduced or
eliminated, some producers could not profitably remain in cotton
farming. In addition, because lower government support would cause
declines in land values, some producers and rural economies would be
negatively affected. Among the producers most adversely affected
would be those who are heavily in debt for land or machinery. Thus,
if significant changes are made to the program, the Congress may want
to consider options to give producers and other affected parties time
to make adjustments in their investment decisions. The Congress
could, for example, reduce or phase out payments over a number of
years, perhaps over the life of the next farm bill.
The implementation of GATT's and NAFTA's requirements over a period
of years will give the cotton industry time to make adjustments
during the transition from a government program to greater reliance
on the market. The elimination of trade barriers for both raw cotton
and cotton textiles will push the U.S. cotton industry and program
toward greater market orientation and reliance on market prices.
However, because the economic changes imposed by GATT and NAFTA will
be phased in gradually, the U.S. cotton industry will have some time
to adjust to an environment of a more competitive world market.
AGENCY COMMENTS AND OUR
EVALUATION
---------------------------------------------------------- Chapter 0:6
In commenting on a draft of this report, USDA stated that it is in
fundamental disagreement with certain conclusions reached by GAO. In
particular, USDA took exception to GAO's statement that the economic
situation that led to the cotton program no longer exists--that the
economic emergency and related problems that generated a cotton
program in the 1930s have ended. USDA indicated that such a
statement questions the need for a cotton program. USDA noted that
some problems have not gone away, especially the continuing problem
of trade protectionism and subsidization by other cotton-producing
countries.
GAO believes that the economic conditions affecting farming have
changed significantly since the 1930s. For example, the average
recipient of farm payments is no longer poor; the number of farms has
declined dramatically, while the size of the average farm is much
larger; and farmers can now insure themselves against price declines.
Regarding the trade issue raised by USDA, GAO recognizes that certain
of our foreign competitors intervene in various ways in the world
cotton market. Some of these actions may put our cotton at a price
disadvantage. GAO believes that, ultimately, it is up to the
Congress to determine whether it is in the nation's best interest to
respond to these countries in-kind by continuing to subsidize our
cotton industry. Subsidizing the U.S. cotton industry to make it
more competitive in the world market has been very costly--about $3
billion between 1986 and 1993 for marketing loan and step 2 payments.
USDA also disagreed with the methodology used by GAO to measure
producers' revenues and costs, the losses to society resulting from
land taken out of production, and the losses to the nation from
export sales.
First, USDA believes that the use of 8-year average revenue and cost
figures are of limited value in explaining a real profit/loss result
in a given year because each cost estimate is based on a yield for a
specific year and each revenue estimate is based on an assumed price
for a given year. GAO believes that its use of averages to summarize
cotton producers' costs and revenues is an acceptable methodology
that has explanatory value. Averaging is particularly relevant in
this case, since GAO is trying to assess the economic condition of
cotton producers in both the long and short terms. For example,
GAO's analysis for each year between 1986 and 1993 showed that in the
short run, revenues from market prices alone as a percent of cost
ranged from 107 to 176 percent. In the long run, revenues from
market prices alone as a percent of cost ranged from 66 to 103
percent. Thus, while producers may sustain variations in returns
from year to year, their long-term condition is determined by the
average across years.
Second, USDA notes that GAO's estimate of the loss to society from
the cotton program, as measured by an economic model, results
primarily from program participants' idling some of their land either
voluntarily or in compliance with the program's requirements. USDA
disagrees with the model's use of acres idled voluntarily. USDA
suggests that program participants decide to idle land in response to
the market price of competing crops and not in response to program
requirements. However, GAO believes that its analysis appropriately
considers idled acres. The model is designed to evaluate the impact
of a commodity program on the economic welfare of society as a
whole--not any particular group--by measuring the inefficiencies that
result from aspects of the program that keep land from being used
productively. The idled acres used in GAO's model, whether idled
because of the program's requirements or voluntarily, were idled
under provisions of the cotton program. From the standpoint of
society, the reasons for cotton producers' deciding to idle acres are
not relevant to whether economic inefficiency results. The key point
is that, as a result of the cotton program, productive land of
potential benefit to society is idled.
Third, USDA questioned GAO's use of total (long-run) costs to show
that the United States loses money when exporting cotton. USDA
believes that, rather than total production costs, short-run costs
would be more appropriate. GAO agrees that short-run costs are
appropriate to use if the question is whether exporting cotton is
profitable in the short term, that is, annually or for a few years.
In other words, the United States might be willing to take losses on
cotton exports in one year if it expected that exports would be
profitable the next year. However, under the cotton program, these
losses have continued over the longer term. GAO's analysis found
that the nation sustained losses from 1986 through 1993.
USDA's specific comments and GAO's evaluation of them are discussed
in chapters 2 through 6. GAO made changes to the report in response
to these comments where appropriate. USDA's comments appear in their
entirety in appendix IV.
INTRODUCTION
============================================================ Chapter 1
Cotton is the single most important textile fiber in the world,
accounting for 47 percent of all fibers produced. Cotton is produced
in about 80 countries and generates international trade, not only in
raw cotton but also in value-added products such as yarn, cloth, and
finished textile goods. The United States produces about 18 percent
of the world cotton supply and is generally the largest cotton
exporter.
Like other commodities, cotton farming has become more concentrated
in fewer, larger farms with fewer producers and farm workers. In
1949, the United States had more than 1.1 million cotton farms that
grew an average of 24 acres of cotton per farm. On more than half of
these farms, the family provided almost all of the labor. Less than
one-third of these farms had a tractor. The mechanical harvester had
been developed, but most farms were too small to support one.
Mechanization of cotton farms increased rapidly in the 1950s and
1960s. By 1970, virtually all of the U.S. cotton crop was
mechanically harvested. Technology has increased yields per acre and
nearly eradicated the perennial cotton pest, the boll weevil. The
number of cotton farms declined from over one million in the 1940s to
about 147,000 in 1993.
WORLD COTTON PRODUCTION AND
TRADE
---------------------------------------------------------- Chapter 1:1
Two major types of cotton are produced in the world: upland and
extra-long staple.\1 Upland cotton, the type most commonly grown
throughout the world, accounts for about 98 percent of the U.S.
crop. In 1993, extra-long staple cotton accounted for almost 5
percent of total world production. This report deals with upland
cotton.
The United States produces more upland cotton than any country in the
world except China. China, the United States, and India account for
54 percent of world cotton production. From 1986 through 1993, world
production averaged 36.5 billion pounds (76 million bales) per
year.\2
Table 1.1 shows the world's major cotton-producing countries and
their average upland cotton production over this 8-year period.
Table 1.1
World Cotton Producers by Rank and
Average Production, Crop Years 1986-93
Average annual
amount Percen
produced t of
(billions of world
Rank Country pounds) total
-------------------- -------------- -------------- ------
1 China 9.2 25.2
2 United States 6.7 18.4
3 India 3.9 10.7
4 Pakistan 3.3 9.0
5 Uzbekistan 2.9\a 7.9
Rest of the 10.9 29.9
world
World total 36.5 100.0
------------------------------------------------------------
Note: Totals may not add because of rounding and because data for
Uzbekistan were not available for all years.
\a The average for Uzbekistan is based on 7 years because data for
1986 were not available.
Source: GAO analysis of USDA data.
Cotton exports in recent years have averaged about 40 percent of U.S.
cotton production. The United States usually exports the largest
quantity of cotton. However, the U.S. share of world exports
fluctuates from year to year. For example, in 1988, the U.S. share
of exports was 18 percent, and in 1993, it was 26 percent.
Uzbekistan, Australia, India, Pakistan, China, and a consortium of
African nations are the major competitors of U.S. cotton exports.
Most U.S. cotton exports go to Pacific Rim countries with large
textile and apparel industries. During the 1992 crop year, the
largest importers of U.S. cotton were Korea, Japan, Mexico, and
Indonesia.
Although the United States is a major exporter of raw cotton, it is a
net importer of processed and finished cotton goods. To illustrate,
in 1992 and 1993, the United States exported an average of about 3
billion pounds of raw cotton annually and in return had net imports
of textile and apparel products containing the equivalent of about
2.5 billion pounds of raw cotton.\3 The value of cotton textile and
apparel products is several times greater than the value of raw
cotton alone.
--------------------
\1 Besides the United States, extra-long staple cotton is grown
primarily in India, Egypt, and the former Soviet Union countries.
Extra-long staple cotton is characterized by fineness and high fiber
strength, contributing to finer and stronger yarns and is more
expensive than upland cotton. Textile products made from extra-long
staple cotton are considered to be of higher quality and more
luxurious and are also higher priced than similar products made from
upland cotton.
\2 Cotton is processed at gins into standard 480-pound bales.
\3 Imports for 1992 and 1993 averaged 3.4 billion pounds, while
exports averaged .9 billion pounds, for a net import of 2.5 billion
pounds.
U.S. COTTON PRODUCTION AND
DOMESTIC USE
---------------------------------------------------------- Chapter 1:2
Cotton is grown in 17 states from Virginia to California, with the
greatest concentrations in four distinct regions, as shown in figure
1.1. Cotton production from 1986 to 1993 averaged 6.7 billion pounds
per year, ranging from 4.6 billion pounds in 1986 to a high of 8.3
billion pounds in 1991. The U.S. Department of Agriculture (USDA)
projects 1994 cotton production at 9.3 billion pounds, which if
realized, would be the largest crop on record.
Figure 1.1: Average Percent of
U.S. Upland Cotton Production
by State, Crop Years 1986-93
(See figure in printed
edition.)
Note: Numbers do not add to 100 percent due to rounding.
Source: USDA's Cotton and Wool Situation and Outlook Report.
Yields per harvested acre vary widely from year to year and from
state to state. Production depends on many factors, including soil
productivity, climate, producer management skills, cost of
production, market conditions, and government programs.
About 60 percent of the cotton grown in the United States is used
domestically. In 1993, mill use of upland cotton was about 5 billion
pounds, almost all of which was domestically grown. Cotton is
typically used to make yarn, thread, cords, and rope and is woven
into fabrics that are then used to make apparel, house furnishings,
and floor coverings.
U.S. COTTON POLICY
---------------------------------------------------------- Chapter 1:3
The federal government intervened in agricultural markets in the
1930s because the Great Depression had severely disrupted the
domestic economy and resulted in low prices for goods sold. U.S.
farm families, whose income at that time was only about one-half that
of nonfarm families, saw their income decline further as farm
commodity prices dropped by 50 percent. This income reduction was
felt throughout the nation because over 31 million people--or
one-fourth of the U.S. population--then lived on farms.
To correct the income imbalance between farm and nonfarm families and
stabilize the agricultural market, the Congress enacted the
Agricultural Adjustment Act of 1933. The 1933 act established a
comprehensive program aimed at controlling production of designated
"basic" commodities, including cotton. The programs mandated by the
1933 act and its amendments were intended to be temporary; they were
to be terminated as soon as the President declared an end to the
national emergency.
The goals and provisions of current cotton legislation trace back to
this act and the Agricultural Adjustment Act of 1938. After World
War II, support provisions were recodified in the Agricultural Act of
1949, which still serves as the main U.S. farm law. At intervals, a
new farm act is passed that amends the 1949 act and supersedes the
previous farm act. Each of these acts added to the complexity of the
program--new features were incorporated or existing features were
deleted or revised. The most recent of these acts was the Food,
Agriculture, Conservation, and Trade Act of 1990 (known as the 1990
Farm Bill).
The major objectives of USDA's cotton program are to protect U.S.
farm income, manage cotton supply levels for domestic mill use and
export, and maintain competitive U.S. cotton prices in world
markets. Under the current U.S. farm policy, USDA tries to
accomplish these objectives through deficiency payments, nonrecourse
loans, acreage reduction, and various import and export provisions.
Each of these program components is discussed below.
About 103,000 of the existing 147,000 U.S. cotton-producing farms
participated in the cotton program in 1993. From 1986 through 1993,
between 84 to 92 percent of cotton acreage was enrolled in the cotton
program.
DEFICIENCY PAYMENTS
-------------------------------------------------------- Chapter 1:3.1
One element of the cotton program as authorized in the 1990 Farm Bill
provides direct government payments to cotton producers when market
prices are low. These payments are known as deficiency payments.
Under this program component, a minimum target price is legislatively
set for each year in a 5-year period, and deficiency payments are
made to support producers' incomes when the calendar year national
average price producers receive for their cotton falls below the
target price.\4 For 1986 to 1993, cotton deficiency payments averaged
almost $1 billion annually. The Food Security Act of 1985 reduced
target prices from 81 cents per pound to 72.9 cents per pound from
1986 to 1990. The 1990 Farm Bill set target prices at 72.9 cents per
pound for each of the 5 years covered by the act.
--------------------
\4 The deficiency payment rate is the difference between the target
price and either the national average market price or the loan rate,
whichever is higher.
NONRECOURSE LOANS
-------------------------------------------------------- Chapter 1:3.2
In conjunction with target prices and deficiency payments, the 1990
Farm Bill also continued the use of nonrecourse loans to cotton
producers. USDA's Commodity Credit Corporation (CCC)\5 makes these
loans at an established loan rate, and producers, in turn, pledge
their stored cotton as collateral. The nonrecourse loan rate is
calculated by USDA following a statutory formula that is based on
historical market prices. This rate is expressed in cents per pound
of cotton and, for 1993, it was 52.35 cents. For 1994, the rate was
50 cents per pound, the lowest rate allowed under the 1990 Farm Bill
for any crop year.
Essentially, these loans support the prices farmers receive by
establishing a minimum price for cotton. The loans are nonrecourse
because producers may forfeit their stored cotton to CCC as payment
of their loan in full, regardless of the market value of cotton at
that time. In this situation, producers keep the loan proceeds, and
the government bears the costs of storing, transporting, and
disposing of the forfeited cotton. These costs become a loan loss to
USDA.
Nonrecourse loans for cotton mature 10 months from the first day of
the month in which they were made. At the end of the 10-month loan
period, producers can elect to (1) repay the loan, (2) forfeit their
pledged cotton as full loan repayment, or (3) depending upon average
market prices,\6 request that the loan maturity date be extended for
8 months. Thus, producers may have 18 months in which to keep their
cotton under loan and off the market.
The Food Security Act of 1985 introduced a new concept, the marketing
loan provision, that allows producers to redeem nonrecourse loans at
a discount, referred to as a marketing loan gain. As an alternative,
producers may receive an equivalent amount, referred to as a loan
deficiency payment, by agreeing to forego nonrecourse loans. The
marketing loan was devised to help keep U.S. cotton prices
competitive in world markets, thus encouraging producers to sell
their cotton instead of keeping it under loan and off the market.
The marketing loan changes the nonrecourse loan repayment process by
permitting producers to repay their loans at the lower of the loan
rate or the USDA-calculated adjusted world price.\7 When the adjusted
world price is below the loan rate, producers have an opportunity to
sell their cotton at any price they can obtain from the market and
receive the marketing loan gain from USDA for the difference between
the loan rate and the USDA-calculated adjusted world price. In
addition, when producers receive a marketing loan gain, they also
benefit because USDA pays for storage. Cotton is the only USDA
commodity program in which USDA pays these storage costs.
--------------------
\5 CCC is a wholly owned, government corporation created in 1933 to
(1) stabilize, support, and protect farmers' incomes and prices; (2)
maintain balanced and adequate supplies of agricultural commodities;
and (3) assist in the orderly distribution of those commodities. CCC
finances its operations by borrowing from the U.S. Treasury.
However, the agency is reimbursed for losses resulting from its
operations by annual appropriations. CCC has no staff itself but
administers its programs through a separate USDA agency, the
Consolidated Farm Service Agency, that includes the former
Agricultural Stabilization and Conservation Service, which deals with
farmers through a network of county offices.
\6 Loans may be extended, provided the spot market average price for
average (base) quality cotton does not exceed 130 percent of its
price for the preceding 36 months.
\7 The adjusted world price is an estimate of the prevailing world
price--the average of the five lowest quoted prices for cotton from
various countries in the northern European market--as calculated
weekly and adjusted to U.S. quality and location by the Secretary of
Agriculture.
LIMITATIONS ON PAYMENTS
-------------------------------------------------------- Chapter 1:3.3
In response to concerns about large payments to farm operations and
the overall cost of federal farm programs, beginning with the 1971
crop, the Congress limited the annual amount of certain program
payments a person could receive. Today, a "person" may receive up to
$50,000 in deficiency payments and up to $75,000 in marketing loan
gains and loan deficiency payments annually. Both of these limits
are included within an overall limit of $250,000 that also includes
disaster payments and other adjustments. "Persons" may be not only
individuals--including those participating in general partnerships or
joint ventures--but also entities such as corporations. Each
individual or entity who qualifies as a separate "person" and meets
additional "actively engaged in farming" requirements is eligible to
receive payments up to the applicable limits. Some cotton farming
operations are so organized as to have numerous "persons" associated
with them. This effectively increases the amount of payments that
the operation receives beyond the amount normally available to an
individual.
ACREAGE REDUCTION
-------------------------------------------------------- Chapter 1:3.4
USDA attempts to manage the domestic supply of cotton through several
program provisions to support prices and limit deficiency payments.
Each of these provisions results in land being idled. The provisions
are an acreage reduction program; flexibility provisions, also called
"flex acres;" and the 50/92 program.\8 Producers who participate in
the cotton program must follow the requirements of the acreage
reduction program but may voluntarily idle additional acres through
the flex acres and 50/92 programs. Figure 1.2 shows the extent to
which these programs have been used to take land out of production
since crop year 1986, including those flex acres that remain idle.
Figure 1.2: Use of Acreage
Reduction Program, Idled Flex
Acres, and the 50/92 Program,
Crop Years 1986-93
(See figure in printed
edition.)
Source: GAO analysis of USDA data.
--------------------
\8 The Omnibus Budget Reconciliation Act of 1993 changed this program
to a 50/85 program for crop year 1994. However, during the period
covered by our evaluation, it was the 50/92 program, and for this
report we will refer to it as the 50/92 program.
ACREAGE REDUCTION PROGRAM
-------------------------------------------------------- Chapter 1:3.5
Under this program, producers are required to remove acreage from
production as a condition for participating in USDA's cotton program.
The 1990 Farm Bill provides for the Secretary to use the acreage
reduction program to ensure that U.S. stocks at the end of the crop
year are at about 30 percent of domestic consumption. For 1993, the
Secretary set the acreage reduction program rate at 7.5 percent of
producers' base acreage;\9 for 1994, the acreage reduction program
rate was 11 percent. (Base acreage is used by USDA to determine
deficiency payments for the cotton program.) In December 1994, the
Secretary announced that because of significant increases in export
sales of cotton, farmers would not have to idle any of their cotton
acreage to qualify for program benefits in 1995.
--------------------
\9 The cotton base equals a farm's 3-year average acreage of cotton
planted for harvest, plus land not planted because of acreage
reduction programs during a period specified by law.
FLEX ACRES
-------------------------------------------------------- Chapter 1:3.6
The Omnibus Budget Reconciliation Act of 1990, together with the 1990
Farm Bill, authorized a two-part flex-acre provision. Under the
first provision--called "normal" flex acres--producers do not receive
deficiency payments on 15 percent of their enrolled base acreage. In
lieu of these payments, producers are permitted to plant other crops
(except fruits or vegetables) and maintain their cotton base. They
can also continue to plant cotton on the flex acres, in which case
they remain eligible to receive marketing loans with any associated
marketing loan gains.
In addition to the normal flex acres, producers can plant crops other
than cotton on another 10 percent of their land--known as optional
flex acres--without a reduction in their cotton base acreage. As is
the case with normal flex acres, however, deficiency payments will
not be paid for those acres used to grow crops other than cotton.
Both the normal and optional flex acres allow producers to plant
alternative crops on the basis of market signals without losing any
of their cotton base acres, on which future government payments will
be determined.
50/92 PROGRAM
-------------------------------------------------------- Chapter 1:3.7
Cotton producers who plant at least 50 percent of their maximum
payment acres (acres enrolled in the program less acreage reduction
program acres and other program requirements) and devote the rest to
conserving uses or approved nonprogram crops, are allowed to receive
deficiency payments on 92 percent of their maximum payment
acreage.\10 The purpose of this provision is to reduce cotton stocks
while allowing producers to retain most of their deficiency payments
and protect their cotton base. In crop year 1993, farmers received
almost $34 million in deficiency payments under this program.
--------------------
\10 The 50-percent minimum is disregarded if the Secretary of
Agriculture determines that producers are prevented from planting
because of drought, flood, or other natural disasters.
IMPORT AND EXPORT PROVISIONS
-------------------------------------------------------- Chapter 1:3.8
The 1990 Farm Bill included three provisions (referred to as steps)
to help ensure that U.S. cotton would be competitive in world
markets. Step 1 permits the Secretary to lower the adjusted world
price under certain price conditions; however, this step has not been
used since April 1992. Step 2 provides for USDA payments to
exporters and to domestic mills when U.S. prices have been higher
than world prices for 4 consecutive weeks--this step has cost about
$553 million since 1991. Step 3 is an import quota that must be
implemented when U.S. prices exceed world prices for 10 consecutive
weeks and the spot market quota, described below, is not in effect.
On two occasions, the price conditions met the 10-week requirement,
but the quota was not implemented because the spot market quota was
already in effect.
Import quotas have been a part of the cotton program since the 1930s.
Since then, farmers' incomes have been supported by an annual cotton
import quota of about 125,000 bales (60 million pounds) imposed
pursuant to section 22 of the Agricultural Adjustment Act of 1933, as
amended (so-called section 22 quota restrictions). This quota was
imposed to prevent U.S. textile mills from purchasing unlimited
supplies of cotton from foreign sources. However, this provision was
superseded by passage of the Uruguay Round of the General Agreement
on Tariffs and Trade (GATT) in late 1994. In addition, the Food
Security Act of 1985 provides for a temporary import quota based on
spot market prices. This spot market quota permits entry into the
United States of a quantity of cotton equal to 21 days of cotton use
by U.S. textile mills. It is to be implemented during 90-day
periods when the current U.S. spot market price for cotton exceeds
historical price averages by specific amounts. The U.S. spot price
represents the average of quoted prices for cotton in seven U.S.
geographical areas, as designated by the Secretary of Agriculture.
EXPORT PROMOTION AND
INTERNATIONAL TRADE AGREEMENTS
---------------------------------------------------------- Chapter 1:4
In addition to supporting import and export provisions for cotton,
the United States supports promotion programs for agricultural
exports and an open commercial trading system. USDA promotes cotton
exports through a series of initiatives, administered by its Foreign
Agricultural Service, to develop foreign markets. Also, the United
States is a party to GATT and the North American Free Trade Agreement
(NAFTA). GATT will allow increasing levels of cotton imports and
replace section 22 quota restrictions with tariffs. NAFTA,
implemented on January 1, 1994, will also result in import quotas
being replaced with tariffs. Both of these agreements will push the
industry toward a greater reliance on the market.
PLAYERS IN THE COTTON PROGRAM
---------------------------------------------------------- Chapter 1:5
Many participants have to interact to make the cotton program
function. Figure 1.3 lists the key players and briefly describes
their roles.
Figure 1.3: Key Players in the
U.S. Cotton Program
(See figure in printed
edition.)
OBJECTIVES, SCOPE, AND
METHODOLOGY
---------------------------------------------------------- Chapter 1:6
In response to a request from Representative Richard K. Armey, we
evaluated the cotton program's cost and complexity, distribution of
payments and benefits, effects on producers' costs and returns, and
effectiveness in enhancing U.S. cotton exports. We interviewed USDA
officials, industry representatives, and producers. We analyzed USDA
data on cotton production, prices, payments, costs of production, and
exports for 1986 through 1993. We selected this period for our
review of the cotton program to focus on recent program changes made
by the 1985 and 1990 Farm Bills, including the marketing loan
provision and the three step provisions to address competitiveness.
Also, we used an economic model developed by Bruce L. Gardner,
University of Maryland, to assess the economic impact of the program
on cotton buyers' costs and producers' benefits. (App. I provides a
detailed explanation of the economic model.) In addition, we
evaluated the potential effects on the cotton program of NAFTA and
GATT. As necessary, we adjusted figures in this report to 1993
dollars to more accurately compare prices and costs over time. For
this adjustment, we used the gross domestic product implicit price
deflator, with 1993 being equal to 1.00. Further details on our
objectives, scope, and methodology are provided in appendix II.
We conducted our review from February 1994 through March 1995 in
accordance with generally accepted government auditing standards.
However, we did not independently verify data provided to us by USDA.
USDA's comments on a draft of this report and our responses to them
are included in chapters two through six and appendix IV.
THE COTTON PROGRAM IS COSTLY AND
COMPLEX
============================================================ Chapter 2
The cotton program contains a complex series of provisions that cost
the government hundreds of millions of dollars each year. For crop
years 1986 through 1993, the program cost the government about $1.5
billion (1993 dollars) annually. These costs do not include the cost
of other government programs that also support cotton producers.
These other programs, which include farm credit, disaster assistance,
export assistance, and water subsidies, substantially increase the
cost of the government's involvement in cotton production. In
addition, federal budgetary costs were slightly offset because U.S.
cotton buyers gained an estimated average annual savings of $16
million as a result of lower cotton prices.
Modifications to the program since 1985 have added new features, such
as the marketing loan, flex acres provisions, and export subsidies,
that have added to the cost of the program and made it more
complicated.
COTTON PROGRAM IS COSTLY
---------------------------------------------------------- Chapter 2:1
As shown in table 2.1, during crop years 1986-93, government costs
averaged about $1.5 billion annually in support of cotton crops that
were valued at an average of about $4.5 billion annually. Thus, for
every dollar producers realized from the sale of cotton, the
government spent 33 cents in support of the program.
Deficiency payments accounted for the largest single government cost
category, averaging $995 million annually. Other major government
cost categories included marketing loan gains, step 2 payments, and
loan deficiency payments that together averaged $497 million
annually. These costs were slightly offset by cotton buyers' gains.
Cotton buyers paid less for cotton in 1986, 1992, and 1993 than they
would have without a cotton program. These gains averaged an
estimated $16 million annually from 1986 through 1993.
Table 2.1
Annual Cotton Program Costs, Crop Years
1986-93
(1993 dollars in millions)
Cost category Total Average
------------------------------------ ---------- ----------
Deficiency payments\a $7,958 $ 995
Loan deficiency payments 942 118
Marketing loan gains 1,558 195
Step 2 payments\b 553 184
Loan losses 915 114
Storage costs\c 149 19
============================================================
Total government costs $12,075 $1,509
Cotton buyers' gain (131) (16)
============================================================
Total program cost $11,944 $1,493
Market value of production $35,908 $4,488
------------------------------------------------------------
\a Deficiency payments include the costs of the 50/92 program, which
averaged $47.4 million for the 8 crop years covered by our analysis.
\b Step 2 payments were authorized in the 1990 Farm Bill and were
made only during crop years 1991 through 1993.
\c These amounts are for storage costs paid for cotton owned by CCC.
Dollars were provided by USDA on a fiscal year basis but have been
presented here by crop year. For example, fiscal year 1987
corresponds to crop year 1986. For the 8 years in our analysis, USDA
also paid an estimated $448 million (1993 dollars) for storage costs
for cotton under the loan program. These costs were included in the
marketing loan category for 1991-93, but for earlier years, USDA
personnel were unsure about what category absorbed these costs.
Source: GAO calculations based on USDA data.
Over the 8-year period, 1986-93, government costs varied
considerably. For example, government costs ranged from $439 million
in 1990 to $2.8 billion in 1986. In general, government costs were
low when cotton prices were high and program features, such as the
marketing loan, were not in effect. For example, in the current crop
year, 1994, government costs are expected to be reduced because of
high market prices, low deficiency payments, and minimal marketing
loan gains.
Costs for administering the cotton program are not included in table
2.1 because USDA does not normally accumulate such costs by program.
USDA budget officials estimated that for fiscal year 1993,
administrative costs amounted to about $14.4 million.
On average, cotton buyers gained an estimated $16 million annually by
paying lower prices for cotton than they would have without the
program in 1986, 1992, and 1993. Cotton buyers' gains in 1992 and
1993 derived from step 2 payments made to domestic mills, which
contributed to lower prices. From 1987 to 1991, however, cotton
buyers incurred costs ranging from $28 million to $233 million by
paying higher prices for cotton than they would have without the
program. Cotton buyers' costs are incurred through several
mechanisms that the government uses to limit cotton acreage and
production. The primary mechanisms include the acreage reduction
program, the 50/92 program, and idled flex acres. To the extent that
these mechanisms result in lower production and less supply, cotton
buyers pay higher prices with the program than they would without it.
Appendix I describes cotton buyers' costs and gains in more detail.
NUMEROUS OTHER GOVERNMENT
PROGRAMS SUPPORT COTTON
PRODUCERS AT ADDED COSTS
---------------------------------------------------------- Chapter 2:2
Multiple USDA agencies administer other agriculture programs that
support cotton and other crops and are intended to, among other
things, minimize adverse impacts on the environment, promote cotton
exports, provide disaster assistance, control pests, and provide
credit assistance. In addition, the Department of the Interior
administers a program that provides subsidized irrigation water to
producers in western states. A brief description of these programs
and estimates of their benefits to cotton farmers, based on available
data, are provided below. Amounts for these programs are given in
1993 dollars, except where noted otherwise.
Conservation Reserve Program. This program authorizes USDA to
contract with farmers to take highly erodible and other
environmentally sensitive cropland out of production. Cotton
producers received an estimated $80 million to $100 million from this
program in fiscal year 1993.
Export Promotion. The Foreign Agricultural Service administers
several programs to help the U.S. agriculture sector sell products,
including cotton, abroad. These programs, collectively referred to
as "export promotion," are intended to help maintain existing markets
and develop new markets. Export promotion costs for cotton totaled
$428 million from fiscal years 1986 to 1994, an average of $48
million per year.
Disaster Payments and Crop Insurance. Since 1980, USDA has provided
disaster assistance to farmers through direct cash payments and
subsidized insurance. From crop years 1988 through 1993, disaster
payments for cotton losses totaled $804 million, an average of $134
million per year. In addition, costs associated with crop insurance,
including subsidized premiums, insurance commissions, administration,
and excess losses, totaled about $896 million during crop years
1986-94, an average of $100 million per year.
Pest Management. The Animal and Plant Health Inspection Service
conducts cooperative programs with state and local agencies to
control and eradicate the boll weevil and the pink bollworm. From
fiscal years 1986 through 1994, funding for these two programs
totaled about $126 million, an average of about $14 million annually.
Market News Service. The Agricultural Marketing Service collects,
analyzes, and disseminates market information for numerous
agricultural commodities, including cotton. Federal outlays for the
Cotton Market News Service have averaged about $2 million annually
since 1986.
Farm Credit. The Farmers Home Administration,\1 a lending agency
within USDA, provides assistance to financially troubled farmers
through direct government-funded loans and federal guarantees on
loans made by other agricultural lenders. Because of loan defaults,
the agency lost about $6.3 billion on its farm loan programs during
fiscal years 1991-94, some of which was attributable to cotton
farmers. However, the Farmers Home Administration does not maintain
accounting records that show losses by type of crop.
Agricultural Research. Two USDA agencies--the Agricultural Research
Service and the Cooperative State Research Service--conduct
agricultural research, some of which benefits cotton farmers. The
Agricultural Research Service estimates that it spent about $38
million for cotton-related research in fiscal year 1994. The
Cooperative State Research Service's funding for cotton research
totaled $3.7 million in fiscal year 1994.
Water Subsidies. The Department of the Interior's Bureau of
Reclamation plans, constructs, and operates water resource projects
to provide irrigation water to arid and semiarid lands in the 17
western states. Generally, interest on the federal government's
costs incurred in constructing the irrigation component of the
project facilities is not included in water irrigation rates. Cotton
farmers, particularly in California, benefit from these irrigation
subsidies. However, the Bureau does not maintain information by crop
on federal irrigation subsidies. Although recent studies are not
available, Interior estimated that the subsidy in 1986 was $534
million, and cotton production accounted for about 7 percent of the
irrigated acres.
--------------------
\1 In 1994, the responsibility for administering farmer program loans
was transferred from the Farmers Home Administration to the newly
created Consolidated Farm Service Agency.
COTTON PROGRAM HAS BECOME
INCREASINGLY COMPLEX IN AN
EFFORT TO ACHIEVE MULTIPLE
OBJECTIVES
---------------------------------------------------------- Chapter 2:3
The cotton program has become increasingly complex because of new
features and changes to existing ones, especially since 1985. As
with many pieces of agricultural legislation, the government has
attempted to manage cotton production and marketing through a
combination of policies affecting prices, payments, acreage, and
import restrictions. Making the program work requires the interplay
of a wide variety of factors throughout the year. The magnitude of
the program's complexity is indicated in the nine figures presented
in appendix III.
In administering the cotton program and determining the payments to
be made to the various participants, USDA must deal concurrently with
at least six levels of domestic and international prices, eight
acreage reduction and yield mechanisms, and three different types of
import restrictions--plus the effects of GATT and NAFTA. The outcome
of this interplay of prices and supply control mechanisms results in
USDA's making at least seven types of payments to cotton producers,
exporters, millers, and other buyers. USDA headquarters personnel
generally perform the price calculations, sending the results on to
its county offices, which calculate acreage determinations and
payments to producers.
The interactions required by the cotton program are often at
cross-purposes. For example, the program provides a target price
that serves as a guaranteed price for producers, who respond by
increasing production. This response would have the effect of
lowering market prices for cotton. To counteract this effect, the
program then imposes supply-control features, such as the acreage
reduction program, to reduce production. These features tend to
increase prices. Also, when prices are low, the program supports the
price the producers get, then subsidizes buyers to reduce their
costs. In addition, the program attempts to make cotton prices
competitive, in order to move cotton quickly to market. However, it
also allows producers to put their cotton under the loan program and
hold it off the market for up to 18 months (free of storage costs for
10 of those months) and to ultimately forfeit their cotton if prices
do not rise sufficiently.
Two of the cotton program's many complex processes are discussed
below. These two processes are the calculation of the adjusted world
price and the step 2 payment rate.
CALCULATION OF ADJUSTED
WORLD PRICE
-------------------------------------------------------- Chapter 2:3.1
USDA, both daily and weekly, analyzes domestic and international
cotton price quotations to calculate the prices used to determine
program payments. The adjusted world price is an important component
in establishing rates for marketing loan payments, loan deficiency
payments, and step 2 payments. The method for calculating the
adjusted world price, as described in USDA publications, is
inherently difficult:
"The adjusted world price (AWP) is equal to the Northern Europe
price, 'NE Price,' (an average of the 5 lowest-priced growths
for Middling (M) 1-3/32 inch cotton, CIF [cost, insurance, and
freight] northern Europe), adjusted to average U.S. quality and
location. The AWP for individual qualities is determined by
applying the 1994 schedule of loan premiums and discounts, and
location differentials. An additional 'coarse count adjustment'
(CCA) may be applicable for cotton with a staple length of
1-1/32 inches or shorter and for certain specific lower grades
with a staple length of 1-1/16 inches and longer. The AWP and
CCA for the subsequent week are announced each Thursday.
"Under certain conditions, the AWP may be adjusted downward.
These conditions are when 'the lowest U.S. growth quote for M
1-3/32 inch cotton, CIF northern Europe' (USNE price) exceeds
the NE price and the AWP is within 115 percent of the 1993-crop
loan level. This is known as the Step 1 competitiveness
provision. When these conditions are triggered the Secretary
has the discretion to lower the AWP by up to the difference of
USNE price minus NE price."
Figure 2.1 depicts this process.
Figure 2.1: Process for
Calculating the Adjusted World
Price
(See figure in printed
edition.)
.
(See figure in printed
edition.)
CALCULATION OF STEP 2
PAYMENT RATE
-------------------------------------------------------- Chapter 2:3.2
Computations used to determine whether step 2 payments should be
triggered and at what rate are similarly complex. USDA publications
describe the procedures as follows:
"Whenever, for 4 consecutive weeks, the USNE [U.S. northern
European] price exceeds the NE price by more than 1.25 cents per
pound and the AWP is within 130 percent of the base loan rate
for the 1994-crop upland cotton, marketing certificates will be
issued to eligible domestic users and exporters on eligible
cotton designated during the week following the consecutive
4-week period. For domestic users, eligible cotton is any bale
'opened' at the mill during the subsequent week. For exporters,
eligible cotton is any 'contracted' for shipment by specified
dates. The payment rate is equal to the fourth week's USNE
price minus NE price minus 1.25 cents/pound."
Step 2 payments will continue only if the USNE price, adjusted by any
step 2 payment announced on the previous Thursday, has not exceeded
the NE price by at least 1.25 cents per pound for 10 consecutive
weeks, according to USDA staff.
Figure 2.2 depicts this process.
Figure 2.2: The Step 2 Process
(See figure in printed
edition.)
AGENCY COMMENTS AND OUR
EVALUATION
---------------------------------------------------------- Chapter 2:4
USDA agreed with the data in our report showing that program costs
averaged $1.5 billion annually for crop years 1986-93 and that for
every dollar producers realized from the sale of cotton, the
government spent 33 cents in support of the program. USDA explained
that these costs were lower than for many other program crops and
that costs were high in several years because of atypical
circumstances. Specifically, USDA explained that costs were high
because (1) in 1986 and 1987, USDA had to dispose of stocks that had
accumulated under the previous program before marketing loans were in
effect and (2) in 1991 and 1992, a worldwide recession and the
breakup of the Soviet Union drove world cotton prices down, which
added to the cost of the cotton marketing loan program. In addition,
USDA noted that projected program costs for 1995-2000 will average a
little more than half of what they were for crop years 1986-93.
Finally, USDA agreed that the cotton program is undeniably complex.
USDA provided various explanations for the program's complexity,
including requirements in the legislation. In addition, USDA noted
that some aspects of the program were developed after "exhaustive
consultations with the cotton industry concerning the appropriate
variables to include."
We have not reviewed projections of future cotton program costs.
However, we would point out that international commodity markets are
inherently volatile because of weather, domestic and international
political events, disease and pest outbreaks, and major economic
forces in individual countries and regions. As a result, any such
projections are subject to much uncertainty. We would also point out
that USDA's projections of cotton program expenditures for 1991-95
proved to be well below the actual experience. For example, in its
Commodity Credit Corporation Commmodity Estimates Book, USDA
projected expenditures of $3 billion, or about $600 million annually
over that period. However, actual costs through 1994--$5.7
billion--about $1.4 billion annually, have nearly doubled the total
projected costs for the period. In addition, although prices are
expected to be high and program costs low in 1995, it is likely that
producers, both in the United States and elsewhere, will respond to
these prices with increased production, which will probably result in
low prices and high program costs in subsequent years.
MOST PROGRAM PAYMENTS GO TO A
RELATIVELY FEW LARGE PRODUCERS
============================================================ Chapter 3
Since the 1940s, the number of cotton farms in the United States has
declined, and cotton production has become concentrated among a
relatively few large producers. A producer can be an individual or a
farm operation organized as a joint venture, partnership, or
corporation. One producer can operate several farms. Cotton
production in the United States is dominated by the largest
producers, who control most of the cotton acreage and receive most of
the cotton program payments. Legislation generally limits the amount
of payments a producer can receive, but producers may organize their
operations in ways that result in their receiving more than the
$250,000 limit applicable to a single "person."
COTTON PAYMENTS AND PRODUCTION
HAVE BECOME CONCENTRATED
---------------------------------------------------------- Chapter 3:1
The number of cotton farms has decreased, from more than one million
in 1949 to about 147,000 in 1993. With this decrease, control of
cotton production has become concentrated among a relatively few
producers. USDA's data bases for crop year 1993 showed that 95,479
producers controlled the 103,000 farms participating in the program.
During that year, the top 20 percent of producers grew an estimated
5.7 billion pounds of cotton, which was 79 percent of production, and
received 79 percent of the government payments, as shown in table
3.1.
Table 3.1
Concentration of Cotton Farms, Acres,
Production, and Government Payments
Among U.S. Producers, Crop Year 1993
(cumulative)
Estimate Governme
d nt
producti payments
Percent Number Farms Acres on \a
---------- -------- -------- -------- -------- --------
1 955 4.0 11.4 16.2 16.1
10 9,548 32.1 50.0 57.4 57.1
20 19,096 56.6 74.3 79.1 78.6
50 47,740 88.4 95.6 96.7 96.7
100 95,479 100.0 100.0 100.0 100.0
------------------------------------------------------------
\a Amount of payments does not include about $262 million in loan
deficiency payments and marketing loan gains paid through
cooperatives to producers. We did not have the data available to
allocate these payments to producers.
Source: GAO analysis of USDA cotton farm and producer data bases.
SOME PRODUCERS RECEIVED MORE
THAN $250,000
---------------------------------------------------------- Chapter 3:2
For the 1993 crop year, 295 cotton producers received thousands of
dollars more than the overall limit of $250,000, including 4
producers who received more than $1 million. Such payments occur
because the legislation that defines limits also allows producers to
organize their operations in ways that enable them to receive
payments of more than $250,000.
USDA's 1993 data bases disclosed that of the 95,479 producers, 295
received over $250,000 in cotton payments. Table 3.2 shows the
distribution of payments for crop year 1993.
Table 3.2
Distribution of Cotton Payments, Crop
Year 1993 (cumulative)
Producer Percen Total Percen
Payment ranges s\a t Payments\b t
------------------ -------- ------ -------------- ------
> $1,000,000 4 .004 $7,647,785 .5
> 250,000 295 .31 $119,107,558 7.9
> 100,000 2,099 2.20 $381,061,563 25.3
> 50,000 8,045 8.43 $788,336,778 52.2
> 10,000 29,860 31.27 $1,357,703,468 90.0
> 1 95,479 100.00 $1,508,884,835 100.0
------------------------------------------------------------
\a A producer can be an individual or a farm operation organized as a
joint venture, partnership, or corporation. One producer can operate
several farms.
\b Amount of payments does not include about $262 million in loan
deficiency payments and marketing loan gains paid through
cooperatives to producers. We did not have the data available to
allocate these payments to producers.
Source: GAO analysis of USDA data.
All of the 295 producers who received in excess of $250,000 were
entities organized as joint ventures, partnerships, and corporations.
For example, the operation that received the most in cotton program
payments ($4.4 million) for 1993 was a general partnership with 39
members who formed 66 corporations, covering more than 20 farms that
produced an estimated 16 million pounds of cotton in three counties
within two states.
AGENCY COMMENTS AND OUR
EVALUATION
---------------------------------------------------------- Chapter 3:3
In its comments on a draft of this report, USDA referred to table
3.1, which shows that about 79 percent of government cotton payments
were made to producers who controlled about 79 percent of cotton
production. USDA stated that it is clear from table 3.1 that
government payments are directly related to the amount of production,
not to the number of producers. USDA also commented that other than
the rules concerning payment limitations, government payments are
designed to be based on production or on base acreage on the farm.
We recognize that government payments are related to production. At
the same time, the data also show that the majority of government
payments go to a small number of producers. Both production and the
number of producers are related to government payments because
control of cotton production has become concentrated among a
relatively few producers. (As we stated, "A producer can be an
individual or a farm operation organized as a joint venture,
partnership, or corporation.") As we said earlier in this chapter,
cotton production in the United States is dominated by the largest
producers who control most of the cotton acreage and receive most of
the cotton program payments.
COTTON PROGRAM HAS PROVIDED HIGH
RETURNS TO PRODUCERS AND CREATED
TROUBLING ECONOMIC CONSEQUENCES
============================================================ Chapter 4
Cotton farming under the U.S. cotton program is a profitable
venture. Revenues from the domestic market alone are substantially
higher than producers' average annual short-run fixed and variable
production costs and are nearly sufficient to cover total (long-run)
costs of operation, including a return on assets. When government
payments are added to producers' revenue, total revenues are well
above all costs of production.
In addition to providing substantial returns, the program has
produced two troubling economic consequences. First, it has reversed
producers' relationship to the market: Producers receive higher
returns when market prices are low. Second, program features, such
as the acreage reduction program, require land to be taken out of
production, creating economic inefficiencies. This results in a net
economic loss to society, sometimes called a social welfare loss.
DOMESTIC MARKET RETURNS PLUS
GOVERNMENT PAYMENTS EXCEED
PRODUCTION COSTS
---------------------------------------------------------- Chapter 4:1
Domestic prices alone have exceeded the short-run production costs of
the average producer by 38 percent from 1986 through 1993. Over the
same period, domestic prices, on average, have provided sufficient
revenue to cover 85 percent of USDA's estimate of the average
producer's existing total production costs. This
market-revenue-to-costs comparison is conservative--if the cotton
program did not exist, production costs would be lower. When the
cotton program's deficiency and marketing loan payments are added to
revenues from the market, the average producer's revenues exceed
total production costs by 17 percent. Table 4.1 summarizes these
costs and revenues over the 8-year period, 1986-93.
Table 4
1: Average U.S. Cotton Production Costs
and Revenues, 1986-93
(In 1993 dollars per pound)
Producti Governme Mark
on costs Market nt Total et Total
---------- -------- ------ -------- ------ ---- ------
Short-run $.48 $.66 $.25 $.91 138% 190%
Total $.78 $.66 $.25 $.91 85% 117%
(Long-
run)
------------------------------------------------------------
Source: GAO analysis of USDA data.
Short-run production costs, both variable and fixed, represent money
paid out by producers each crop year to grow, harvest, and market
cotton. Variable costs--seed, fertilizer, fuel, irrigation, hired
labor, chemicals, and other inputs--depend upon farm operation
management and practices, operation size, and input quantities and
prices. Fixed production costs, including general farm overhead,
taxes, and insurance, are allocated to each crop on the basis of its
relative share of total production value and also must be paid in the
short-run, or each year, if a producer is to stay in business.
To estimate the amount of revenue average producers must receive to
stay in business over the long run, USDA's Economic Research Service
combines short-run production costs with its estimates of the amount
of profit or returns average producers should earn on each cotton
production asset, such as capital, land, equipment, and unpaid labor.
This total estimate is also known as the long-run cost.
The cotton program benefits producers significantly and has allowed
some to depend on government payments. This is not to say that all
producers would be similarly affected in the absence of a cotton
program. Producers told us that some would not survive as cotton
farmers without the program. They suggested that producers who are
heavily financed by debt loans for land and machinery, have
consistently low yields and high cost, and/or who refuse to modernize
will not survive without the program. However, the number of
producers who would cease cotton farming is difficult to measure
without looking at individual producers' financial records and
assessing their operational efficiency.
ESTIMATED PERCENTAGE OF
PRODUCERS COVERING
PRODUCTION COSTS FROM MARKET
PRICES ALONE DURING 1993
-------------------------------------------------------- Chapter 4:1.1
The estimated percentage of producers who were able to cover
short-run and long-run production costs from market prices alone are
shown in tables 4.2 and 4.3. For example, table 4.2 shows that 77
percent of producers, who grew 90 percent of the cotton, had a
short-run production cost of 58 cents per pound, the average market
price for cotton in 1993.
Table 4.2
Range of the Estimated Short-run
Production Cost Under the Cotton Program
and Producers at or Below That Cost,
1993
Percent of
Short-run producers Percent of total
production at or below that production at or
cost per pound cost below that cost
-------------------- ------------------ ------------------
$.40 29 42
.47 50\a 67
.50 58 75
.58\b 77 90
.60 81 93
.70 93 98
.80 97 99
.90 and higher 100 100
------------------------------------------------------------
\a Median.
\b Average domestic price for 1993 was $.58 per pound.
Source: GAO analysis of USDA data. Estimates are based on each
producer's 1981-85 historical average yield and the applicable
regional average production cost.
Table 4.3 shows, for 1993, a range of total production costs with
corresponding percentages of producers and production at or below
those costs. For example, 27 percent of producers, who grew about 39
percent of the cotton, had a total production cost of 58 cents per
pound, the average price for cotton in 1993.
Table 4.3
Range of the Estimated Total Production
Cost Under the Cotton Program and
Producers at or Below That Cost, 1993
Percent of
producers Percent of total
Total production at or below that production at or
cost per pound cost below that cost
-------------------- ------------------ ------------------
$.40 3 4
.50 14 19
.58\a 27 39
.60 31 44
.70 50\b 67
.78 64 80
.80 67 83
.90 and higher 100 100
------------------------------------------------------------
\a Average domestic market price for 1993 was $.58 per pound.
\b Median.
Source: GAO analysis of USDA data. Estimates are based on each
producer's 1981-85 historical average yield and the applicable
regional average production cost.
In 1990, USDA last estimated the percentage distribution of cotton
farms and production by per pound variable cash cost and total
long-run economic cost. USDA's estimate was based on its 1987
national production costs and returns survey data. We found our
percentage distribution of cotton producers and production data and
returns to be similar to USDA's estimates. For example, USDA
reported that about two-thirds of the 1987 cotton crop was grown at
long-run economic production costs below the then target price of 79
cents a pound. In our analysis, we estimated that about 50 percent
of the producers had the potential to grow over two-thirds of the
1993 cotton crop at a long-run economic cost of 70 cents a pound, or
about 2 cents less than the current legislated target price. USDA
officials said that they had not officially updated their
distribution estimates for unit costs since they first reported them
in 1990.
Our estimates of production costs in tables 4.2 and 4.3 are
conservative because we were required to use USDA's 10-year-old yield
data, called "program yields," to estimate costs for each farm. Had
we been able to use current actual yields, which are higher, costs
per pound would have been lower. Since 1985, yields have increased
considerably because of increased irrigation, USDA's program to
eradicate the boll weevil, technology, and other improvements.
Because the program increases producers' revenues, it also raises the
total costs of production. Total costs include a value for the
earning capacity of land. The value of the land is determined not
only by what it earned in the previous year but also by what it is
expected to earn in the future. Since the cotton program
significantly raises the returns to farmers, the program increases
the value of the land, as well as other costs of operation.
PROGRAM HAS GENERATED TROUBLING
ECONOMIC CONSEQUENCES
---------------------------------------------------------- Chapter 4:2
As currently designed, the cotton program results in two troubling
economic consequences for producers and society. First, between 1986
and 1993, the program created a situation in which cotton producers
were better off when prices were low. This occurred primarily
because at low price levels during this period, farmers were eligible
for an array of government payments, including deficiency payments,
marketing loan gains, and loan deficiency payments, that more than
made up for the revenues lost due to the low price. As prices rose,
producers were no longer eligible to receive these payments.
As a second consequence, program features such as acreage control
resulted in economic inefficiencies as land was withdrawn from
production. These net economic losses to society, sometimes called
social welfare losses, averaged an estimated $738 million a year from
crop year 1986 through 1993.
PRODUCERS GAIN MORE WHEN
PRICES ARE LOW
-------------------------------------------------------- Chapter 4:2.1
The cotton program is meant to protect farmers' income during periods
of low prices. However, program benefits from 1986 through 1993 were
so generous that they more than offset lower prices. As shown in
figure 4.1, in years when prices were low, farmers received higher
total revenue per pound on their payment acres.
Figure 4.1: Relationship
Between Producer Revenues and
Price, 1986-93
(See figure in printed
edition.)
Source: GAO analysis of USDA data.
This condition occurs when domestic and world prices are such that
producers receive both marketing loan gains and deficiency payments.
When these amounts are added to the market price, the total is more
than the legislatively set target price. Under current program
requirements, the marketing loan gain is not incorporated into the
deficiency payment calculations. Figure 4.2 illustrates how farmers'
revenues were affected in 1992 when market prices were low and
compares this situation with 1990 when market prices were high.
Figure 4.2: Prices and
Producer Receipts in a Low- and
High-Price Year, 1992 and 1990
(price per pound in nominal
dollars)
(See figure in printed
edition.)
\a Calendar year market price.
\b The adjusted world price weighted by the quantity of loans
redeemed.
PROGRAM CAUSES ECONOMIC
INEFFICIENCIES
-------------------------------------------------------- Chapter 4:2.2
Although the cost of the cotton program to the federal government
averaged about $1.5 billion for crop years 1986 through 1993, we
estimate that cotton producers' annual economic benefits were only
about $754 million, or 51 percent of these costs. The $738 million
per year, or 49 percent, difference between the program's cost and
the producers' benefits represents a net loss to society (sometimes
called "social welfare loss"), as shown in table 4.4.
Table 4.4
Economic Impact of the Cotton Program,
Crop Years 1986-93
(In millions of 1993 dollars)
Total
governme Producer Social
nt and s' gain welfare
net Producer as loss as
cotton s' net percent Social percent
buyer economic of total welfar of total
Crop year costs gain cost e loss costs
------------ -------- -------- -------- ------ --------
1986 $2,364 $ 688 29% $1,676 71%
1987 1,680 625 37 1,055 63
1988 1,741 1,405 81 336 19
1989 981 522 53 460 47
1990 489 162 33 327 67
1991 1,251 982 79 269 21
1992 1,761 887 50 875 50
1993 1,676 766 46 910 54
Average $1,493 $ 754 51% $ 738 49%
------------------------------------------------------------
Note: Totals and averages may not add due to rounding.
Source: GAO analysis of USDA data.
This social welfare loss measures the inefficiencies that result
primarily from aspects of the program that keep land from being used
productively--such as the acreage reduction program, 50/92, and the
portion of flex acres that remain idle. (Although land has been left
idle under 50/92 and flex acres, producers are not required to leave
this land idle to receive program benefits as they are in the case of
the acreage reduction program.) Because of these inefficiencies,
producers receive economic benefits worth less than a dollar for
every dollar given up by the nation. Figure 4.3 shows the
relationship between the social welfare loss and idled land.
Figure 4.3: Social Welfare
Losses and Idled Land (in 1993
dollars)
(See figure in printed
edition.)
Source: GAO analysis of USDA data.
Idled acreage reduces producers' opportunities to earn additional
revenue and in some years increases cotton buyers' costs because
reduced supplies lead to higher prices. In addition, by providing
program benefits, the government is using resources to pay producers
not to produce on the idled acres. The magnitude of the social
welfare loss derives from (1) the number of idled acres and (2)
government costs, in terms of program benefits, that the government
incurs to induce producers to leave those acres idle. Additional
losses accrue through government stock-holding
activities--particularly the release of large stocks at prices less
than the government paid for them, as occurred in 1986.
The largest social welfare losses occurred in 1986 because of two
factors. First, the number of idled acres peaked that year. Second,
the social welfare loss increased because the government released
stocks, which had been accumulated in previous years as a result of
the program, for prices less than it paid for them.\1
The number of idled acres and social welfare loss has generally
declined since 1986. In 1992 and 1993, however, social welfare loss
increased because of increases in program benefits, particularly
through the marketing loan provision and step 2 payments.
--------------------
\1 Social welfare losses from the sale of CCC stocks would have been
even greater had they not been partially offset by the gains that
domestic cotton buyers received through the purchase of cotton at the
reduced price.
AGENCY COMMENTS AND OUR
EVALUATION
---------------------------------------------------------- Chapter 4:3
In its comments on a draft of this report, USDA questioned (1) our
use of averages in table 4.1 to analyze cost and revenue data over
the period 1986-93, (2) our finding that producers can receive higher
returns when market prices are low, and (3) our application of an
economic model used to determine the social welfare effects--economic
gains and losses--of the cotton program.
First, USDA believes that the use of 8-year average revenue and cost
figures are of limited value in explaining a real profit/loss result
in a given year because each cost estimate is based on a yield for a
specific year and each revenue estimate is based on an assumed price
for a given year.
We believe our use of averages to summarize cotton producers' costs
and revenues is a reasonable methodology that has explanatory value.
Averaging is particularly relevant in this case, since we are trying
to assess the economic condition of cotton producers in both the long
and short term. For example, our analysis for each year during
1986-93 showed that in the short run, revenue from market prices
alone as a percent of cost ranged from 107 to 176 percent. In the
long run, revenue from market prices alone as a percent of cost
ranged from 66 to 103 percent. Thus, while producers may sustain
variations in returns from year to year, their long-term condition is
determined by the average across years.
Second, USDA states that while some farmers may be better off as
prices decline, there are many farmers who, by employing futures and
options markets, would be able to earn nearly as much income from a
higher market through timely trading as they would in a lower market
through government payments. Nevertheless, USDA does agree that
there is a potential for a producer to realize a higher return in a
year with lower prices.
We believe that USDA's comments on producers' use of futures markets
are not relevant to whether the cotton program, as currently
designed, allows producers to receive higher returns from the
government when market prices are low. The fact that some farmers
may take actions on their own by using private market mechanisms,
such as futures markets, is not relevant in evaluating whether the
cotton program creates unreasonable economic incentives.
Finally, USDA notes that the social welfare loss measured by our
model is determined, in part, by program participants' idling part of
their land. USDA does not appear to object to our use of acres idled
as a result of the acreage reduction program. However, USDA
disagrees with our model's use of idled acres voluntarily idled under
flex acres and 50/85. (The 50/85 program was a revision to the 50/92
program, which was in place during the period of our review.) USDA's
comments also include a discussion of possible impacts of eliminating
the cotton program on the basis of cost data contained in tables 4.2
and 4.3. USDA suggests that although supply may expand without the
cotton program as indicated by our model, it would not be long before
farmers leave the business.
We believe that we have correctly applied our theoretical model. Our
model is well tested, has withstood close internal and external
scrutiny, and has proved to be an effective analytical tool to
evaluate the economic effects of commodity programs. First, our use
of the model and our interpretation of its results were reviewed by
Dr. Bruce L. Gardner, who developed the model. He is currently a
professor at the University of Maryland and formerly was a USDA
assistant secretary. Dr. Gardner agrees that we have properly
applied the model to the cotton program. Second, we have used this
same model in essentially the same way to evaluate the rice and wheat
commodity programs, and the model and results were described in
detail in published reports.\2 Third, we recognize that our model,
like any other model, is a simplification of the real world and has
limitations. However, this chapter and appendix I clearly set forth
how the model was constructed, relevant variables, and any
limitations of the model. Fourth, the model is designed to evaluate
the impact of a commodity program on the economic welfare of society
as a whole--not any particular group--by measuring the inefficiencies
that result from aspects of the cotton program that keep land from
being used productively. For example, although farmers may respond
to program provisions by making certain decisions that benefit them,
our model looks at whether these decisions provide an overall benefit
or cost to society.
Regarding our use of voluntarily idled acres under flex acres and
50/92, we clearly explain in this chapter that producers are not
required to leave land idle under flex acres and 50/92 to receive
program benefits. Also, the idled acres used in our model, whether
idled because of program requirements or voluntarily, were idled
under provisions of the cotton program. From the standpoint of
society, the reasons cotton producers decide to idle acres are not
relevant to whether economic inefficiency results. The key point is
that, as a result of the cotton program, productive land of potential
benefit to society was idled.
USDA suggests the possibility that although cotton supply would
initially expand without the program, it would not be long before
farmers would leave the business because they could not meet their
total costs from the market prices alone, as shown by the long-run
cost data in table 4.3. Our report does recognize in chapter 6 that
there will be negative consequences for some farmers if the program
is reduced or eliminated. We should also point out that where
turnover occurs among producers, some producers leaving the business
could be replaced by those with lower costs.
--------------------
\2 Rice Program: Government Support Needs to Be Reassessed
(GAO/RCED-94-88, May 26, 1994) and Wheat Commodity Program: Impact
on Producers' Income (GAO/RCED-93-175BR, Sept. 8, 1993.
FEDERAL EFFORTS TO ENHANCE COTTON
EXPORTS ARE COSTLY
============================================================ Chapter 5
The cotton program's provisions to maintain and expand exports when
world prices are low--the marketing loan and step 2--have not
succeeded despite government expenditures of about $3 billion since
1986. Even with these provisions, U.S. cotton prices remained
significantly higher than world prices, and U.S. exports fell when
world prices were low. This situation occurred because the marketing
loan and step 2 provisions could not counteract the effect of other
cotton program provisions that keep U.S. prices higher than world
prices. In addition to expenditures for the marketing loan and step
2, USDA spent $428 million under other agricultural programs for
cotton export promotion between 1986 and 1994.
Although exports are generally beneficial to a nation, this has not
been the case with cotton, which has continually sold at a loss.
Under the cotton program, exports have regularly been sold for less
than the combined costs of government expenditures and cotton
production.
In the future under NAFTA and GATT, the United States may find it
difficult to enhance exports, support domestic prices at higher than
world levels, and maintain other aspects of the present cotton
program. Under these trade agreements, the U.S. cotton program will
be pushed toward a greater market orientation in response to the
gradual elimination of trade barriers and increased foreign
competition, particularly in textiles.
COSTLY EXPORT PROGRAM
PROVISIONS HAVE NOT
COUNTERACTED EFFECTS OF OTHER
PROVISIONS ON U.S. PRICES
---------------------------------------------------------- Chapter 5:1
While the marketing loan and step 2 provisions were used to try to
support exports by making U.S. cotton available at lower prices on
world markets, other cotton program provisions worked to keep U.S.
prices higher, and, in fact, exports dropped. Furthermore, exports
fell in 1988, 1991, and 1992 even though USDA expended an additional
$428 million for export promotion and market development in foreign
countries from 1986 through 1994.
MARKETING LOAN AND STEP 2
PROVISIONS HAVE NOT
PREVENTED PERIODIC DECLINES
IN EXPORTS
-------------------------------------------------------- Chapter 5:1.1
During the 3 years when world prices were low--1988, 1991, and
1992--U.S. cotton exports declined significantly: by 222 million
pounds, or 7 percent, in 1988; by 500 million pounds, or 14 percent,
in 1991; and by 700 million pounds, or 23 percent, in 1992. For
these 3 years, the U.S. share of the world export market dropped
from 22 to 18 percent, 26 to 24 percent, and 24 to 20 percent,
respectively, as shown in figure 5.1.
Figure 5.1: Export Volume and
Market Share, 1986-93
(See figure in printed
edition.)
Source: GAO analysis of USDA data.
During 1988, 1991, and 1992, the adjusted world price was at or below
the U.S. loan rate and below the U.S. domestic price.\1 This
situation activated the marketing loan at a cost of $1.4 billion and
step 2 at a cost of $357 million. The cost for these 3 crop years
represented about 60 percent of the total $3 billion in expenditures
for the marketing loan and step 2 provisions between 1986 and 1993.
The price of U.S. cotton remained above the adjusted world price
during 1986-93. (See fig. 5.2.) This situation does not appear to
affect U.S. cotton exports when world prices are relatively high,
indicating a tight world supply. However, when world prices drop to
or below the loan rate, as they did in 1988, 1991, and 1992, U.S.
exports decline. The marketing loan and step 2 provisions were
intended to alleviate this situation by making U.S. cotton more
price-competitive. By allowing producers to redeem cotton from loan
at the adjusted world price, the marketing loan provision was
expected to encourage producers to sell cotton on the market because
they would gain the difference between the loan rate and the adjusted
world price. In addition, these sales were expected to lower the
U.S. price for cotton, eliminating the loan rate as the floor price
and bringing U.S. prices in line with world prices. However, as
figure 5.2 shows, the U.S. price (also referred to as the U.S. spot
price) remained above the loan rate even when the adjusted world
price approached or fell below the loan rate in 1988 and 1991-92.\2
Figure 5.2: Relationship
Between Adjusted World Price,
U.S. Price, and Loan Rate,
1986-93 (in 1993 cents/pound)
(See figure in printed
edition.)
Source: GAO analysis of USDA data.
In the 1990 Farm Bill, the Congress enacted the step 2 provision,
which compensates domestic buyers and exporters for the higher price
of U.S. cotton. However, cotton exports and market share still
declined in 1991 and in 1992, when the adjusted world price was below
the loan rate. These periodic declines in U.S. cotton exports
indicate that the marketing loan and step 2 provisions have been
unable to overcome other provisions of the cotton program that keep
U.S. prices high.
During our review, USDA officials and cotton industry representatives
pointed out that the breakup of the Soviet Union was an important
factor in explaining declining U.S. cotton exports in 1991 and 1992.
According to these officials and representatives, Uzbekistan and
other cotton-producing Central Asian countries in the former Soviet
Union, which had traditionally sold their cotton to textile mills in
Russia, now offered their cotton for sale on the world market in
order to obtain needed hard currency. Because of market uncertainty
regarding the reliability of these Central Asian countries in meeting
contract commitments, their cotton could be sold only at heavily
discounted prices. The end result was that world cotton prices
dropped during this period.
The breakup of the Soviet Union was a factor in the decline in world
price during the 1991-92 period; however, such world events are not
relevant in assessing whether the marketing loan and step 2
provisions are effective. These provisions were designed
specifically to help make U.S. cotton competitive by bringing the
U.S. price more in line with a lower world price--whatever the
reason for the lower world price. As shown in figure 5.2, the
marketing loan provision, in particular, has been unable to bring
down the U.S. price to the lower world price. In fact, figure 5.2
shows that the gap between the U.S. price and the adjusted world
price is actually widening.
--------------------
\1 The U.S. domestic price is also referred to as the spot price.
Data on U.S. spot prices are collected and published by the
Agricultural Marketing Service.
\2 U.S. prices were below the loan rate in crop year 1986 during the
transition to the marketing loan program. During this transition
year, the government released to market the stocks it had accumulated
during previous years. In addition, loans were repaid under plan A
rather than the present plan B. Under plan A, the repayment rate was
fixed at 80 percent of the loan rate regardless of how high market
prices were. Under the present system--plan B--the loan repayment
rate equals the adjusted world price or the loan rate, whichever is
lower.
OTHER PROGRAM PROVISIONS
KEEP U.S. PRICES HIGHER
THAN THE ADJUSTED WORLD
PRICE
-------------------------------------------------------- Chapter 5:1.2
Five features of the cotton program reduce producers' incentive to
sell cotton to the market and thereby keep the U.S. price above the
world price: import restrictions, production restrictions, the loan
rate, loan extension, and free storage.
Import restrictions. Because domestic textile mills are largely
prohibited from importing cotton, U.S. prices are insulated
from world competition. Therefore, producers have a captive
domestic market and do not have to compete against foreign
producers selling cotton at lower world prices. This makes it
possible for other program features to operate.
Production restrictions. To the extent that cotton production is
reduced through such program provisions as acreage set-asides,
50/92, and idled flex acres, prices will be higher because less
cotton will be available to the market.
Loan rate. By guaranteeing a minimum price to producers, the loan
rate enables them to keep cotton off the market unless they are
offered a price higher than the loan rate. In effect, despite
the introduction of the marketing loan, the loan rate still acts
as the floor price for the U.S. market, as figure 5.2 shows.
Loan extension. The 8-month extension of the basic 10-month
nonrecourse loan makes it easier for producers to be selective
in the price they accept for their cotton. In total, producers
may have up to 18 months in which to sell their cotton.
Free storage. Because producers can receive free storage for the
first 10 months of their nonrecourse loan when the adjusted
world price nears or drops below the loan rate, producers can
keep cotton off the market at no cost. USDA does not provide
free storage for any other commodity.
With these provisions, U.S. cotton producers can hold their cotton
under loan until pricing conditions are favorable or forfeit cotton
to the government. Therefore, to get cotton to the market, cotton
buyers (domestic textile mills and exporters) have to overcome the
disincentives created under the domestic program by paying premiums,
known as equities.\3 For example, as shown in table 5.1, a producer
put cotton under loan and received the loan rate of 52 cents per
pound from USDA. Cotton buyers offered the producer 11 cents per
pound in equity payments for the right to redeem cotton held under
loan. The producer therefore received 63 cents per pound for the
cotton--52 cents from the loan rate and 11 cents in equity. Using
the marketing loan, the buyer then redeemed the cotton at 43 cents
per pound, which was the adjusted world price when the cotton was
redeemed. Therefore, the cotton buyer's total purchase price was 54
cents per pound--43 cents to redeem the cotton from loan and 11 cents
for the equity payment to the producer.
Table 5.1
Illustration of the Effect of Equity
Sale on Payments and Receipts of
Producer, Government, and Buyer
(Cents/pound)
Governme
nt
payments
Producer and Buyer
Producer and buyer actions receipts receipts payments
------------------------------ -------- -------- --------
Producer puts cotton under 52 cents (52 Not
loan cents) applicab
le
Buyer pays equity for right to 11 cents Not (11
redeem cotton applicab cents)
le
Buyer redeems cotton from loan Not 43 cents (43
at adjusted world price applicab cents)
le
============================================================
Net result 63 cents (9 (54
cents) cents)
------------------------------------------------------------
As shown in table 5.1, although buyers redeemed cotton from the
government at the adjusted world price of 43 cents, the additional 11
cents equity payment resulted in a domestic price of 54 cents. Our
analysis found that these equity payments to producers kept the
domestic price above the adjusted world price. As shown in figure
5.3, equity payments are closely related to the amount by which the
domestic price exceeded the adjusted world price between August 1990
and July 1994.
Figure 5.3: Role of Equities
in Explaining the Gap Between
the U.S. Price and the
Adjusted World Price (in 1993
cents/pound)
(See figure in printed
edition.)
Source: GAO's analysis was based on unpublished equity data
collected by USDA's National Agricultural Statistics Service as part
of its monthly survey of market prices received by farmers.
Domestic cotton buyers are able to pay equities to producers because
the government compensates them through several components of the
cotton program. First, buyers receive step 2 payments. Second, the
government's provision for free storage under the marketing loan
represents a cash value associated with cotton for which cotton
buyers are willing to pay. Third, under the marketing loan program,
buyers receive a price advantage as a result of the method used to
calculate the adjusted world price. This advantage occurs because
the adjusted world price is calculated as if all cotton were
transported to Europe at a cost of 12 cents per pound. Domestic
buyers, however, actually incur only the 5-cent cost of transporting
cotton to domestic mills. Therefore, domestic buyers gain a benefit
of 7 cents per pound on the value of the cotton they purchase.
--------------------
\3 Under the cotton program, after taking out a loan, a producer has
three marketing options: the producer may (1) allow the government
to keep the cotton, called "forfeiting"; (2) repay the loan and
market the cotton; or (3) sell the right to redeem the cotton from
loan to a merchant or domestic mill. The value of the right to
redeem cotton from loan is typically referred to as "buying an
equity." In an equity transaction, the producer retains beneficial
interest (ownership) in the cotton until it is actually redeemed from
loan by the buyer.
OTHER AGRICULTURAL PROGRAMS
PROMOTE COTTON EXPORTS
-------------------------------------------------------- Chapter 5:1.3
In addition to the approximately $3 billion spent since 1986 on
marketing loan and step 2 payments, USDA spent another $428 million
to enhance cotton exports through its export promotion and market
development programs for agricultural commodities. (See table 5.2.)
Despite these programs, cotton exports periodically declined. While
each export promotion and market development program has specific
objectives, such as countering the actions of foreign competitors who
subsidize their own exports, all of these programs are used to
support the development of commercial markets in foreign countries.
These export promotion programs include Foreign Market Development,
the Market Promotion Program, Public Law 480, and General Sales
Manager 102 and 103 programs. The Foreign Market Development and the
Market Promotion Program promote exports in specified markets.
Public Law 480 has the multiple objectives of developing and
expanding U.S. agriculture export markets, encouraging economic
development, providing humanitarian assistance, and promoting U.S.
foreign policy. The General Sales Manager programs guarantee
repayment of private short- and intermediate-term credit to potential
foreign customers who cannot otherwise obtain commercial credit.
General Sales Manager 103, added in 1985, created an alternative
program with a longer repayment period than that available under
General Sales Manager 102. These credit guarantee programs incur
losses when loans are defaulted on. USDA has guaranteed loans of
about $2.6 billion to foreign cotton buyers from fiscal years 1986
through 1994 and incurred estimated losses of about $154 million from
defaulted loans.
Table 5.2
Export Program Costs, Fiscal Years 1986-
94
(In millions of 1993 dollars)
General Market Foreign
Sales Promotio Market
Fiscal Manager Public n Developm Total
year defaults Law 480 Program ent cost
---------- -------- -------- -------- -------- --------
1986 $17 $15 $8 $3 $42
1987 16 16 0 0 42
1988 15 12 9 1 37
1989 15 17 19 1 52
1990 15 25 15 2 57
1991 19 0 13 2 34
1992 19 26 16 1 62
1993 19 30 14 2 65
1994 19 9 8 1 37
============================================================
Total $154 $160 $102 $12 $428
------------------------------------------------------------
Source: GAO analysis of USDA data.
COTTON EXPORTS ARE SOLD AT A
LOSS
---------------------------------------------------------- Chapter 5:2
While exports are generally beneficial, this has not been the case
for the cotton program. Although the cotton industry does benefit
from selling cotton on the world market, it has been able to do so
largely because of government assistance. When the benefits of
exporting cotton are analyzed from a national viewpoint and both the
producers' and the government's costs are combined, cotton exports
occur at a loss.
From a national standpoint, government payments combined with
production costs are higher than the adjusted world price. As a
result, the United States sells cotton on the world market for less
than its cost. For example, to be competitive in world markets in
1993, U.S. exporters would have had to sell cotton at the adjusted
world price of 56 cents per pound; however, U.S. cotton cost 90
cents per pound--66 cents for the cost of production and 24 cents in
federal payments.
As figure 5.4 shows, the cost of production plus government payments
have been higher than the world price in every year from 1986 to
1993. As a result, the nation has exported cotton at a loss in each
of those years.
Figure 5.4: Comparison Between
Total Cost Plus Government
Payments and Adjusted World
Price, Crop Years 1986-93 (in
1993 cents/pound)
(See figure in printed
edition.)
Source: GAO analysis of USDA data.
Furthermore, the value of cotton exports has been declining as the
real price of cotton has fallen. Cotton prices, in real terms, have
declined generally from 1981 to 1993, falling by more than 30
percent. As a result, the value of U.S. exports has declined even
though export volume has shown a slight upward trend. For example,
even though the volume of cotton exports in 1993 was up by 2 percent
from its 1981 level, the value of these exports fell by about 30
percent, from about $2.7 billion in 1981 to less than $1.9 billion in
1993.
TRADE AGREEMENTS WILL PUSH
COTTON INDUSTRY TOWARD GREATER
MARKET ORIENTATION
---------------------------------------------------------- Chapter 5:3
The gradual elimination of trade barriers brought by provisions of
NAFTA and GATT will open the United States to increased market
opportunities and challenges from foreign competition and thereby
push the U.S. cotton industry toward a greater market orientation.
These agreements convert quotas into tariffs, reduce tariffs over
time, and allow greater market access. In an environment of open,
competitive markets, it may be increasingly difficult to maintain
price supports and other aspects of the present U.S. cotton program.
The effect of these agreements upon the cotton industry will depend
upon changes in the raw cotton trade as well as the trade in cotton
textiles and apparel. Although the precise impact is unclear, the
impact on the raw cotton trade is expected to be small in comparison
to the potential impact on cotton textiles and apparel.
THE IMPACT OF REDUCING
BARRIERS ON RAW COTTON TRADE
ARE EXPECTED TO BE SMALL
-------------------------------------------------------- Chapter 5:3.1
Both the NAFTA and GATT agreements are phased in over a number of
years, and the impact on the raw cotton trade will be gradual.
Little change occurred during NAFTA's first year, and impacts are not
likely to be significant for the next few years. The direct effects
of GATT on the raw cotton trade are also expected to be modest.
NAFTA IMPACT
-------------------------------------------------------- Chapter 5:3.2
Under NAFTA, import restrictions shaping present cotton trade flows
will be gradually phased out by the United States and Mexico over 5
to 10 years. Initially, the impact on trade is expected to be small.
In the longer term, the impact will depend on such factors as future
Mexican agricultural policy and the relative competitiveness of the
Mexican and U.S. textile and apparel industries.
Mexico will gradually phase out its 10-percent tariff on cotton over
the next 10 years. Likewise, U.S. section 22 import quotas on
Mexican cotton--currently about 9 million pounds--will be replaced by
a tariff-rate quota\4 that will remain in place during a 10-year
transition period. This duty-free quota, set at 22 million pounds,
is to grow by 3 percent compounded annually over the transition
period. Any additional cotton imported will face a tariff of 26
percent of the import price. This 26-percent tariff will be phased
out over 10 years.
NAFTA has been in effect since January 1, 1994. Exports from January
to October 1994 were about the same as exports during the same period
in 1993. Mexican consumption of cotton has increased; however, this
increase has been met by additional domestic production rather than
by imports from the United States. This is because farm acreage in
Mexico shifted from corn production to cotton production as a result
of the changes in agricultural policy that the Mexican government
made in 1993. U.S. imports of cotton from Mexico continue to be
minimal. Since Canada does not grow cotton and has only a relatively
small textile and apparel industry, NAFTA is expected to have little
effect on cotton trade between the United States and Canada.
--------------------
\4 Under a tariff-rate quota, a certain amount of the product is
imported free or nearly free of duty, while a heavier import duty is
levied on additional quantities.
GATT IMPACT
-------------------------------------------------------- Chapter 5:3.3
As in the case of NAFTA, the direct effects of GATT on raw cotton
trade appear to be modest over the 6-year implementation period.
GATT does not require any reduction in the domestic support of the
U.S. cotton program. GATT, however, will result in a reduction of
section 22 import quotas and, as a result, the potential exists for
U.S. imports to increase. The quota under section 22 was about 60
million pounds, or less than 2 percent of domestic consumption during
the GATT base period--1986 to 1988.\5 Under GATT, the United States
will initially implement a tariff-rate quota for cotton of about 114
million pounds, or 3 percent of the base period's domestic
consumption. By 2001, this import quota will increase to 191 million
pounds, or 5 percent of the base period's domestic consumption.
Imports under GATT are not likely to exceed the 3 to 5 percent level
of domestic consumption during the base period because the tariff on
additional cotton imports will be relatively high. Under the
tariff-rate quota--the 3 to 5 percent of GATT base-period domestic
consumption--imports will be subject to a maximum tariff of 2 cents
per pound. Any additional cotton will be imported at a higher tariff
of 16.7 cents per pound. This tariff will be reduced by 15 percent
in equal annual installments over 6 years, beginning in 1995, to a
tariff of about 14.2 cents per pound in 2001.
GATT's impacts on exports, like those for imports, are expected to be
modest. Increased exports under GATT are expected to result from
increased demand for cotton caused by trade liberalization and
resulting increases in world income. Studies by the Economic
Research Service and the International Trade Commission have
projected that GATT will cause a small increase in cotton exports.
Specifically, the Economic Research Service projected that cotton
exports would increase by 7 to 14 percent by 2005. The International
Trade Commission projected a negligible increase.
We discussed the Economic Research Service estimates of GATT impacts
on agricultural products in a July 1994 report.\6 We pointed out that
GATT's projected impacts on U.S. agriculture are based on
assumptions about future events, which are subject to substantial
uncertainty. For example, uncertainty exists concerning assumptions
about (1) what projections are for world income growth, (2) how
governments of other countries would implement GATT requirements, and
(3) how agricultural producers in the United States and other
countries would respond to the expected changes in agricultural
policies.
--------------------
\5 In the past, these quotas have not been filled. According to
representatives of the U.S. textile industry, domestic textile mills
chose not to import cotton because the annual import quota equates to
less than 1 week's consumption by their mills. According to these
representatives, it is not practical for domestic mills to import
such a small quantity of cotton.
\6 General Agreement on Tariffs and Trade: Agriculture Department's
Projected Benefits Are Subject to Some Uncertainty (GAO/RCED-94-272,
July 22, 1994).
LIBERALIZED TEXTILE AND
APPAREL TRADE WILL AFFECT
COTTON INDUSTRY
-------------------------------------------------------- Chapter 5:3.4
NAFTA and GATT will reduce the highly protective restrictions on the
cotton textile and apparel trade. These changes will provide
increased opportunities for world trade in textile and apparel and
for the sectors that supply those industries, including cotton.
Prior to NAFTA and GATT, most U.S. imports of cotton textiles and
apparel were subject to restrictive quotas associated with the
Multi-fiber Arrangement.\7 For example, we found that at any one
time, the United States maintained quotas covering about two-thirds
of U.S. textile and apparel imports. Reductions in these trade
barriers under NAFTA and GATT mean that markets and market prices
will play a greater role in shaping trade. These changes in textiles
and apparel trade will, in turn, affect trade in the raw cotton from
which textiles and apparel are produced--pushing the cotton program
toward greater market orientation.
It is expected that movements toward free trade will increase world
income, which would lead to increased world consumption of textiles
and apparel. This in turn, will increase the demand for fibers such
as cotton that are used in textile and apparel production. To the
extent that U.S. cotton is competitive on the world market, U.S.
producers could increase their exports of cotton. Domestic cotton
sales could also be affected by increasingly open trade in textiles
and apparel. U.S. textile and apparel producers will be subject to
growing competitive pressures from abroad as import barriers are
reduced and will have difficulty competing if the domestic price for
cotton is higher than the world price. The net effect of changes in
textiles and apparel on raw cotton trade is difficult to predict
because it is unclear, for example, how U.S. textile and apparel
producers will respond to the new market conditions, particularly
potential increases in textile and apparel imports. However they
respond, this new environment, in the long run, will make it
difficult to support cotton prices through the present cotton
program.
Under NAFTA, U.S. quotas on various textile and apparel imports from
Mexico will either terminate immediately or be phased out in three
stages over a 10-year period. Canada and Mexico, the two largest
single-country markets for U.S. textile exports, are phasing out
their tariffs on U.S. textiles under NAFTA. In the first year of
NAFTA, trade has grown in both directions as apparel components are
shipped to Mexico and goods that have had some further processing in
Mexico are in turn shipped to the United States. From January to
September 1994, the United States maintained an export surplus, in
value terms, with Mexico in yarn and fabric trade that just offset an
increase in the trade deficit in apparel and made-up goods. Under
GATT, the Uruguay Round Agreement on Textiles and Clothing, according
to the U.S. International Trade Commission\8 and GAO,\9 will have a
greater impact on the U.S. clothing and apparel sectors. GATT will
integrate textile and clothing into the agreement by phasing out
quotas under the Multi-fiber Arrangement and accelerating quota
growth rates for products not yet integrated into GATT. This process
is to occur over 10 years in three stages.
The International Trade Commission has estimated that over the long
run the United States will experience a sizable increase (over 15
percent) in apparel imports as a result of GATT. Increased import
penetration in this sector substitutes foreign-produced yarn and
fabric for domestic materials. As apparel imports increase, the
demand for domestic textile and cotton inputs decreases.
While the net effect of these forces are uncertain, competitive
pressures resulting from trade liberalization could make it difficult
to support domestic cotton prices at higher than world levels.
Therefore, while NAFTA and GATT do not formally require any changes
in U.S. internal support for cotton, market forces may in fact
demand some long-term changes in the present program.
--------------------
\7 The Arrangement Regarding International Trade in Textiles, known
as the Multi-fiber Arrangement, has governed world trade in textiles
and apparel since 1974. The Multi-fiber Arrangement allows
signatories to place quantitative limits, or quotas, on most imports
of textiles and apparel. These quotas are a departure from the
general principles of GATT in that they are bilateral rather than
multilateral and they lower restrictions for some countries rather
than all countries. The present Uruguay Round of GATT supersedes the
Multi-fiber Arrangement by setting a 10-year phaseout of these
bilateral quotas, resulting in a reduction in the trade restrictions
imposed by the Multi-fiber Arrangement.
\8 U.S. International Trade Commission. Potential Impact on the
U.S. Economy and Industries of the GATT Uruguay Round Agreements.
Investigation No. 332-353, June 1994.
\9 The General Agreement on Tariffs and Trade: Uruguay Round Final
Act Should Produce Overall U.S. Economic Gains (GAO/GGD-94-83B,
Volume 2, July 1994).
AGENCY COMMENTS AND OUR
EVALUATION
---------------------------------------------------------- Chapter 5:4
In commenting on a draft of this report, USDA raised two concerns
relating to our analysis of export costs and world prices. We found
that from a national viewpoint, the United States exports cotton at a
loss when both government payments and total (long-run) production
costs are taken into account. First, USDA said that, rather than
long-run production costs, short-run cost would be more appropriate
since the marginal cost of an additional bale shipped would be of
most interest. (In table 4.1, we reported that the average short-run
costs over 1986-93 were 48 cents and the long-run costs were 78
cents.) Second, USDA questioned whether we properly accounted for the
impact of step 2 and marketing loan provisions in measuring the loss
on export sales. USDA is concerned that we have included the costs
of step 2 and marketing loan payments in determining the government's
costs but have not properly reflected that these provisions may
result in lowering the price of cotton and result in sales that may
not produce losses.
We agree that short-run costs are appropriate to use if the question
is whether exporting cotton is profitable in the short-term, that is,
annually or for a few years. In other words, the United States might
be willing to take losses on cotton exports in one year if it
expected that exports would be profitable the next year. However,
the cotton program has created a situation in which these losses
continue over the longer term. Our analysis found that as a nation,
we sustained losses from 1986 through 1993. This is the context in
which the use of long-run costs is appropriate.
We believe our analysis properly incorporates the impact of marketing
loan and step 2 provisions in the measurement of losses through
export. Our analysis is based on three data elements: (1) the cost
of producing cotton in the United States; (2) government payments
made under the cotton program; and (3) the adjusted world price.
Step 2 and marketing loan costs are appropriately included in
government payments made under the program. Although these
provisions may also influence the adjusted world price on those
occasions when the U.S. northern European price is actually used in
the computation of the world price, the difference between the
adjusted world price and producer costs plus government subsidies
appropriately reflects losses from exports. The adjusted world price
is USDA's official measure of prevailing world prices and is used to
calculate step 2 and marketing loan payments under the cotton
program. To the extent that step 2 and marketing loan payments
contribute to lowering the adjusted world price, they also contribute
to increasing government costs, thereby further increasing losses.
CONCLUSIONS AND MATTERS FOR
CONGRESSIONAL CONSIDERATION
============================================================ Chapter 6
The cotton program has so evolved over the years that its provisions
work at cross-purposes in an attempt to meet conflicting goals. As a
result, the government spends money to do one thing and then spends
money to undo it. The program
protects producers' income by guaranteeing a minimum price to
encourage production, then pays farmers to idle land to reduce
production;
supports the price the producers get, then subsidizes buyers to
reduce their costs;
attempts to make cotton prices competitive and move cotton quickly
to market, yet allows farmers to hold their crop off the market
for up to 18 months (free of storage cost for 10 of those
months) and ultimately forfeit their cotton if prices do not
rise sufficiently; and
reduces payments to producers by lowering and freezing the target
price and introducing flex acres, then increases payments by
allowing producers to earn revenues in excess of the target
price from the marketing loan provision.
In an effort to accomplish all of these goals simultaneously, the
program has become complex and costly, concentrates benefits among a
small percentage of producers, and results in troubling economic
consequences. This program will be difficult to sustain in the era
of greater fiscal constraint and increased international competition
being imposed by NAFTA and GATT.
We recognize that if government support were reduced or eliminated,
some producers could not profitably remain in cotton farming. In
addition, because lower government support would cause declines in
land values, some producers and rural economies would be negatively
affected. Among the producers most adversely impacted would be those
who are heavily in debt for land or machinery.
MATTERS FOR CONGRESSIONAL
CONSIDERATION
---------------------------------------------------------- Chapter 6:1
The Congress may wish to consider whether benefits from the cotton
program are worth its costs and whether the program should be
continued. The severe economic conditions and many of the
motivations that led to the cotton program in the 1930s no longer
exist. If the program were eliminated, the Congress might want to
consider options to give producers and other affected parties time to
make adjustments in their investment decisions. The Congress could,
for example, reduce or phase out payments or outlays over a number of
years, perhaps over the life of the next farm bill. One way to cut
outlays would be for the Congress to include marketing loan gains in
the calculation of the deficiency payments and to resume the process
of reducing the target price that it began under the 1985 farm bill.
The gradual implementation of GATT and NAFTA requirements will give
the cotton industry time to make adjustments during the transition
from reliance on the government program to greater reliance on the
market. The elimination of trade barriers for both raw cotton and
cotton textiles will push the U.S. cotton industry and program
toward greater market orientation and reliance on market prices.
However, because the economic changes imposed by GATT and NAFTA will
be phased in over a number of years, the U.S. cotton industry will
have time to adjust to more competitive world markets.
AGENCY COMMENTS AND OUR
EVALUATION
---------------------------------------------------------- Chapter 6:2
In commenting on a draft of this report, USDA took issue with our
statement that "the economic situation that led to the cotton program
no longer exists." USDA indicated that such a statement questions the
need for a cotton program if the economic emergency and related
problems that generated a cotton program in the 1930s no longer
exist. USDA noted that some problems have not gone away, in
particular the continuing problem of trade protectionism and
subsidization by other cotton-producing countries.
We believe that the economic conditions affecting farming have
changed significantly since the 1930s. Nevertheless, we did not mean
to imply by our statement that every economic problem of the 1930s
has been solved, and we have qualified our statement accordingly to
say that "the severe economic conditions and many of the motivations
that led to the cotton program in the 1930s no longer exist." This
revised statement is consistent with the following views expressed by
the Council of Economic Advisers in its 1995 Annual Report:\1
"Today's agricultural commodity support programs are rooted in
landmark New Deal legislation that followed the agricultural
depressions of the 1920s and 1930s. These programs were
designed to sustain prices and incomes for producers of cotton,
milk, wheat, rice, corn, sugar, tobacco, peanuts, and other
crops. However, changing economic conditions and trends in
agriculture over the past half-century suggest that many of the
original motivations for farm programs no longer apply."
The Council cited the following changing economic conditions and
trends to support its views:
The farm sector is no longer a major factor in the macroeconomy.
Commodity programs were originally instruments of macroeconomic
policy as well as a means of sustaining farm families' incomes.
In the 1930s, farm households accounted for 25 percent of the
U.S. population and generated over 10 percent of the gross
domestic product. Today, these households account for less than
2 percent of the population and generate less than 2 percent of
the gross domestic product.
The average farm payment recipient is no longer poor. In the
1930s, per capita farm income was only one-third the per capita
income of the rest of the population. Commodity programs were
intended to reduce this disparity. Today, recipients of farm
program payments (about one-third of all farm operators) tend to
have higher incomes than the average American. Moreover,
two-thirds of program payments go to the largest 18 percent of
farms.
The number of farms has fallen by more than 60 percent since 1950,
while the size of the average farm has doubled. Moreover, 92
percent of farm households (according to the Bureau of the
Census definition) operate small farms and receive almost all
their income from off-farm sources and only a small share of
government farm program payments.
Farmers now can insure themselves against price declines. In the
early 1930s, farm incomes were at the mercy of year-to-year
fluctuations in farm prices. Commodity programs provided price
floors for agricultural producers, insuring them against adverse
price swings. The growth of futures and options markets now
lets farmers protect against short-term price declines without
the need for a government program.
USDA also cited the use of futures and options as a form of insurance
against price fluctuations in its comments on a draft of this report.
USDA stated, "Many cotton farmers today employ marketing strategies
which involve locking in prices through the futures and options
markets."
Regarding the trade issue USDA raised, we recognize that certain of
our competitors in the world market intervene in various ways in the
cotton market. Some of these actions put our cotton at a price
disadvantage in world markets. We believe that, ultimately, it is up
to the Congress to determine whether it is in the nation's best
interest to respond to these countries in kind by continuing to
subsidize our cotton industry. As the last 8 years have shown,
subsidizing the U.S. cotton industry to make it more competitive on
world markets has been very costly. For example, as discussed in
chapter 5, the marketing loan and step 2 provisions, which were
designed specifically to help make U.S. cotton competitive in world
markets, cost $3 billion between 1986 and 1993. Yet, when these
provisions were in effect, U.S. exports declined. Moreover, as a
nation we have been exporting cotton at a loss when both production
costs and government payments are taken into account.
--------------------
\1 Economic Report of the President. Together With the Annual Report
of the Council of Economic Advisers. Transmitted to the Congress.
Washington, D.C.: U.S. Government Printing Office, Feb. 1995.
GAO'S ECONOMIC WELFARE ANALYSIS OF
THE COTTON PROGRAM
=========================================================== Appendix I
This appendix discusses the economic welfare analysis we used to
measure the economic gains and losses of the U.S. cotton program we
have reported on. The first section of the appendix discusses the
methodology used to estimate the market price and the amount of
cotton that would be produced if the cotton program were not in place
(the "no-program" scenario). These estimates are called the
equilibrium price and quantity. The second section explains the
methodology used to measure welfare gains and losses to cotton
buyers, cotton producers, and taxpayers.\1 According to this
methodology, gains and losses are measured by using the estimated
equilibrium price and quantity as reference points against which we
estimated changes caused by the program in real income for market
participants. The third section presents the results derived from
the methodologies described in the first two sections.
This analysis shows that between 1986 and 1993, the average annual
cost of the cotton program to the government (taxpayers) and to
cotton buyers was $1.5 billion.\2 The majority of this cost was borne
by taxpayers. About 51 percent of this cost was transferred to
producers as income; the remaining proportion represented a loss in
social welfare (deadweight loss) resulting from inefficiencies in
production and consumption caused by the cotton program. For the
most part, these social welfare losses are attributed to the
opportunity cost associated with land left idle because of the
program.
--------------------
\1 In this analysis, "cotton buyers" refers to buyers at the first
processor stage because the analysis is based on prices for
unprocessed cotton.
\2 These estimates of economic gains and losses are in 1993 dollars.
METHODOLOGY FOR NO-PROGRAM
EQUILIBRIUM PRICE AND QUANTITY
--------------------------------------------------------- Appendix I:1
We used a methodology developed by Gardner (1989)\3 to determine the
price and quantity of cotton if there were no program that could be
compared to prices and quantities with the program in effect in order
to estimate the economic gains and losses from the cotton program.\4
We conducted the analysis for 1986 through 1993.\5 We chose these
years because they incorporated the major and most recent changes
made to the program by the 1985 and 1990 Farm Bills.
--------------------
\3 Bruce L. Gardner, "Gains and Losses from the Wheat Program,"
Department of Agricultural and Resource Economics, Working Paper
88-11, University of Maryland, 1989.
\4 This model, like most models used for welfare analyses, is a
static partial equilibrium model. It does not consider a movement to
worldwide free trade, nor does it consider a complete absence of
governmental intervention in agriculture (such as an absence of
disaster payments, research and development.)
\5 These years, 1986 to 1993, correspond to crop years 1986-87 to
1993-94, respectively, throughout the analysis.
GRAPHIC PRESENTATION OF THE
COTTON MODEL
------------------------------------------------------- Appendix I:1.1
The basic model used in the analysis is shown in figure I.1. It is a
simplified economic representation of how the cotton market
operates.\6 The demand curve, identified as D, shows the quantity of
cotton that cotton buyers will demand at each price. The supply
curve, identified as S, represents the no-program supply curve.\7
With no program in effect, the market clears at point e
(equilibrium). At this point, cotton buyers purchase and producers
sell Qe quantity of production at Pe price.\8
With the cotton program in effect, however, prices and quantities
diverge from equilibrium. Because the focus of the cotton program is
on producers' income and therefore supply, the program does not cause
a shift in the demand curve.\9 With the program in effect, cotton
buyers purchase what they want at the market price, just as they
would in the absence of the program. Therefore, the major impact of
the cotton program on demand is through its effect on price.
Figure I.1: The
Program/No-Program Cotton
Supply and Demand Curves
(See figure in printed
edition.)
Legend
S' = Supply curve with program acreage controls
S = No-program supply curve
D = Demand curve for cotton
Pt = Target price
PIP = Producer incentive price
Pm = 12-month-season average market price
Pe = No-program equilibrium price
Qd = Total quantity demanded
Qe = No-program equilibrium quantity
Qb = Quantity without acreage constraints at point B
Q2 = Quantity of cotton that would be consumed at target price
Q1 = Quantity of cotton that would be produced at target price
Qp = Quantity of production for which producers receive deficiency
payments under program yields that are frozen
Note: The shape of the supply curve S' is uncertain because it has
to account for participants entering and leaving the program in
response to expectations about price. However, this uncertainty does
not affect the calculations because they are based on point A, which
remains the same regardless of the shape of the curve. In addition,
to the extent that the program has caused higher profits, it may have
attracted more resources into the industry than otherwise would have
occurred. In any event, there will still be social welfare loss due
to the inappropriate allocation of resources.
Under the U.S. cotton program, the government supports producers'
income and, in so doing, causes a leftward shift of the supply curve
from S to S'. This shift in the supply curve results from the effect
of the acreage reduction program (ARP) on land use and consequently
on the quantities produced. In simplified terms, under the program
(in the absence of acreage restrictions), producers participating in
the program do not receive Pe; instead, they receive Pt, which is
much higher. At this higher price, producers would supply much more
cotton (Q1) than cotton buyers would purchase at that price (Q2). In
order to maintain the support price at its high level, the difference
between what farmers produced and cotton buyers bought would have to
be purchased (through the Commodity Credit Corporation--CCC--loan
program) and held in storage by the government, at taxpayers'
expense.\10
The government reduces the costs it would incur, as well as the
quantities of cotton put into storage, by restricting supply through
acreage controls. The government does this by requiring producers to
reduce the acreage on which they produce cotton by a specified amount
in order to be eligible to receive the support price on the remaining
amount. These acreage reductions have the effect of reducing the
quantity produced under the program to Qd.
During the period covered by this analysis, the total acreage left
idle under the program ranged from a high of 29 percent of the
complying base in 1986 to a low of 13 percent of the base in 1993.
During the first few years covered by the analysis, ARPs accounted
for the majority of the idled acres. Although ARPs were reduced in
subsequent years, other aspects of the program, particularly the
50/92 program and planting flexibility (flex acres), have provided
producers with incentives to leave land idle.\11
The government further reduces the costs associated with the cotton
program by limiting the quantity on which deficiency payments are
made. This is done by limiting the yield and/or acreage eligible for
payment. For example, the program yield used to calculate the
deficiency payment has been frozen since 1985. This limits
deficiency payments to the Qp level of production shown in the
figure. Actual yields, however, have continued to increase and in
1992 were about 15 percent higher than program yields (the yields
used to calculate deficiency payments). Production above program
yield is the participants' marginal production and is sold at the
market price represented by Pm in the figure.\12
Furthermore, under the flex acre program, acreage on which producers
receive the deficiency payment is reduced by 15 percent (after ARPs)
for normal flex acres and an additional 10 percent for optional flex
acres. Production on this acreage can be sold for the market price.
In the figure, Qd-Qp represents the quantity produced and sold at the
market price by participants as a result of production on flex acres
and the production of nonparticipants, who also sell their product at
the market price.\13
Under the cotton program, government-held stocks have been greatly
reduced as a result of the marketing loan program. Under this
provision, producers are permitted to redeem their loan at the
USDA-calculated world price, which is lower than the loan rate. The
difference between the loan rate and this lower redemption rate
represents a payment from the government to producers, which is
called the marketing loan gain. Producers receive this gain in
addition to the target price. In some years, this additional payment
has resulted in producers' receiving more than the target price on
eligible production.
--------------------
\6 The figure is a theoretical construct that represents a
generalization of the U.S. cotton market. In any given year,
specific details may differ from those in the figure. In addition,
participation in the cotton program is between 84 to 92 percent. The
fact that there are both program participants responding to the
program and nonparticipants responding to the market price makes
graphic depiction somewhat difficult.
\7 The no-program scenario assumes no deficiency payments, no
marketing loan program, no acreage reduction program (ARP), no flex
acreage, and no 50/92 program.
\8 Consistent with the treatment in Gardner's model, our analysis
incorporates the assumption that in a no-program scenario, annual
beginning and ending stocks would cancel each other out under normal
market conditions, so that stocks would not accumulate.
\9 This analysis does not incorporate export promotion programs that
may or may not affect demand.
\10 Government costs will in part depend on the quantity of cotton
put into storage, which is related to the price elasticity of supply
and demand as well as the level of support in relation to the
no-program price.
\11 Land is also removed from production through the Conservation
Reserve Program (CRP). Although the point can be made that without
the CRP, ARPs would have been higher, it is assumed in this model
that the CRP would continue in the absence of the cotton program for
environmental reasons. The ARP was not considered for environmental
benefits because, unlike the CRP, it is a short-term idling of land,
and benefits, if measurable, are likely to be minimal.
\12 Participants can sell this marginal production at the loan rate
or the market price, whichever is higher. During most of the period
covered by the analysis, the market price has been higher than the
loan rate. Therefore, the market price is depicted in the figure.
\13 The marginal production of program participants sold in the open
market is based on actual yield, which is higher than program yield.
DERIVATION OF THE NO-PROGRAM
SUPPLY AND DEMAND FUNCTIONS
------------------------------------------------------- Appendix I:1.2
In order to calculate the economic welfare effects of the cotton
program, it is necessary to know more about the no-program supply and
demand curves as well as equilibrium price and quantity. This is
because these equilibrium prices and quantities are used as reference
points against which changes in the market caused by the program are
measured. Unfortunately, much of this information is not observable
(particularly on the supply side) in today's market because today's
market operates under the program. Therefore, the no-program supply
and demand curves as well as equilibrium price and quantity must be
estimated.
According to the Gardner method, this estimation is done by using
current available data (with the program in effect) to estimate a
single point on each of the no-program supply and demand curves. (In
the figure, these points are represented as point B for the supply
curve and point A for the demand curve.) Then, using the assumption
of constant elasticity in the relevant range of the function, the
identified points are extended so that the entire no-program supply
and demand functions can be approximated. These extended supply and
demand functions are then used to calculate the no-program
equilibrium price and quantity.
CALCULATION OF POINT A ON
THE NO-PROGRAM DEMAND
FUNCTION
------------------------------------------------------- Appendix I:1.3
As stated above, the components of the cotton program included in
this analysis do not cause a shift in the demand curve. Therefore,
the most readily observable point on the demand curve is the one at
today's current price-quantity combination represented by point A in
the figure. At this point, Qd quantity of production is sold at Pm
(defined as the farm-level, 12-month-season average price).
CALCULATION OF POINT B ON
THE NO-PROGRAM SUPPLY
FUNCTION
------------------------------------------------------- Appendix I:1.4
As stated above, the cotton program does cause a shift in the supply
curve, making the no-program supply curve more difficult to estimate
than the no-program demand curve. To locate the no-program supply
curve, we identified one price-quantity combination (point B)
representing a point on the curve. As the first step in this
process, we estimated a no-program market price (with no acreage
restrictions in place) that would leave producers as well off as the
current situation (with acreage restrictions in place). This price,
called the producer incentive price (PIP), is the weighted average of
the price that program participants receive from the cotton program
(called returns from participation) and the market price that
nonparticipants receive. The PIP therefore can be thought of as a
price faced by an aggregated "composite" producer made up of both
program participants and nonparticipants. Instead of responding
solely to returns from participation in the program or to the market
price, producers respond to a blend of the two prices.
We then located the appropriate no-program quantity, called Qb, that
corresponds to the PIP. Starting from observed production data under
the program, we calculated the quantity of cotton that would have
been produced in the absence of the program by using information on
yearly ARP levels, 50/92 acres, flex acres idled, and estimates of
slippage.\14 These acres would come back into production because,
adjusting for slippage, producers would have an economic incentive to
plant on them at the market price equivalent to the average return
the producers earn when the program is in effect. Given the
producers' original commitment of land under the program provisions
at the PIP, producers would be likely to produce on these additional
acres because by doing so they would earn the same return as they
were earning with the program. This quantity, in combination with
the PIP, identifies point B on the no-program supply curve. We then
used estimates of elasticities of supply to identify the remainder of
the curve and find its intersection with the demand curve.
The following section describes how we calculated the PIP and its
major component, the return from participation in the program. The
subsequent section describes how we found the no-program quantity
that corresponds to the PIP.
--------------------
\14 Slippage occurs when the level of commodity production decreases
by a smaller percentage than the number of idled acres under a
program such as ARP. Although slippage ranges for ARP, 50/92, and
idled flex may differ, data on slippage for specific program
provisions is not available. Therefore, we used a range of slippage
estimates (0.30-0.42) which includes acreage and yield slippage for
all program provisions.
CALCULATION OF THE PIP,
THE PRICE COORDINATE FOR
POINT B
------------------------------------------------------- Appendix I:1.5
The PIP is the weighted average of two prices: (1) a price
representing net returns from participation in the program and (2)
the market price that represents the expectations of nonparticipants.
The PIP elicits the quantity that is produced by a representative or
"average" producer, accounting for both participants and
nonparticipants. It is lower than the target price because it
incorporates the cost to participants of idled land as well as the
market price weighted by nonparticipants. This price would produce
the equivalent net returns, without acreage constraints and other
program provisions, that producers obtain under the program with
acreage restraints. The expression for the PIP is:
(1) PIP = (Participation Rate * Net Returns From Participation) +
((1 - Participation Rate) * Market Price)
Several terms in equation 1, such as the participation rate and
market price, are data that are readily available. However, net
returns from participation must be calculated. This calculation
incorporates aspects of the program, such as the target price, frozen
program yields, marketing loan gains, 50/92 payment on idled acres,
revenues foregone on idled acres, and the return from flex acres
planted to crops other than cotton, which affect producers' returns
under the program.
CALCULATION OF RETURNS FROM
PARTICIPATION, USED TO
DETERMINE THE PIP
------------------------------------------------------- Appendix I:1.6
The returns from participation are calculated as the difference
between the expected revenues from the program and the costs of
participating. Producers derive revenues from the program through
the target price and marketing loan gains. However, in order to be
eligible to receive this income support, producers must agree to
leave a specified portion of their land idle under the ARP.
Additional land is left idle because of the economic incentives
provided by the 50/92 and flex acre programs. By leaving land idle,
producers incur costs represented primarily by the opportunity costs
of not producing on the idled land.
The calculation of the returns from participation derives from the
fact that producers have an incentive to join the program if they
receive more from the program (after accounting for program costs)
than they would if they did not join and received only the market
price for their production. On a per-acre basis, producers would
join the program if\15
(2) (PT*YP*(1-ARP)-TFC-VC(1-ARP))>(MP*YA-TFC-VC)
where
PT = Target price per pound (lb.)
YP = Program yield in lb. per acre
ARP = Effective acreage reduction as a percentage of the complying
base acres (percentage)
TFC = Total fixed costs per acre
VC = Variable cost per acre
MP = Expected market price per lb.
YA = Average yield in lb. per acre
Substituting and rearranging the terms,
(3) (PT*YP*(1-ARP)-TFC+TFC+(1-1+ARP)VC)>(MP*YA)
or
(4) (PT*YP*(1-ARP)+ARP*VC)>(MP*YA)
Dividing through by yield to obtain the revenue per pound (the
per-unit price):
(5) ((PT*(YP/YA)*(1-ARP))+(ARP*VC)/YA)>(MP)
The calculation of returns for participation (RP) in equation 6 is
based on equation 5. The left-hand side of equation 5 specifies that
the return for participation equals the revenue received on program
acres plus saved variable costs on the idled acreage.\16
The actual calculation, however, is further modified to account for
additional aspects of the program, such as frozen program yields,
marketing loan gains, 50/92 payments on idled acres, and, starting in
1991, foregone deficiency payments on flex acres and returns from
flex acres planted to crops other than cotton, all of which affect
producers' returns under the program.
One such modification is shown in equation 6.
(6) [RP=(PT*(YP/YA)*(1-ARP))+(1-ARP)*
(MP*(YA-YP)/YA)+((ARP*VC)/YA)]>MP
Equation 6 differs from equation 5 by the addition of the term
(6a) (1-ARP)*(MP*(YA-YP)/YA)
This term reflects the fact that program yield and actual yield are
not equal. As shown by the first term in equation 6, producers
receive the target price only on program yield. The program yield
for cotton has been frozen at historic levels since the 1985 Farm
Bill. However, actual yields have continued to increase and in 1992
were 15-percent greater than frozen program yields. Producers
receive returns from the market on this additional production. These
additional returns affect their incentive to participate in the
program. Other similar adjustments were made to account for other
components of the program and their effect on participants' returns.
In addition, because the data were unavailable, we did not take into
consideration the costs associated with cover crops on the idled
acres. The impact of these costs are likely to be very small.
--------------------
\15 This equation is a simplification of a producer's decision about
participation. Other components are discussed below.
\16 Saved variable costs are defined as variable costs plus unpaid
labor.
CALCULATION OF QB, THE
QUANTITY COORDINATE FOR
POINT B
------------------------------------------------------- Appendix I:1.7
The PIP provided the price coordinate for an estimated point on the
no-program supply curve (point B in the figure). Additional
calculations are necessary, however, to identify the corresponding
quantity coordinate, Qb. This quantity represents the amount that
farmers would produce, in the absence of the program's acreage
constraints, if they received from the market the same return (as
indicated by the PIP) in the absence of a program that they currently
receive under the program. To calculate this quantity, we used
actual cotton production adjusted for program participation and the
percentage of idled ARP, flex acres, and 50/92 acres. However, not
all acres currently left idle would be expected to be brought back
into production. In addition, the idled acres are likely to be
lower-yielding acres. Therefore, the estimate of no-program
production was further adjusted by estimates of production slippage.
The resulting estimated production, in the case of cotton, was
greater than present quantities.
Under this scenario, quantity in conjunction with the PIP located a
point (B) on the no-program supply curve. We found the remainder of
the no-program supply curve by using estimates of supply elasticities
from other studies and the assumption of constant elasticity in the
relevant range of the supply function.
NO-PROGRAM EQUILIBRIUM PRICE
AND QUANTITY
------------------------------------------------------- Appendix I:1.8
After finding a probable point on each of the no-program supply and
demand curves--points A and B--we used constant elasticity functional
forms and elasticities of supply\17 and demand\18 to extend the
points:
(7) Qd=KdP\n
(8) Qs=KsP\E
where
Qd = Quantity demanded
Qs = Quantity supplied
Kd = Shift parameter or intercept term for demand equation
Ks = Shift parameter or intercept term for supply equation
P = Price
n = Price elasticity of demand
E = Price elasticity of supply
The shift parameter for the demand equation, Kd, was found by
substituting the data for point A in the figure into equation 7 and
then solving for Kd. For example, the actual quantity demanded was
substituted for Qd and the 12-month-season average price was
substituted for P. These values were then used to solve for the
intercept. The same procedure was used to determine the supply
intercept, Ks, in the supply equation. The data for point B in the
figure were substituted into equation 8. In this case, the PIP was
used for P and the quantity supplied at point B (QB) was used for Qs.
These values were then used to solve for Ks.
Once we determined the intercepts, we solved for the no-program
equilibrium price and quantity, Pe and Qe. This was done by equating
supply and demand, substituting the estimated values for the shift
parameters into the equations, and solving for Pe:
(9) Pe=(Ks/Kd)\1/(n-E \)
We calculated equilibrium quantity by substituting the appropriate
demand/supply shift parameters into the appropriate demand/supply
function and solving for Qe.
The resulting estimates are shown in table I.1. On average, over the
period, no-program production would have been greater than program
levels, while prices would have been about the same. This implies
that over the period covered by the analysis, the cotton program,
through its reductions in acreage, has generally had a restrictive
impact on production, despite the incentives to increase production
provided by the target price.\19
Table I.1
Estimated No-Program Price and Quantity,
Crop Years 1986-93
(Prices in 1993 dollars; quantities in
millions of pounds)
No- No-
program program Program Program
Crop year price quantity price quantity
---------------- ---------- ---------- -------- --------
1986 $0.75 5,773 $0.65 4,572
1987 0.75 7,006 0.78 6,948
1988 0.61 6,950 0.65 7,237
1989\a 0.66 8,278 0.72 5,522
1990 0.71 7,801 0.72 7,271
1991 0.59 7,703 0.59 8,264
1992 0.60 6,687 0.55 7,541
1993 0.61 7,837 0.58 7,567
Average $0.66 7,254 $0.66 6,865
------------------------------------------------------------
\a No-program quantities for 1989 are estimated on the basis of
average acreage because adverse weather in the Southwest and Delta
regions severely reduced harvested acreage in this year. These
estimates indicate what would have been produced in the absence of
adverse weather.
--------------------
\17 We used supply elasticities ranging from 0.5 to 0.74, with 0.62
as the average. This range was based on elasticities presented in
the economic literature.
\18 We used a weighted average of domestic and export demand
elasticities, which ranged from -1.05 to -0.78, depending on the
years. These elasticities were the midpoints of the ranges presented
in the economic literature.
\19 If in fact the analysis had shown that on average the no-program
prices were lower and quantities were higher than program levels, the
program would still have resulted in social welfare losses. In this
case the losses would have been due to overproduction.
METHODOLOGY FOR MEASURING GAINS
AND LOSSES
--------------------------------------------------------- Appendix I:2
After we calculated the no-program price and quantity, we measured
the economic welfare effects of the cotton program on cotton buyers,
cotton producers, and taxpayers. We measured these effects by using
the estimated no-program equilibrium price and quantity as reference
points against which we measured, as gains or losses, changes caused
by the program in real income for market participants.
COTTON BUYERS' GAIN OR LOSS
------------------------------------------------------- Appendix I:2.1
Cotton buyers' gain or loss\20 as a result of the cotton program is
determined by the relationship between market prices with and without
the program. Buyers of cotton gain if they pay a lower price under
the program than they would have paid if there were no program.
Conversely, they lose if they pay a higher price. In 1986, 1992, and
1993, buyers gained because they paid less for cotton than they would
have in the absence of the program. Cotton buyers' gains in 1992 and
1993 derived, in part, from the extent to which step 2 payments made
to domestic mills contributed to lower prices. From 1987 to 1991,
however, cotton buyers incurred costs by paying higher prices for
cotton than they would have without the program. Mathematically,
domestic cotton buyers' losses or gains was estimated using the
following expression:
(10) DCB=((SAP-Pe)*(0.5*(Qd+Qe)))*DD
where
DCB = Domestic cotton buyers' loss or gain
SAP = 12-month-season average price
Pe = No-program equilibrium price
Qd = Quantity demanded under program
Qe = No-program equilibrium quantity
DD = Domestic quantity demanded as a percentage of total quantity
demanded
--------------------
\20 To the extent that the cotton program affects world prices,
international consumers would also experience gains and losses. The
present analysis focuses only upon the program's impact on domestic
buyers.
PRODUCERS' GAIN OR LOSS
------------------------------------------------------- Appendix I:2.2
Producers' gain or loss under the cotton program is determined by the
net welfare effect of the program on participants and
nonparticipants. As is the case with cotton buyers, nonparticipants'
gain or loss depends upon the relationship between market prices
under the program and no-program prices. This is because, as defined
above, nonparticipants receive the market price for their production.
Nonparticipants gain if the program market price is higher than the
no-program price. Conversely, they lose if this price is lower.
Since participants respond to program prices, participants' gain
depends on the relationship between the returns from participation
per pound, which was used above to calculate the PIP and the
no-program price.
Average producer gain for both participants and nonparticipants is
shown in figure I.1 by area PIPBePe, which is the difference between
the PIP and the no-program price to the left of the no-program supply
curve. The PIP/Qb price-quantity combination is used to determine
producers' surplus because it represents the quantity that would be
produced, in the absence of the program, if producers received the
PIP (which is the price they currently receive under the program,
adjusted for the costs of ARPs).
Mathematically, producers' net gain or loss was determined using the
following expressions for participants' gain and nonparticipants'
gain or loss:
(11) PG=((RP-Pe)*PR)*(0.5*(Qe+Qb))
(12) NGL=((SAP-Pe)*(1-PR))*(0.5*(Qe+Qb))
where
PG = Participants' gain
RP = Returns from participation (used to calculate the PIP)
PR = Participation rate
Pe = No-program equilibrium price
Qe = No-program equilibrium quantity
Qb = Observed quantity without program set-asides
NGL = Nonparticipants' gain or loss
SAP = 12-month-season average price
GOVERNMENT COSTS
------------------------------------------------------- Appendix I:2.3
We calculated the budgetary costs as the sum of deficiency payments,
marketing loan gains, storage, transportation, handling of CCC
stocks, and losses on the sale of CCC stocks. Deficiency payments
are represented by the rectangle Pt-Pm for the volume of Qp in the
figure. Other budgetary costs, however, are not represented.
SOCIAL WELFARE LOSS
------------------------------------------------------- Appendix I:2.4
The social welfare loss is the amount of revenue that taxpayers or
cotton buyers give up but that producers do not gain. For every
dollar paid by cotton buyers and the government, the gains to
producers are less than a dollar. This revenue is lost to society
and actually measures the economic inefficiencies of income transfer
associated with having a cotton program. In the case of cotton, most
of the deadweight loss is due to the lost returns from idled land
that can be approximated by the area PtfcPIP in figure I.1.\21 This
area is based on the supply curve S', which takes into account the
cost of idle land. In addition, some social welfare loss occurs when
ARPs result in cotton buyers' consuming less cotton at higher prices
than they would in the absence of the cotton program. This social
welfare loss is approximately represented by area Aeg in the figure.
Another component of social welfare loss, which was important during
the period covered by our analysis, was CCC stocks that had been
accumulated prior to 1985 but that were sold at a loss as part of the
transition to the marketing loan program.\22
We used the following equation to arrive at our estimates of
deadweight loss:\23
(13) DWL=(GOVT+DCB)-(PG+NGL)
where
DWL = Deadweight loss
GOVT = Government budgetary cost
DCB = Domestic cotton buyers' gain or loss
PG = Participants' gain
NGL = Nonparticipants' gain or loss
--------------------
\21 Again, because of the complexity of the program, graphic
representation is an approximation.
\22 The sale of CCC stocks reduced market prices, resulting in a loss
for the government and society. This loss was partially offset by a
gain for the cotton buyers who purchased cotton at the reduced
prices.
\23 Since this equation contains the gains or losses for domestic
cotton buyers only, the deadweight loss is domestic deadweight loss
only.
RESULTS OF ECONOMIC WELFARE
ANALYSIS
------------------------------------------------------- Appendix I:2.5
The results of our economic welfare analysis of the cotton program
appear in tables I.2, I.3, I.4, and I.5. The analysis shows that
from 1986 to 1993, cotton buyers gained an average of $16 million
from the cotton program. In 5 of the 8 years analyzed, the cotton
program resulted in additional costs to cotton buyers. These costs
occurred because the cotton program generally restricted production.
In 3 of the 8 years, cotton buyers gained as a result of the program,
and these gains outweighed the costs incurred the other years. In
1986, cotton buyers gained $329 million. This occurred as a result
of the transition to the marketing loan program, when the government
released previously accumulated stocks onto the market, reducing
market prices. In crop years 1992 and 1993, cotton buyers also
received a gain because ARPs were relatively low and cotton yields
were relatively high, resulting in low prices. In addition, cotton
buyers gained by the extent to which step 2 payments made to domestic
mills those years contributed to the lower prices. From 1986 to
1993, taxpayers' costs averaged $1,509 million over the period.
Total cotton buyers' and taxpayers' costs amounted to $1,493 million.
Only about 51 percent of these costs were actually transferred to
producers. Social welfare loss accounted for the remaining
percentage. Producers gained an average of $754 million during the
period, while the social welfare loss averaged $738 million.
Table I.2
Gains and (Losses) to Cotton Buyers,
Crop Years 1986-93
(In millions of 1993 dollars)
Baseli Minimu Maximu
Crop year ne m m
------------------------------------ ------ ------ ------
1986 $329 $275 $376
1987 (134) (194) (82)
1988 (156) (224) (97)
1989 (233) (288) (186)
1990 (50) (75) (30)
1991 (28) (85) 21
1992 270 202 327
1993 134 84 177
Average $16
------------------------------------------------------------
Note: The baseline estimate represents the average estimate for the
year, calculated using the average elasticity and slippage estimates.
We used three supply and demand elasticity combinations (high, low,
average) and two slippage rates to produce a total of six different
estimates of gains or losses to cotton buyers, taxpayers, and
producers and the social welfare loss for each year. Of these six,
the average estimate was calculated using the average elasticity and
slippage factor. The minimum and maximum represent the high and low
of the six estimates.
Table I.3
Net Gain to Cotton Producers, Crop Years
1986-93
(In millions of 1993 dollars)
Baseli Minimu Maximu
Crop year ne m m
------------------------------------ ------ ------ ------
1986 $ 688 $ 583 $ 809
1987 625 515 752
1988 1,405 1,270 1,561
1989\a 522 424 634
1990 162 122 209
1991 982 885 1,096
1992 887 787 1,005
1993 766 684 861
Average $ 754
------------------------------------------------------------
Note: The baseline estimate represents the average estimate for the
year, calculated using the average elasticity and slippage estimates.
We used three supply and demand elasticity combinations (high, low,
average) and two slippage rates to produce a total of six different
estimates of gains or losses to cotton buyers, taxpayers, and
producers and the social welfare loss for each year. Of these six,
the average estimate was calculated using the average elasticity and
slippage factor. The minimum and maximum represent the high and low
of the six estimates.
\a Producer gains in 1989 are estimated on the basis of average
acreage because adverse weather in the Southwest and the Delta
regions severely reduced harvested acreage that year. These
estimates indicate what the gains would have been in the absence of
the adverse weather.
Table I.4
Government Costs, Crop Years 1986-93
(In millions of 1993 dollars)
Crop year Baseline
-------------------------------------------------- --------
1986 ($
2,693)
1987 (1,546)
1988 (1,585)
1989 (748)
1990 (439)
1991 (1,223)
1992 (2,031)
1993 (1,810)
Average ($
1,509)
------------------------------------------------------------
Table I.5
Social Welfare Loss, Crop Years 1986-93
(In millions of 1993 dollars)
Baseli Minimu Maximu
Crop year ne m m
------------------------------------ ------ ------ ------
1986 ($ ($ ($
1,676) 1,734) 1,609)
1987 (1,055 (1,113 (987)
) )
1988 (336) (412) (248)
1989 (460) (510) (402)
1990 (327) (347) (305)
1991 (269) (317) (212)
1992 (875) (917) (824)
1993 (910) (949) (865)
Average ($
738)
------------------------------------------------------------
Note: The baseline estimate represents the average estimate for the
year, calculated using the average elasticity and slippage estimates.
We used three supply and demand elasticity combinations (high, low,
average) and two slippage rates to produce a total of six different
estimates of gains or losses to cotton buyers, taxpayers, and
producers and the social welfare loss for each year. Of these six,
the average estimate was calculated using the average elasticity and
slippage factor. The minimum and maximum represent the high and low
of the six estimates.
OBJECTIVES, SCOPE, AND METHODOLOGY
========================================================== Appendix II
In response to a request from Representative Richard K. Armey, we
conducted a broad-based review of the federal cotton program. We
evaluated the program's cost and complexity, distribution of payments
and benefits, effects on producers' costs and returns, and
effectiveness in enhancing U.S. cotton exports.
We obtained an understanding of how the cotton program works and how
it has evolved to its current level of complexity by interviewing
Deputy Directors, Branch Chiefs, and analysts in several divisions of
the USDA's Consolidated Farm Service Agency. In addition, we met
with agricultural economists from USDA's Economic Research Service;
the Director and Deputy Director of the Tobacco, Cotton, Seeds
Division of USDA's Foreign Agricultural Service; and the Deputy
Director, Cotton Division of USDA's Agricultural Marketing Service.
We also spoke with executive officers of the National Cotton Council
and member associations. To obtain cotton producers' perspectives on
the program, including such issues as the benefits and environmental
effects of cotton farming, we spoke with executive officers of the
three largest producer marketing cooperatives and with individual
producers in California, Texas, Georgia, and Mississippi. In
addition, we analyzed legislation, regulations, and USDA reports on
the history of the cotton program.
To develop cost estimates of the cotton program and of other
government programs that support cotton, we interviewed personnel in
the Consolidated Farm Service Agency's Fibers and Rice Analysis
Division and Budget Division and collected and analyzed cost data
from various USDA offices for crop years 1986-93.
To understand the program's many elements, we reviewed regulations,
documents, and reports, and interviewed USDA County Executive
Directors and several Branch Chiefs in the Cotton, Grain, and Rice
Price Support Division at USDA headquarters. Using this information,
we prepared graphs and flow charts to illustrate how the program
works and obtained USDA concurrence on the accuracy of our depiction.
To determine the distribution of program payments and benefits among
cotton producers, we used USDA's records of payments for all U.S.
cotton farms for crop year 1993.
To evaluate the effects of the program on producers' cost and
revenue, we obtained and analyzed USDA data on average U.S. domestic
prices and U.S. and geographic regional production costs for upland
cotton for crop years 1986-93. To compute aggregate U.S. per pound
production cost, we used USDA's reported total planted acres and
production and computed the aggregate U.S. per pound yield per
planted acre for each of the 8 crop years. Our analysis assumes that
each producer has full ownership of all assets required to produce
and market cotton. In addition, our analysis considers only cotton
produced on acreage participating in the program for which the
government was liable to make some type of benefit payment to the
producer, including deficiency, marketing loan gains, or loan
deficiency payments.
To compute the per pound market revenues received by producers, we
used USDA's July 1994 and October 1994 Agricultural Prices Summaries
for upland cotton, which reported the U.S. average market price per
pound received by farmers. The U.S. average government payment
rates per pound used in our analysis came from a USDA-prepared
schedule showing the U.S. average deficiency payment and loan
deficiency payment rates USDA paid on a pound of cotton participating
in the program. Our analysis was conservative, in that it did not
include disaster payments that producers might have received.
To estimate the range of cotton producers' short-run and long-run
total production cost, we computed each producer's average per pound
production cost using the 1981-85 historical per harvested acre yield
for each farm (as recorded in USDA files). Also, we assumed that the
producer's per acre cost was equal to the average for a producer's
regional location. We used the 1981-85 per harvested acre yields for
each cotton farm, called "program yields," because under the Food
Security Act of 1985, USDA "froze" the yields for each farm and still
uses them today to compute deficiency payments. USDA no longer
collects actual up-to-date yields, neither of planted nor harvested
acres, on each cotton farm. Therefore, the 1981-85 yields are the
only yield data available within USDA that are based on the actual
cotton production for each farm participating in the cotton program.
To assess the economic impact of the program on cotton buyers' costs
and producers' benefits, we identified and analyzed economic studies
of the U.S. cotton program. We used a methodology developed by
Bruce L. Gardner of the University of Maryland to determine the
welfare effects--economic gains and losses--of the program. We
worked with him to ensure that we accurately applied the model to the
cotton program. A discussion of how we measured the welfare gains
and losses of the cotton program and how we modified the Gardner
model is included in appendix I. The Gardner model, like most models
used for welfare analyses, is a static partial equilibrium model. It
does not consider a movement to worldwide free trade, nor does it
consider a complete absence of governmental intervention in
agriculture (such as an absence of disaster payments, research and
development, etc.). We used program cost data provided by USDA. We
used this model in our previous evaluations of the rice and wheat
programs.\1
To evaluate the impact of program features on cotton exports and on
the U.S. share of cotton export markets, we obtained and analyzed
USDA data on U.S. cotton exports and on U.S. export promotion
efforts. We assessed the effect of the cotton program on domestic
and international cotton marketing by reviewing and analyzing program
features dealing with controlling the supply of cotton (such as
acreage reduction requirements) and producers' use of the nonrecourse
loan and the effect of these program features on the marketing of
cotton.
To assess prospects for cotton exports under changing international
trade conditions, we reviewed analyses of the potential effects of
NAFTA and GATT. We also discussed possible future opportunities for
U.S. cotton exports under increased free-trading conditions with the
(1) Director and Deputy Director of the Tobacco, Cotton, Seeds
Division of the Foreign Agricultural Service; (2) Executive Director
of Cotton Council International (Cotton Council International is the
export promotion arm of the Cotton Council); and (3) Executive
Director of the International Cotton Advisory Committee (an
association of governments having an interest in the production,
export, import, and consumption of cotton). We also talked with
officers of the American Textile Manufacturers Institute, including
the Director, International Trade.
We did not independently verify the USDA data used in this report.
As necessary, we adjusted figures in this report to 1993 dollars to
more accurately compare prices and costs over time. For this
adjustment, we used the gross domestic product implicit price
deflator, with 1993 being equal to 1.00.
--------------------
\1 Rice Program: Government Support Needs to Be Reassessed
(GAO/RCED-94-88, May 26, 1994) and Wheat Commodity Program: Impact
on Producers' Income (GAO/RCED-93-175BR, Sept. 8, 1993).
GRAPHIC ILLUSTRATION OF COTTON
PROGRAM OPERATIONS
========================================================= Appendix III
This appendix presents nine flow charts showing how the cotton
program operates at USDA county offices and in headquarters. Figures
III.1 through III.4 illustrate county office operations. Figures
III.5 through III.9 show USDA headquarters operations.
Figure III.1: The County
Office's General Processing
Steps
(See figure in printed
edition.)
.
(See figure in printed
edition.)
Figure III.2: Advance
Deficiency Payment Calculation
(See figure in printed
edition.)
Figure III.3: Loan Deficiency
Payment Processing Steps
(See figure in printed
edition.)
Figure III.4: Loan Processing
Steps
(See figure in printed
edition.)
.
(See figure in printed
edition.)
Figure III.5: Loan Rate
Calculations
(See figure in printed
edition.)
Figure III.6: Adjusted World
Price Calculations
(See figure in printed
edition.)
.
(See figure in printed
edition.)
Figure III.7: Step 2 Payment
Calculations
(See figure in printed
edition.)
Figure III.8: Step 3 Import
Quota Calculations
(See figure in printed
edition.)
Figure III.9: Spot Price
Import Quota Calculations
(See figure in printed
edition.)
(See figure in printed edition.)Appendix IV
COMMENTS FROM THE U.S. DEPARTMENT
OF AGRICULTURE
========================================================= Appendix III
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
MAJOR CONTRIBUTORS TO THIS REPORT
=========================================================== Appendix V
Robert C. Summers, Assistant Director
Jerry W. Coffey, Evaluator-in-Charge
Carol E. Bray
Barbara J. El Osta
W. Fred Mayo
John F. Mitchell
Paul J. Pansini
Stuart Ryba
Louis J. Schuster
Pamela Armstrong Scott
Loren Yager
RELATED GAO PRODUCTS
============================================================ Chapter 1
Wheat Support: The Impact of Target Prices Versus Export Subsidies
(GAO/RCED-94-79, June 7, 1994).
Rice Program: Government Support Needs to Be Reassessed
(GAO/RCED-94-88, May 26, 1994).
Dairy Industry: Potential for and Barriers to Market Development
(GAO/RCED-94-19, Dec. 21, 1993).
Wheat Commodity Program: Impact on Producers' Income
(GAO/RCED-93-175BR, Sept. 8, 1993).
Peanut Program: Changes Are Needed to Make the Program Responsive to
Market Forces (GAO/RCED-93-18, Feb. 8, 1993).
Cotton Program: The Marketing Loan Has Not Worked (GAO/RCED-90-170,
July 31, 1990).
Sugar Program: Changing Domestic and International Conditions
Require Program Changes (GAO/RCED-93-84, Apr. 16, 1993).
General Agreement on Tariffs and Trade: Agriculture Department's
Projected Benefits Are Subject to Some Uncertainty (GAO/RCED-94-272,
July 22, 1994).
General Agreement on Tariffs and Trade: Uruguay Round Final Act
Should Produce Overall U.S. Economic Gains (GAO/GGD-94-83B, Volume
2, July 1994).