Maritime Security Fleet: Many Factors Determine Impact of	 
Potential Limits on Food Aid Shipments (13-SEP-04, GAO-04-1065). 
                                                                 
Food aid cargo must generally be carried on U.S.-flag ships under
requirements set by the cargo preference program. Two groups of  
carriers compete for this cargo: (1) those that participate in	 
the Maritime Security Program and receive an annual government	 
subsidy--generally liners operating on scheduled routes and (2)  
those that do not--generally carriers operating on a charter	 
basis. Congress directed GAO to study (1) how the cargo 	 
preference and Maritime Security programs are designed and who	 
participates;(2) the nature and extent of MSF and non-MSF carrier
participation and competition in the food aid program; and (3)	 
how a tonnage limitation on bagged preference cargo for MSF	 
vessels could affect MSF, other U.S.-flag ships, the cargo	 
preference food aid program, and the ports servicing these ships.
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-04-1065					        
    ACCNO:   A12455						        
  TITLE:     Maritime Security Fleet: Many Factors Determine Impact of
Potential Limits on Food Aid Shipments				 
     DATE:   09/13/2004
  SUBJECT:   Maritime Security 				        
                                                                 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO 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.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-04-1065

                 United States Government Accountability Office

                     GAO Report to Congressional Committees

September 2004

MARITIME SECURITY FLEET

    Many Factors Determine Impact of Potential Limits on Food Aid Shipments

                                       a

GAO-04-1065

Highlights of GAO-04-1065, a report to congressional committees

Food aid cargo must generally be carried on U.S.-flag ships under
requirements set by the cargo preference program. Two groups of carriers
compete for this cargo: (1) those that participate in the Maritime
Security Program and receive an annual government subsidy-generally liners
operating on scheduled routes and (2) those that do not-generally carriers
operating on a charter basis. Congress directed GAO to study (1) how the
cargo preference and Maritime Security programs are designed and who
participates;(2) the nature and extent of MSF and non-MSF carrier
participation and competition in the food aid program; and (3) how a
tonnage limitation on bagged preference cargo for MSF vessels could affect
MSF, other U.S.-flag ships, the cargo preference food aid program, and the
ports servicing these ships.

While GAO makes no recommendations in this report, we observe that a
decline in bulk food aid shipments since 2000 suggests greater competition
for bagged food aid, but clear trends in how industry segments are
responding cannot yet be determined. Our analysis suggests that under
certain conditions a tonnage limit would not lead to large shifts in food
aid to non-MSF carriers, and could result in lower levels of subsidy
payments and increased agency burdens. We note significant limits of
efforts to predict the future course of events where key factors are so
volatile.

www.gao.gov/cgi-bin/getrpt?GAO-04-1065.

To view the full product, including the scope and methodology, click on
the link above. For more information, contact Celia Thomas at (202)
512-8987 or [email protected].

September 2004

MARITIME SECURITY FLEET

Many Factors Determine Impact of Potential Limits on Food Aid Shipments

The cargo preference program and the Maritime Security Program provide
incentives to retain privately owned U.S.-flag ships and their U.S.
citizen mariners for commercial and national defense purposes. The cargo
preference program is open to all U.S.-flagged vessels, while the Maritime
Security Fleet (MSF) subsidy is only available to certain militarily
useful vessels. Of the 47 ships currently in the MSF, 37 have participated
in cargo preference food aid shipments.

MSF and non-MSF carriers compete for food aid shipped as bagged cargo,
which averaged 33 percent of food aid shipments by tonnage from fiscal
years 1999 to 2003. There is no competition for bulk food aid shipments
because MSF carriers do not carry bulk cargo. Changes in food aid spending
have contributed to a shift from bulk to bagged cargo and increased
reliance on bagged cargo by some non-MSF carriers. From 1999 to 2003, MSF
carriers shipped about 45 percent and non-MSF carriers 55 percent of
bagged food aid cargo. Competition between MSF and non-MSF carriers for
bagged food aid is affected by certain cargo preference requirements.

Establishing a tonnage limitation on MSF vessels would likely reduce their
share of food aid shipments, but the extent would depend on factors such
as the level of the limit and the options MSF carriers have in responding
to it. We examined three proposed limits and found that the percentage of
food aid voyages carrying more than the proposed limit rises from 3
percent with a limit of 7,500 tons to 19 percent above 2,500 tons,
according to fiscal year 2001 to 2003 data. The actual impact on MSF
carriers will be smaller if they are able to (1) carry some food aid up to
the limit, (2) replace some food aid above the limit with other cargo,
and/or (3) elect to carry food aid even without the subsidy. Food aid
agencies are concerned about the impacts of a tonnage limit, including
increased delays in providing food aid, administrative burdens, and higher
shipping costs. Major ports would generally experience a limited overall
impact of a tonnage limitation, but specific food aid terminals could be
affected.

Non-MSF and MSF Bagged Food Aid Shipments, Fiscal Years 1999-2003

Tons in thousands

1,000

800

600

Non-MSF

MSF

400

200

0 1999 2000 2001 2002 2003a Year

                           a2003 data are estimated.

Contents

  Letter

Results in Brief
Background
Cargo Preference and Maritime Security Programs Provide

Incentives to Retain U.S.-Flag Ships and Mariners MSF and Non-MSF Carriers
Compete for Bagged Food Aid Shipments Multiple Factors Determine How
Tonnage Limit Changes Would

Impact Food Aid Carriers Observations Agency Comments and Our Evaluation

1 3 5

11

16

26 38 39

Appendixes

Appendix I: Appendix II:

Appendix III: Appendix IV:

                                       Appendix V: Appendix VI: Appendix VII:

Objectives, Scope, and Methodology

Simulating Potential Impacts of a Bagged Tonnage Limitation

Profile of Maritime Security Fleet

Comments from the U.S. Agency for International Development

Comments from the Department of Agriculture

Comments from the Department of Defense

GAO Contacts and Staff Acknowledgments

GAO Contacts Acknowledgments

                                       45

                                     49 56

59

64

66

67 67 67

Tables	Table 1: Table 2:

Table 3:

Table 4:

Characteristics of U.S.-Flag Vessels Participating in Cargo
Preference Food Aid, Fiscal Years 1999-2003 6
Tonnage of Bagged Food Aid Cargo Loaded by MSF and
Non-MSF Vessels at Major Food Aid Ports, Fiscal Years
1999-2003 25
GAO Simulation of MSF Voyages with Food Aid above a
2,500 Tonnage Limit, 3-Year Annual Average for Fiscal
Years 2001-2003 32
Assumptions in the Simulation Model and the Probability
Distributions Reflecting Those Assumptions 51

                                    Contents

Table 5: GAO Simulation of MSF Voyages with Food Aid above a

5,000 Ton Limit, 3-Year Annual Average for Fiscal Years

2001-2003 54 Table 6: Maritime Security Fleet, as of December 2, 2003 56

Figures	Figure 1: Figure 2: Figure 3:

Figure 4: Figure 5: Figure 6: Figure 7:

Figure 8: Figure 9:

A Bulk Vessel 7
A Containership 8
USDA Purchases of Bulk and Bagged Agricultural
Commodities for the Food Aid Program, Fiscal Years
1996-2003 17
Non-MSF Total Food Aid Shipments, Fiscal Years
1999-2003 19
Non-MSF and MSF Bagged Food Aid Shipments, Fiscal
Years 1999-2003 20
Criteria for MARAD's Prioritization of U.S.-Flag Vessel
Service for the Cargo Preference Program 21
Estimated Share of Title II Food Aid Shipments by
U.S.-Flag Liner Carriers at Each Priority Level, Fiscal
Years 1999-2003 23
Number of MSF Food Aid Voyages at Different Tonnage
Levels, Fiscal Years 2001-2003 28
Percentage of MSF Food Aid Revenues from Voyages
with Tonnage above a Potential Limitation, Fiscal Years
2001-2003 29

Contents

Abbreviations

DOD Department of Defense
LASH Lighter Aboard Ships
MARAD Maritime Administration
MSF Maritime Security Fleet
TEU Twenty-foot Equivalent Unit
USAID U.S. Agency for International Development
USDA U.S. Department of Agriculture

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

A

United States Government Accountability Office Washington, D.C. 20548

September 13, 2004

Congressional Committees:

Food aid shipments must generally be transported on U.S.-flag ships under
the requirements set by the cargo preference program.1 Two broad groups of
U.S.-flag carriers compete for these food aid shipments:

1.	those that participate in the Maritime Security Fleet (MSF)2 (generally
liners that carry containers on scheduled routes) and receive a $2.1
million annual subsidy from the federal government and

2.	those that do not (generally carriers of bulk and bagged cargo that
operate on a charter basis).

During the debate on renewing the Maritime Security Program and MSF, the
non-MSF carriers complained that the MSF subsidy put them at a competitive
disadvantage, particularly for bagged food aid cargo. They supported a
provision to limit the amount of bagged cargo that MSF ships could carry
so that, on days when the limit was exceeded, MSF ships would have to
forfeit their subsidy. Since no analysis had been conducted on the effects
of such a change, the Maritime Security Program was authorized without
this provision, and we were directed to study the impact of placing a
limitation on MSF transportation of bagged food aid preference cargo.3

1Cargo preference is the reservation, by law, for transportation on
U.S.-flag vessels, of all or a portion of all ocean borne cargo that moves
in international trade, either as a direct result of the federal
government's involvement or indirectly because of the financial
sponsorship of a federal program or guarantee provided by the federal
government. These preference cargoes include agricultural, military, and
Export-Import Bank cargoes. This report focuses only on cargo preference
food aid shipments; cargo preference shipments of military or other cargo
were outside the scope of our review. Agricultural preference cargoes
accounted for about a quarter of all preference cargoes, by tonnage,
according to Maritime Administration data for 2002.

2MSF comprises vessels that participate in the Maritime Security Program,
a program established by the Maritime Security Act of 1996 that provides
funding to U.S. vessels participating in international trade, to support
the Department of Defense (DOD).

3Section 3535 of the National Defense Authorization Act for fiscal year
2004, P.L. 108-136, 117 Stat. 1818-19.

As discussed with representatives of the House and Senate Committees on
Armed Services and the Senate Committee on Commerce, Science, and
Transportation, we have focused on answering the following questions:

1.	How are the cargo preference and Maritime Security Programs designed to
meet their objectives and who participates in them?

2.	What are the nature and extent of MSF and non-MSF carrier participation
and competition in the food aid program?

3.	How would establishing a bagged cargo preference tonnage limitation on
MSF vessels be expected to affect MSF, other U.S.-flag ships, the cargo
preference food aid program, and the ports servicing these ships?

To answer these questions, we gathered and analyzed food aid shipment data
for fiscal years 1999 to 2003 from the U.S. Agency for International
Development (USAID), the Department of Agriculture (USDA), and the
Maritime Administration (MARAD), as well as MSF and cargo preference
vessel data from MARAD and military contingency participation data from
MARAD and the Department of Defense (DOD).4 Through electronic data
testing, verification of data against other sources, and interviews with
agency officials managing the data, we found the data to be sufficiently
reliable and appropriate for our purposes. We conducted interviews and
obtained financial data for fiscal years 2001 to 2003 from representatives
of 15 carriers that transported the majority of cargo preference food aid,
including 5 MSF and 10 non-MSF carriers.5 We also conducted interviews
with representatives of eight ports through which a major share of food
aid shipments were shipped. Consistent with the terms of our mandate, we
analyzed the potential impact of daily tonnage limits on MSF vessels of
2,500, 5,000, and 7,500 tons.6 While our analysis provides a range of
impact estimates based on data in recent years and several additional
assumptions, if future market conditions differ from those reflected in
this recent data, impacts may differ from those illustrated by our
analysis. We conducted our work in accordance with generally accepted
government

4We did not conduct an evaluation of programs related to this study,
including the cargo preference program and the Maritime Security Program.

5Self-reported financial data is subject to limited verification.
Heightened uncertainties in our analysis relating to financial data are
discussed in appendixes I and II.

6Our analysis and findings are expressed in terms of metric tons.

auditing standards. (See app. I for details about our scope and
methodology.)

Results in Brief 	The cargo preference and Maritime Security Programs both
provide incentives to retain privately owned U.S.-flag ships and their
U.S.-citizen mariners for commercial and national defense purposes. By
allocating a percentage of federal cargoes to U.S.-flag vessels, the cargo
preference program creates a protected market that provides an economic
incentive for vessel owners to pay the higher costs associated with
U.S.-flag registry and employ U.S.-citizen crews. We found that a total of
190 privately owned U.S.-flag vessels carried cargo preference food aid
shipments at some point during the fiscal year 1999 to 2003 period. In
addition, the Maritime Security Program provides a subsidy for MSF
carriers with particular types of militarily useful vessels. MSF currently
has 47 participating ships, of which 37 have also participated in cargo
preference food aid shipments. DOD strongly supports both programs and
said it has benefited from both during the recent wars in Afghanistan and
Iraq.

MSF carriers compete with non-MSF carriers for the 33 percent of the food
aid tonnage that is currently shipped as bagged cargo. However, MSF
carriers do not compete with non-MSF carriers for the 67 percent of food
aid tonnage that is shipped as bulk cargo because MSF vessels do not carry
bulk cargo. Changes in the amount of bulk and bagged commodities purchased
for food aid programs have contributed to a decline in bulk shipments and
a relative increase in bagged shipments from fiscal years 1999 to 2003.
The reduction in bulk food aid shipments has caused some non-MSF bulk
carriers to rely more on bagged food aid shipments in recent years. From
fiscal years 1999 to 2003, MSF carriers shipped about 45 percent and
non-MSF carriers 55 percent of bagged food aid cargo. Competition between
MSF and non-MSF carriers for bagged food aid is affected by certain cargo
preference requirements. For example, MARAD's priority system for
U.S.-flag service guarantees preference for carriers that use only
U.S.-flag vessels to transport food aid cargo. This system has tended to
favor non-MSF carriers because they provide charter service from the port
of origin to the port of destination solely on a U.S.-flag vessel, whereas
MSF liners may use a foreign-flag vessel on one leg of their scheduled
service route. However, a provision of the Maritime Security Act of 1996
allocates a certain amount of bagged food aid cargo to Great Lakes ports,
which has favored MSF carriers and other carriers that offer intermodal
services. MSF and non-MSF carriers ship most of the bagged food aid cargo
from U.S. ports in the Gulf of Mexico.

Establishing a bagged cargo tonnage limitation on MSF vessels would likely
reduce their market share in food aid, but the extent will depend on
factors such as the level of the limit and the options MSF carriers have
in responding to it. Recent data show that only 3 percent of MSF food aid
voyages would be affected by a 7,500-ton limit; 19 percent of MSF food aid
voyages would be affected by a 2,500-ton limit. Almost all MSF voyages
during fiscal years 2001 to 2003 with more than 7,500 tons were on a small
number of noncontainership, specialized vessels that are being phased out
of MSF. If MSF vessels were to lose all food aid shipments on
containership voyages with more than 2,500 tons, the MSF would carry on
average 160,000 tons less food aid per year-out of their average total of
324,000 tons of food aid per year. However, setting a limit at this level
may not mean a reduction of 160,000 tons of food aid cargo for MSF
vessels, to the degree they are able to respond in the following ways: (1)
carrying some food aid up to the limit, (2) replacing some food aid above
the limit with other cargo, and (3) continuing to carry food aid above the
limit when they determine that is the most profitable decision. Under a
scenario in which these options are considered, for a limit of 2,500 tons
for example, we estimate that under certain assumptions the total annual
decline in food aid carried by the MSF would range from 17,000 tons to
63,000 tons. We estimate that the total annual decrease in MSF net
revenues over all carriers would then range from around $2 million to $5
million.7 While this analysis reflects some range of probabilities in the
flexibility MSF carriers have in responding to a tonnage limit, it may not
fully reflect certain logistical challenges that MSF carriers have
identified, such as agreements between carriers to share vessel space.
Food aid agencies and MARAD are also concerned that tonnage limits would
create increased delivery delays, administrative burdens, and higher
shipping costs. Food aid agencies emphasized concerns that these potential
impacts could impede their ability to meet critical humanitarian needs.
The major food aid ports would generally experience a limited impact on
their overall port activity if a bagged food aid tonnage limit were
established, although some terminals within the ports that service MSF
vessels could potentially be affected, depending on the degree to which
MSF participation in food aid were decreased.

USAID, USDA, and DOD provided written comments on a draft of this report,
which are reproduced in appendixes IV, V, and VI. These agencies generally
stated that our report appropriately identified issues and

7We use the term net revenues to reflect revenues minus costs.

concerns regarding food aid shipments and potential tonnage limits. They
expressed concerns that tonnage limits could negatively impact their
program or mission. DOT provided oral comments, which we summarize in the
agency comments section below. DOT said that the draft report provided a
thoughtful analysis of the potential impact of tonnage limitations.
However, it identified issues with some factors in our simulation model
and emphasized its view that MSF carriers would face constraints in
responding to a tonnage limitation. We agree that MSF carriers would face
such constraints and explicitly included these considerations in our
analysis. We modified our report language in several places to clarify our
discussion of these constraints and the uncertainties regarding potential
tonnage limitation impacts. In addition, USAID and DOT provided technical
comments, which we incorporated in the report as appropriate.

Background 	U.S.-flag fleet participants in cargo preference food aid
shipments comprise two general categories of carriers: charter service and
liner service. The cargo preference and Maritime Security Programs are
intended to support both as part of the U.S.-flag fleet. These programs
are administered by MARAD, while the food aid programs are administered by
USAID and USDA.

U.S.-Flag Fleet Generally Vessels in the privately owned U.S.-flag fleet
engaged in international Comprises Charter and commerce can be placed into
two general categories: charter service and Liner Service Carriers liner
service. While most non-MSF carriers provide charter service, most

MSF carriers provide liner service, as shown in table 1 below.

Table 1: Characteristics of U.S.-Flag Vessels Participating in Cargo
Preference Food Aid, Fiscal Years 1999-2003

MSF vessels Non-MSF vessels

Major type of servicea	Liner vessels with regularly Charter vessels hired
to scheduled sailings on a fixed carry cargo to specific ports route at a
specific time

             Major types of vesselsa Containerships and LASH    Bulk vessels, 
                                                                   break-bulk 
                                                    vesselsb vessels, and     
                                                             tug-barge        
                                                                      vessels 
              Types of food aid that                                          
                                 are                  Bagged  Bulk and bagged
                             carried                         
                   Number of vessels             47 vesselsc      143 vessels 
                 Number of companies             5 companies     38 companies 

Source: GAO analysis of data from USDA and MARAD.

aThis table is a simplification that places vessels into two broad
categories. There are exceptions such as some non-MSF vessels that offer
liner service and some containerships and LASH vessels, operated by MSF
carriers, that are not enrolled in MSF.

bLASH, or Lighter Aboard Ships, are barge-carrying vessels that use barges
like containers.

cThis number reflects 42 containerships and 5 LASH vessels that
participated in MSF over a 5-year period and carried food aid. Currently,
MSF comprises 36 containerships and 1 LASH vessel that carried food aid.

Most charter service vessels are operated by non-MSF carriers. Charter
service means that vessels are hired to carry a cargo to specific ports at
a specific time; these vessels do not provide regularly scheduled service
on a fixed route but typically carry a shipload of cargo for only one or a
few customers at a time. Charter service is primarily provided by bulk,
breakbulk, and tug-barge vessels that can carry either bulk or bagged
cargo. Bulk vessels are designed to carry dry bulk commodities, such as
rice or wheat, in large interior holds. The benefit of bulk shipments is
the economies of scale that can be gained from shipping large amounts of a
single commodity. Figure 1 shows a photograph of a bulk vessel. Break-bulk
vessels are general cargo ships that are designed to carry nonuniform
items packaged as single parcels or assembled together on pallet boards.
Bagged commodities are stacked and secured within interior holds of the
ship. Tugbarge vessels have a tugboat or towboat that propels a separate
barge by pushing or towing it. Barges generally carry bulk or break-bulk
cargo, although some also carry containerized cargo.

Figure 1: A Bulk Vessel

Source: MARAD.

Most MSF vessels are liner service vessels. Liner service means that
vessels have regularly scheduled sailings on fixed routes. These vessels
typically carry small amounts of cargo for many customers at one time and
will sail, even if not completely full. Liner service is primarily
provided by containerships that carry bagged cargo; they do not carry bulk
cargo. Containerships are designed to carry cargo in standard-size,
preloaded containers that are stacked next to and on top of each other on
the ship. The benefit of containers is that they permit rapid loading and
unloading and efficient transportation of cargo to and from the port area.
Containers facilitate intermodal transportation because they can be loaded
by the supplier and sealed, taken by truck or railcar to the port, then
loaded onto the containership without the cargo being handled. In the case
of food aid, generally the suppliers do not load the containers but
instead ship bagged commodities by rail or truck to the port of loading,
where they are loaded into the containers. Figure 2 shows a photograph of
a containership. Liner service is also provided by Lighter Aboard Ships
(LASH), which are bargecarrying vessels that use barges like containers.
They are also intermodal because the barges can use rivers and canals to
pick up and drop off cargo at interior loading docks.

                           Figure 2: A Containership

                                 Source: MARAD.

Cargo Preference and Maritime Security Programs Intended to Support
U.S.-Flag Fleet

The cargo preference and Maritime Security Programs are both intended to
bolster the U.S.-flag market share in international commerce, as well as
to ensure the availability of an adequate number of U.S.-flag ships and
U.S.citizen mariners in the event of a national defense need. The cargo
preference laws are part of the overall statutory program to support the
privately owned and operated U.S.-flag commercial fleet, or merchant
marine.8 DOD and MARAD consider the merchant marine vital to U.S. national
security, providing essential sealift capability in wartime. The ships
that carry these cargoes also provide jobs for American seafarers who are
available in time of national emergency to crew the sizable fleet of
reserve government vessels. As an agency of the Department of
Transportation, MARAD's responsibilities include promoting the development
and maintenance of the U.S. merchant marine. It administers both the cargo
preference and Maritime Security Programs.9

The Maritime Security Program is more targeted than the cargo preference
program in terms of the vessels that can participate. It is intended to
guarantee that certain kinds of militarily useful ships and their crews
will be available to DOD in a military contingency. Under the renewed
program starting in 2005, DOD must approve the proposed vessels as
militarily

8A 1954 amendment to the Merchant Marine Act of 1936 (P.L. 83-664),
requires that at least 50 percent of the gross tonnage of all
government-generated cargo be transported on privately owned, U.S.-flag
commercial vessels to the extent such vessels are available at fair and
reasonable rates. In 1985, the Merchant Marine Act of 1936 was amended to
require that the percentage of certain agricultural cargoes to be carried
on U.S.-flag vessels be increased from 50 to 75 percent. The Cargo
Preference Act of 1904 requires 100 percent of supplies bought for U.S.
military departments to be carried on U.S.-flag vessels available at rates
that are not excessive or otherwise unreasonable. Finally, Public
Resolution 17 of the 73rd Congress calls for all cargoes, resulting from
loans from federal government agencies, such as the Export-Import Bank of
the United States, to be shipped exclusively on U.S.-flag vessels, unless
MARAD grants a waiver.

In addition, Section 901(b) of the Merchant Marine Act of 1936, as
amended, prohibits vessels built or rebuilt outside the United States or
under foreign registry from carrying preference cargo subject to the Act
for 3 years, unless the vessel is in MSF.

9For cargo preference, MARAD establishes the regulations governing how the
cargo preference laws are to be implemented by federal agencies. MARAD
also exercises oversight over how these regulations are carried out in
practice by federal agencies, such as when USAID and USDA determine ocean
transportation for food aid shipments. It also reimburses USAID and USDA
for a portion of the higher costs related to using U.S.-flag vessels. For
the Maritime Security Program, MARAD screens applicants to MSF,
establishes operating agreements with the carriers that are selected, and
provides their monthly subsidy payment.

useful. The program's main focus has been to enable globally competitive
carriers that operate militarily useful vessels to enter or keep U.S.-flag
status. Most MSF vessels are containerships, operated by some of the
largest containership carriers in the world. For instance, MSF carriers
Maersk Sealand, P&O Nedlloyd, and APL were ranked among the top four
containership carriers by volume as of May 2001, according to the Bureau
of Transportation Statistics. These containership carriers have intermodal
systems that also come as part of the package, allowing DOD to benefit
from private sector global transportation and communication networks.
According to MARAD, these networks include not only vessels, but also
logistics management services, infrastructure, terminals and equipment,
communications and cargo-tracking networks, and thousands of trained,
professional U.S.-citizen mariners and shoreside employees located
throughout the world.

The Maritime Security Program also results in the reflagging of new and
more efficient vessels to U.S. registry for participation in MSF. The
program requires that vessels be less than 15 years old to participate
(except that LASH vessels can be 25 years old). From its implementation in
1996 through 2002, a total of 18 modern commercial liner vessels, with an
average age of less than 9 years, were reflagged to U.S. registry for
participation in MSF, according to MARAD.

Food Aid Programs Administered by USAID and USDA

USAID and USDA's Foreign Agricultural Service are responsible for
administering the food aid programs that provide humanitarian food
assistance to countries in need. The food aid programs had an annual
average budget of $1.97 billion during fiscal years 1999 to 2003,
according to USDA. The primary mechanism through which the U.S. government
implements its international food assistance initiatives is P.L. 480.10
Food assistance provided under P.L. 480 is delivered to foreign countries
through three separate programs: Titles I, II, and III. USDA administers
Title I, which provides for government-to-government sales of agricultural
commodities to developing countries on credit terms or for local
currencies. USAID administers Titles II and III. Title II provides for
donation of U.S. agricultural commodities to meet emergency and
nonemergency food needs in other countries, and it is by far the largest
of the food aid programs. Title III provides for government-to-government
grants to support long-term growth in the least developed countries but
has been inactive in recent years. In addition to P.L. 480, food aid is
provided through three smaller programs administered by USDA's Foreign
Agricultural Service: Food for Progress, section 416(b), and the
McGovern-Dole International Food for Education and Child Nutrition
Program.11

Cargo Preference and Maritime Security Programs Provide Incentives to
Retain U.S.-Flag Ships and Mariners

The cargo preference and Maritime Security Programs both provide
incentives to retain privately owned U.S.-flag ships and their
U.S.-citizen mariners for commercial and national defense purposes. Cargo
preference makes available a protected market that provides the economic
incentive for vessel owners to pay the higher costs associated with the
U.S. flag and employ U.S.-citizen crews. We found that a total of 190
privately owned U.S.-flag vessels carried cargo preference food aid
shipments at some point during the fiscal year 1999 to 2003 period. In
addition, the Maritime Security Program provides a subsidy for MSF
carriers with particular militarily useful vessels. MSF currently has 47
ships, of which 37 have participated in cargo preference food aid
shipments. DOD strongly

10The Agricultural Trade Development and Assistance Act of 1954, P.L.
83-480, 68 Stat. 454.

11Food for Progress provides for the donation or credit sale of U.S.
commodities to developing countries and emerging democracies to support
democracy and an expansion of private enterprise. Section 416(b) provides
for overseas donations of surplus USDA commodities. The McGovern-Dole
International Food for Education and Child Nutrition Program provides for
donations of U.S. agricultural products, as well as financial and
technical assistance, for school feeding and maternal and child nutrition
projects in lowincome, food-deficit countries that are committed to
universal education.

supports both programs and said it has benefited from both during the
recent wars in Afghanistan and Iraq.

Cargo Preference Program Provides Protected Market as Incentive for
U.S.-Flag Registry

Preference cargoes are intended to provide the economic incentive for
vessel owners to pay the higher costs associated with U.S.-flag registry
and employ U.S.-citizen crews. According to MARAD, due to high U.S. labor
costs; safety, health, and environmental regulations; and taxes, it is
more expensive for vessels to be U.S.-flagged. For instance, U.S.-flag
vessels generally incur higher labor costs due to higher manning level
requirements, as well as higher wages and benefits for U.S.-citizen
mariners.

The cargo preference laws, by guaranteeing the availability of cargo to
U.S.flag ships, contribute to the financial viability of U.S.-flag vessel
operating companies, thereby helping to ensure that the vessels, trained
crews, and vessel service industries continue to exist, according to
MARAD.12 The cargo preference program provides this incentive by reserving
a portion of the U.S. market for U.S.-flag vessels, despite the higher
prices they typically charge. In the food aid transportation market, a
minimum of 75 percent of food aid shipments must be shipped on U.S.-flag
vessels.13

The U.S.-flag vessels (both MSF and non-MSF) participating in cargo
preference food aid shipments during fiscal years 1999 to 2003 comprised a
variety of vessel types. According to our analysis of USDA data, a total
of 190 individual vessels participated in food aid shipments at some point
between 1999 and 2003.14 This included 111 bulk, break-bulk, tug-barge,
and

12One reason for MARAD support of the U.S.-flag fleet is concern about its
continuing decline. From 1983 to 2003, the number of active privately
owned ships in the U.S.-flag fleet declined by about half and their
carrying capacity fell by about 25 percent, according to MARAD data.
Another concern is related to the declining size of the U.S.-citizen
mariner pool.

13However, cargo preference has also meant that U.S. government-sponsored
cargoes, including food aid, have been required to be shipped on
higher-cost vessels, resulting in historically higher government
transportation costs. See GAO, Cargo Preference Requirements: Objectives
Not Significantly Advanced When Used in U.S. Food Aid Programs,
GAO/GGD-94-215 (Washington, D.C.: Sept. 29, 1994) and GAO, Maritime
Industry: Cargo Preference Laws-Estimated Costs and Effects,
GAO/RCED-95-34 (Washington, D.C.: Nov. 30, 1994).

14There was an annual average of 108 U.S.-flag vessels that carried food
aid during these years.

tanker vessels that provided charter service and 79 containership, LASH,
and other vessels that provided liner service. These vessels were operated
by 38 carrier companies.

We found that the level of dependence on food aid varied significantly
among carriers. We interviewed representatives of 15 of the top carriers
that participated in U.S.-flag cargo preference food aid shipments during
1999 to 2003, comprising 77 percent of food aid revenues. Of the 10
non-MSF carriers we interviewed that generally provided charter service, 4
said that 60 percent or more of their annual revenues came from food aid
shipments, 3 said between 20 and 50 percent, and 3 said less than 10
percent came from these shipments. Most of the five MSF carriers we
interviewed that provided liner service said that food aid revenues
comprised a small percentage of their total revenues.

Maritime Security Program Provides Subsidy for MSF

The Maritime Security Program was authorized for fiscal years 1996 to
200515 and provides about $100 million in annual funding for up to 47
vessels to participate. Each participating vessel receives an annual
subsidy payment of $2.1 million,16 intended to partially offset the higher
operating cost of keeping these vessels under U.S.-flag registry. In
November 2003, Congress passed another 10-year authorization for the
Maritime Security Program,17 starting in fiscal year 2006, that would
expand the program from 47 to 60 vessels. Annual subsidy payments were
increased from a flat $2.1 million payment to an escalating payment of
$2.6 million for fiscal years 2006 to 2008, $2.9 million for fiscal years
2009 to 2011, and $3.1 million for fiscal years 2012 to 2015, always
subject to the availability of congressional appropriations.

According to MARAD officials and MSF carrier representatives we
interviewed, the combination of MSF subsidy and access to cargo preference
shipments, including food aid shipments, enables these containership
carriers to stay in MSF and creates incentives to reflag newer vessels.
While most MSF carriers primarily carry commercial cargo, MSF

15Maritime Security Act of 1996, P.L. 104-239, 110 Stat. 3118.

16The exception was that the subsidy payment was $2.3 million for the
first year of the program, 1996. However, according to MARAD, the Act was
not signed until the end of the first year, so the $2.3 million was never
paid.

17P.L. 108-136, 117 Stat. 1803.

carrier representatives said that they need both MSF subsidy and cargo
preference food aid shipments to offset the higher costs of operating as a
U.S.-flag vessel. MARAD stated in its 2002 annual report that the current
$2.1 million subsidy represents about 13 percent of the cost of operating
a U.S.-flag vessel. According to a MARAD official, the subsidy only
partially offsets the higher cost of employing U.S.-citizen mariners.
However, during the interviews, MSF carrier representatives said that the
subsidy was important to them because it was a guaranteed monthly payment
that provided a level of financial stability.

MSF currently comprises 47 vessels operated by 12 companies, based on data
as of December 2003.18 These vessels include 38 containerships, 1 LASH,
and 8 roll-on/roll-off vessels.19 Of the vessels currently participating
in MSF, 36 containerships and 1 LASH vessel participated in cargo
preference food aid shipments during fiscal years 1999 to 2003.
Approximately 2,162 mariners are employed on these ships, according to
MARAD.20 (See app. III for a profile of the current MSF participants.)

DOD Supports Both Cargo DOD strongly supports both the cargo preference
and Maritime Security Preference and Maritime Programs. DOD officials said
that DOD's priority is to maintain or increase Security Programs the
current level of U.S.-flag ships and mariners and, therefore, it strongly

supports both programs. Through the cargo preference and Maritime Security
programs, an additional manpower pool is maintained that DOD can draw on
to crew the reserve fleet.21 DOD officials said that the Maritime Security
Program, in addition to guaranteeing militarily useful

18These 12 operating companies were owned by 8 carriers. Of these 8
carriers, 5 had participated in cargo preference food aid shipments.

19Roll-on/roll-off ships are vehicle carriers that allow vehicles such as
cars, trucks, or tanks to drive on and off the ship. They do not
participate in food aid shipments.

20According to MARAD, MSF ships have approximately 940 mariner billets.
MARAD calculates that each billet supports 2.3 mariners based on current
industry ratios, so that the number of employed mariners would total
2,162. However, DOD calculates that each billet supports 2 mariners in its
contingency planning, which would result in a total of 1,880 mariners.

21The reserve fleet comprises (1) the Ready Reserve Force, 59 standby
reserve ships maintained by MARAD and (2) DOD's Surge Force, 8 fast
sealift ships and 11 large, mediumspeed roll-on/roll-off ships. Reserve
fleet vessels are kept ready with a skeleton crew but need additional
mariners to make a full complement when activated for service.

U.S.-flag ships and trained U.S.-citizen mariners, provides access to MSF
liner carriers' intermodal systems, which is important to DOD.

In testimony before the House Armed Services Committee on October 8, 2002,
General John W. Handy, Commander of the U.S. Transportation Command,
strongly supported reauthorization of the Maritime Security Program. He
stated that DOD limited its sealift fleet22 to those assets that the
commercial sector could not provide, so that only 33 percent of the
vessels DOD may require resided in its own fleet. The remainder of the
sealift capacity, needed to transport military equipment and supplies,
came from the commercial sector.

DOD officials said that it had benefited from both programs during the
Afghanistan and Iraq wars. For example, during Operation Iraqi Freedom,
DOD did not need to pull MSF vessels out of their normal commercial
service. Instead, it chartered two MSF roll-on/roll-off vessels for DOD
use and used the other MSF vessels in their normal commercial routes,
where appropriate, to meet its needs, according to a DOD official. This
official said that DOD preferred to leave MSF vessels in their normal
commercial service because then DOD would also be able to benefit from use
of their global intermodal systems. MSF carriers may have had to displace
some commercial cargo but otherwise continued business as usual. During
the period January 1 to October 14, 2003, MSF vessels made 135 vessel
voyages of cargo to sustain the Iraqi deployment. This effort included 35
MSF containerships. These vessels transported a total of 8,668 twenty-foot
equivalent units (TEU),23 according to DOD data. According to MARAD, more
than 7,500 merchant mariners served in Operation Iraqi Freedom. Of these,
about 1,470 mariners served on MSF vessels, based on a DOD estimate.

22DOD prepositioned sealift ships are kept fully loaded with military
material and fully crewed at all times to be ready to embark immediately
to a war zone in the initial surge of a deployment.

23A TEU is the unit of measure for containerized cargo. One 20-foot
container would be one TEU, while one 40-foot container would be two TEUs.

MSF and Non-MSF Carriers Compete for Bagged Food Aid Shipments

MSF and non-MSF carriers compete only for bagged food aid shipments
because MSF vessels do not carry bulk food aid. Although the majority of
food aid continues to be shipped as bulk cargo, bulk food aid shipments
decreased from fiscal years 1999 to 2003, partly because of changes in
food aid spending. The recent decline in bulk cargo has caused some
non-MSF bulk carriers to rely more on bagged cargo. Non-MSF carriers
transported about 55 percent and MSF carriers 45 percent of bagged food
aid shipments from fiscal years 1999 to 2003. Cargo preference
requirements affect whether agencies award bagged food aid shipments to
MSF or non-MSF carriers. Most of the bagged food aid cargo carried by MSF
and non-MSF vessels is loaded for export at U.S. ports in the Gulf of
Mexico.

Increasing Share of Food Aid Is Shipped as Bagged Cargo

Although on average approximately 67 percent of food aid was shipped as
bulk cargo and 33 percent as bagged cargo24 from fiscal years 1999 to
2003, the share of food aid shipped as bagged cargo generally increased
during these years. This change was due mostly to a decline in USDA's
purchases of bulk agricultural commodities for the food aid program.
Although USDA purchases of bulk commodities remained relatively stable
from 1996 to 1998, they increased dramatically in 1999 and then declined
steadily from 1999 to 2003.25 As the procurement data in figure 3 show,
purchases of bulk commodities decreased from 5.76 million tons in 1999 to
2.39 million tons in 2003. However, purchases of bagged commodities
equaled about 2 million tons each year during this period. Thus, the
percentage of commodities procured that were bagged increased from 26
percent in 1999 to 46 percent in 2003.

24Bagged cargo includes agricultural commodities that are packaged or
processed. Processed commodities are packaged in bags, tins, or other
containers.

25In 1999, bulk commodity purchases were augmented to provide food aid to
Russia. This donation to Russia was one of the largest single food aid
transfers in U.S. history.

Figure 3: USDA Purchases of Bulk and Bagged Agricultural Commodities for
the Food Aid Program, Fiscal Years 1996-2003

Tons in millions

6

5

4

3

2

1

0 1996 1997 1998 1999 2000 2001 2002 2003

Year

Bulk commodities

Bagged commodities

Source: GAO analysis of USDA data.

Note: Figure is based on procurement data. USDA shipment data are
available from fiscal years 1999 to 2003. They show the same pattern as
the procurement data for those available years. However, the shipment
amounts are generally smaller than the procurement amounts primarily
because food aid cargo is often damaged while it is being handled in
preparation for export shipment, according to USDA officials.

Changes in food aid spending from fiscal years 1999 to 2003 have
contributed to this shift from bulk to bagged cargo. The largest food aid
program is administered under P.L. 480 Title II, which experienced an
increase in funding from 2000 to 2003. The P.L. 480 statute requires that
at least 75 percent of agricultural commodities donated for development,
or nonemergency, purposes be value-added. Value-added commodities are
shipped as bagged cargo, as opposed to bulk.26 Aside from Title II
development assistance, many of the commodities donated by the Food for
Education and section 416(b) food aid programs also have been shipped as
bagged cargo in recent years, according to USDA officials. However,
spending for the P.L. 480 Title I food aid program generally declined from
1999 to 2003.27 Most commodities sold under the P.L. 480 Title I program
are shipped as bulk cargo, such as wheat, corn, and soybeans.

Reduction in Bulk Food Aid Cargo Leads to Increased Reliance on Bagged
Food Aid Cargo

Many non-MSF carriers depend on cargo preference food aid shipments for a
large share of their business; therefore, the decline in bulk cargo has
meant increased reliance on bagged cargo shipments. According to the
interviews we conducted with non-MSF carriers, some non-MSF carriers that
traditionally ship bulk food aid reacted to the decline in bulk food aid
shipments by increasing their participation in bagged food aid shipments.
Figure 4 shows that while total shipments of food aid by non-MSF vessels
decreased over the fiscal year 1999 to 2003 period, the decline in bulk
food aid shipments was partially offset by an increase in bagged food aid
shipments. Among non-MSF carriers that have shipped bulk food aid, 43
percent have also shipped bagged food aid.

26Value-added commodities are processed, fortified, or bagged, and include
such food products as wheat flour, cornmeal, corn-soy blend, and vegetable
oil.

27Title I spending also declined from 1992 to 1998, before the spike in
1999 when food aid was provided to Russia.

Figure 4: Non-MSF Total Food Aid Shipments, Fiscal Years 1999-2003

Tons in millions 5

4

3

2

1

0

1999 2000 Year

Bagged cargo

Bulk cargo Source: GAO analysis of USDA data.

a2003 data are estimated.

2001 2002 2003a

MSF and non-MSF vessels combined carried a total of 6.73 million metric
tons of bagged food aid cargo and earned an average of $430 million each
year from bagged food aid shipments from 1999 to 2003.28 MSF vessels
carried about 45 percent of this cargo and non-MSF vessels carried 55
percent. However, non-MSF carriers' share of the bagged food aid market
was clearly greater in 2002 and 2003 than in the previous 3 years, as
shown in figure 5.

28If foreign-flag vessels are included, the total is $499 million.

  Figure 5: Non-MSF and MSF Bagged Food Aid Shipments, Fiscal Years 1999-2003

                          Tons in thousands 1,000 800

                                      600

                                      400

                      200 0 1999 2000 2001 2002 2003a Year

                                  Non-MSF MSF

                       Source: GAO analysis of USDA data.

a2003 data are estimated.

The MSF cargo was shipped by five companies: four operating 42
containerships and one operating 5 LASH vessels. Each MSF containership
carried an average shipment of 950 tons per voyage, and each MSF LASH
vessel carried an average shipment of 22,440 tons per voyage. The non-MSF
cargo was shipped by 38 companies that operated 143 vessels. Each non-MSF
vessel carried an average shipment of 1,750 tons of bagged cargo per
voyage, almost twice the average shipment of each MSF containership.

Cargo Preference Cargo preference requirements affect the results of
competition between Requirements Affect MSF and non-MSF carriers for food
aid shipments. One requirement that Whether Food Aid has tended to favor
non-MSF carriers is MARAD's interpretation of U.S.-flag

service for the cargo preference program. Figure 6 outlines the
criteriaShipments Are Awarded to agencies are required to follow when
awarding shipments subject to cargoMSF or Non-MSF Carriers preference
laws. As the figure indicates, an ocean carrier that offers to

carry preference cargo on a U.S.-flag vessel can be counted as either

Priority 1 or Priority 2 service. For example, a U.S.-flag vessel
qualifies for Priority 1 service if it offers to transport preference
cargo on a U.S.-flag vessel or transship the cargo to U.S.-flag vessels
for the entire portion of the waterborne voyage. However, a U.S.-flag
vessel would qualify for Priority 2 service if it transshipped the cargo
to a foreign-flag vessel for any leg of the voyage. In the absence of
Priority 1 service availability, agencies may also count Priority 2 as
Priority 1 service by default.

Figure 6: Criteria for MARAD's Prioritization of U.S.-Flag Vessel Service
for the Cargo Preference Program

Priority 1

The following U.S.-flag vessel services have equal status in the selection
by shippers of preference cargoes:

A B

C

Priority 2

Direct U.S.-flag vessel service

U.S.-flag vessel service with transshipment to another U.S.flag vessel to
the final discharge port

Intermodal services to the final destination from the point or port of
origin utilizing only U.S.-flag vessels for any waterborne portion of the
voyage.

U.S.-flag vessel service with transshipment via a foreign-flag vessel to
the final discharge port.

Priority 3

Foreign-flag vessel service.

Source: MARAD.

Most non-MSF vessels qualify for Priority 1 service because they offer the
food aid program charter service entirely on a U.S.-flag vessel. However,
vessels that operate in liner service, such as MSF containerships, often
qualify for Priority 2 service because they transfer shipment of
(transship) food aid cargo to a foreign-flag vessel for a leg of the
voyage. In some locations, however, some MSF carriers have started to
transship food aid cargo to prepositioned U.S.-flag vessels instead of
foreign-flag vessels so that they can qualify as Priority 1 service. In
fact, as figure 7 shows, liner vessels that carried Title II food aid
cargo from fiscal years 1999 to 2003 qualified for Priority 1 service
about 48 percent of the time.29 Liner vessels counted as Priority 1
service by default about 23 percent of the time and Priority 2 or 3
service about 29 percent of the time. Under the cargo preference program,
agencies are required to award food aid shipments to carriers that offer
Priority 1 service over carriers that offer Priority 2 or 3 service, even
if the freight rate charged by the carrier offering Priority 1 service is
higher, unless the rate exceeds MARAD's fair and reasonable rate
calculation.

29This estimate is for food aid shipments by tonnage from both
containership and LASH carriers with U.S.-flag liner service and
consequently includes shipments from some non-MSF carriers. Title II food
aid data only records the carrier rather than the vessel name.

Figure 7: Estimated Share of Title II Food Aid Shipments by U.S.-Flag
Liner Carriers at Each Priority Level, Fiscal Years 1999-2003

Priority 1 service by default

Priority 2 or 3 service

Priority 1 service

Source: GAO analysis of USAID and USDA data.

Other cargo preference requirements tend to favor MSF carriers. An example
of a requirement that has benefited MSF carriers is section 17 of the
Maritime Security Act of 1996.30 This provision allocates up to 25 percent
of the total tonnage of Title II bagged food aid cargo each month to Great
Lakes ports. Moreover, shipments of this cargo are awarded to carriers
without regard to the flag of the vessel offering service and therefore
are not subject to MARAD's priority rules.31 From fiscal years 1999 to
2003, MSF vessels and foreign-flag vessels carried an estimated total of
221,000 tons and 379,000 tons of this cargo, respectively.32 MSF carriers
have shipped much of this cargo because they have incorporated certain
Great Lakes ports facilities into their intermodal networks. They have
created a system for transporting this cargo intermodally in containers by
rail to U.S. ports on the East and West Coast, where the cargo is
ultimately exported.33 MSF carriers have been successful in winning much
of this cargo because these intermodal shipments allow them to offer
competitive freight rates, according to USAID and USDA officials. However,
non-MSF carriers ship this cargo less often than MSF carriers because they
generally lack access to the intermodal infrastructure that enables MSF
carriers to move this cargo efficiently.34

MSF and Non-MSF Carriers U.S. Gulf ports handled about 70 percent of the
average annual tonnage of Load Most Bagged Food Aid bagged food aid cargo
carried by MSF and non-MSF vessels from fiscal Cargo from U.S. Gulf Ports
years 1999 to 2003. Table 2 shows the tonnages of bagged food aid cargo

loaded by MSF and non-MSF vessels at major food aid ports. As the table
indicates, the ports of Lake Charles and Jacintoport handled 1.72 million
tons and 1.43 million tons of bagged food aid cargo from 1999 to 2003,
respectively. These two ports handle bagged food aid mostly as break-bulk,

30P. L. 104-239, 110 Stat. 3138.

31Agencies are required to count all vessels that carry this cargo as
Priority 3 service, or foreign-flag vessel service, according to USDA and
USAID officials.

32These data reflect an estimate because USDA's food aid shipment database
does not designate this cargo differently from other food aid cargo that
is shipped from Great Lakes ports.

33Few U.S.-or foreign-flag carriers offer to transport food aid cargo by
sailing on the Great Lakes.

34According to USAID, non-MSF carriers also indirectly benefit from this
cargo because it is treated as foreign flag, even though it is mostly
carried by U.S.-flag vessels, thereby leaving a higher proportion of
U.S.-flag cargo available for the non-MSF carriers.

or noncontainerized, cargo. Lake Charles is an agricultural port that is
also the only U.S. port approved by USDA to store prepositioned
commodities for the food aid program. Jacintoport has an automated cargo
handling system capable of loading large tonnages of bagged food aid into
break
bulk vessels and bulk vessels at a high rate of speed. Lake Charles will
soon have a similar machine with like capabilities. MSF carriers do not
load food aid directly into their vessels from these two ports. Instead,
they hire stevedores to stuff the food aid cargo into containers and then
move the containers intermodally by barge or rail to nearby ports that
have container terminals where they have regularly scheduled service, such
as the ports of New Orleans and Houston. MSF carriers run a similar
operation from the Port of Chicago, where most of the Title II bagged food
aid cargo subject to section 17 of the Maritime Security Act of 1996 is
loaded. The Port of Chicago handled on average an estimated 35,000 tons of
bagged food aid cargo for MSF carriers each year from 1999 to 2003. Much
of this cargo was transported intermodally by rail to major U.S. container
ports, such as Charleston, South Carolina; Norfolk, Virginia; and Seattle,
Washington.

Table 2: Tonnage of Bagged Food Aid Cargo Loaded by MSF and Non-MSF
Vessels at Major Food Aid Ports, Fiscal Years 1999-2003

                                    Port        Tonnage      Percent of total 
                 Lake Charles, Louisiana      1,716,035 
                      Jacintoport, Texas      1,432,957 
                  New Orleans, Louisiana        799,382 
                          Houston, Texas        346,228 
              Charleston, South Carolina        341,118 
              Litco Memphis, a Tennessee        276,057 
                       Norfolk, Virginia        270,411 
                       Chicago, Illinois        264,121 
                         All other ports        819,402 
                                   Total      6,265,711                   100 

Source: GAO analysis of USDA data.

Note: Table does not include estimated data for 2003. Percentages do not
add up to 100 due to rounding.

aLitco Memphis is a facility owned by the company that operated the five
LASH vessels in MSF that carried food aid from 1999 to 2003.

Multiple Factors Determine How Tonnage Limit Changes Would Impact Food Aid
Carriers

Our analysis of data from program agencies and carriers suggests that
establishing a bagged tonnage limitation could reduce MSF vessels' market
share in food aid, but the extent will depend on the limitation level and
the options MSF carriers have in responding to it.35 Using recent data, we
examined daily limits of 7,500, 5,000, and 2,500 tons and found that the
percentage of MSF food aid voyages affected rises from 3 percent at a
limit of 7,500 tons to 19 percent at a limit of 2,500 tons. Almost all
voyages above 7,500 tons were on the specialized LASH vessels, of which
only one remains in MSF. Total annual food aid for MSF containerships on
voyages above a 2,500-ton limit was around 160,000 tons. However, setting
a limit at this level may not mean a reduction of 160,000 tons that MSF
vessels carry, to the extent they are able to continue to carry some food
aid on affected voyages, replace some food aid with other cargo, and
forfeit their subsidy for food aid shipments that are sufficiently
profitable. A simulation analysis we performed for MSF containerships
suggests that, at a limit of 2,500 tons for example, the total annual
decrease in food aid carried by these vessels could, under certain
assumptions incorporating those options, range from about 17,000 to about
63,000 tons. Structured interviews with the carriers suggest that
considerations such as vessel sharing arrangements could also affect the
outcome and impacts on non-MSF carriers may depend on their market niche.
Further, if the terms of MSF and non-MSF carriers' participation in cargo
preference change, program agencies are concerned that they could face
increased delivery delays, administrative burdens, and shipping costs. The
major food aid ports would generally experience a limited impact on their
overall port activity from a bagged tonnage limit, although specific food
aid terminals could potentially be affected, depending on the extent of
any limitation and the MSF carriers' responses to it.

35Increasing or decreasing a bulk tonnage limitation on current MSF
vessels will have no effect because these vessels currently do not carry
bulk cargo. However, according to MARAD, the new Maritime Security Program
allots space for five tanker vessels that would be impacted by a bulk
tonnage limitation, and would therefore have a disincentive to join MSF.

Share of MSF Food Aid Voyages Affected Would Increase Substantially as
Limitation Level Decreases

While more than 80 percent of MSF food aid voyages fall below a 2,500-ton
limit, establishing a limit at 2,500 tons would be substantially more
constraining for the majority of the fleet than limits at 5,000 or 7,500
tons. According to USDA data from fiscal years 2001 to 2003, only 3
percent of MSF food aid voyages carried more than 7,500 tons, almost all
of which occurred on the five LASH vessels that have participated in the
MSF. 36 However, another 16 percent of MSF food aid voyages carried food
aid tonnage between 2,500 and 7,500 tons. All of these voyages occurred on
containerships, which comprise the majority of current MSF vessels. Figure
8 shows the number of MSF food aid voyages at different tonnage levels.
The average annual tonnage carried by both MSF LASH vessels and
containerships on voyages in excess of 2,500 tons was around 322,000 tons,
of which around 160,000 tons were carried on the containerships.37

36Almost all LASH voyages carried in excess of 7,500 tons of food aid such
that establishing a tonnage limitation at any of the proposed levels would
affect these vessels. However, only one LASH vessel remains in MSF.

37This figure represents a 3-year annual average for data from fiscal
years 2001 through 2003.

Figure 8: Number of MSF Food Aid Voyages at Different Tonnage Levels,
Fiscal Years 2001-2003

Number of voyages 300

250

200

150

100

50

0 Tonnage

Source: GAO analysis of USDA food aid data.

Similar to the percentage of MSF food aid voyages, the share of MSF food
aid revenues affected by a tonnage limit rises substantially as the level
is decreased. As shown in figure 9, 37 percent of MSF total food aid
revenues were earned on voyages carrying more than 7,500 tons of food aid
while 68 percent of MSF total food aid revenues were earned on voyages
with more than 2,500 tons of food aid. In comparison to the percentage of
voyages affected, these revenue shares reflect that MSF voyages above a
potential tonnage limit are earning proportionally more food aid revenues
than those with smaller cargo volumes. MSF food aid revenues earned on the
primarily LASH vessels that carried more than 7,500 tons were around $26
million annually, or $8.5 million per vessel. Not including LASH vessels,
MSF food aid revenues earned on containerships that carried more than
2,500 tons were around $22 million annually, or $1.3 million per vessel.
Nonetheless, while these data indicate how often an MSF vessel could be
restricted by a tonnage limitation, they indicate the potential loss in

revenue from food aid only under the assumption that MSF carriers were to
no longer carry any food aid on these voyages.

Figure 9: Percentage of MSF Food Aid Revenues from Voyages with Tonnage
above a Potential Limitation, Fiscal Years 2001-2003

                           Percentage of revenues 80

                                      $47m

                                       70

The actual food aid tonnage and net revenue impact for MSF vessels under a
tonnage limitation will depend on options available to the carriers and
how they respond to them. Numerous considerations relating to market
conditions, food aid logistics, and carrier characteristics would
ultimately shape the impact of any tonnage limitation. We identified three
factors to explicitly consider in an analysis of a tonnage limitation.
Each of these factors, under certain assumptions, has the potential to
make the impact of a tonnage limit on MSF vessels smaller than suggested
by the share of MSF voyages affected.

                                       60

                                       50

                                       40

                                       30

                                       20

                                      10 0

2,500 5,000 7,500

Potential tonnage limitation level

Source: GAO analysis of USDA food aid data.

Various Factors Will Affect Actual Impact on MSF

First, affected MSF vessels might be able to carry some food aid,
potentially up to the level of the limit, and may not have to give up the
entire tonnage for that voyage to keep their subsidy. This situation can
occur if a carrier can bid on a portion of an offered shipment or if the
food aid tonnage on a voyage comprises multiple shipments, such that the
carrier could bid on those shipments providing tonnage under the limit.38
For example, a carrier that would normally have a voyage with 3,700 tons
of food aid may be able to carry two food aid contracts for 1,000 tons
each and, under a tonnage limit of 2,500 tons, face a potential loss of
food aid cargo of only 1,700 tons.

Second, depending on market conditions, affected MSF carriers may be able
to replace a portion of the food aid above the limit with commercial or
nonfood preference cargo, diminishing their loss in total revenues. For
example, if an MSF vessel were to carry 1,700 tons less food aid due to a
tonnage limit, the carrier may be able to replace a portion of that
tonnage with nonfood aid cargo.

Third, there may be occasions when carrying food aid cargoes above a
tonnage limit is more profitable than reducing food aid to receive the
subsidy for a voyage, thus providing an incentive to carry food aid above
the limit. This may occur for food aid shipments that are particularly
large or earn a particularly high freight rate such that an affected MSF
carrier might choose to carry the food aid, even if it entailed forfeiting
the subsidy otherwise earned during the days of that voyage, as well as
forgoing any net revenues from available replacement cargo.39 For example,
if an MSF vessel were to normally carry 7,000 tons of food aid on a voyage
that lasted 15 days, the carrier would have to give up a subsidy payment
of around $107,000 to carry that entire tonnage. The carrier might choose
to forfeit

38According to USDA and USAID, the food aid tonnage on both MSF and
non-MSF vessel voyages is often comprised of multiple food aid contracted
shipments, often from different agencies and programs, such that a carrier
could decrease the food aid tonnage carried on a voyage. With regard to
carriers' ability to bid on a portion of a contract, program agencies,
carriers, and industry representatives had different opinions. MARAD
officials stated that contracts are typically structured to require
carriage of the total tonnage with specific arrival dates, which prevents
the splitting of cargoes. USAID and USDA noted that carriers' flexibility
is limited because contract terms are primarily determined by the private
voluntary organizations that seek the food aid grants, however, they
acknowledged that bidding on a portion of a contract was possible.

39The five MSF carriers reported that they would not give up their subsidy
to carry food aid on any voyage -a view that was corroborated by MARAD.
However, USDA acknowledged that this practice could be financially
beneficial for large voyages.

the subsidy payment if the net revenues from the food aid effectively
above the 2,500-ton limit exceeded the $107,000 plus potential net
revenues from replacement cargo.

To illustrate the impact of a tonnage limitation when accounting for these
three factors, we created a simulation model that suggests ranges of
possible tonnage and net revenue changes for MSF vessels at different
tonnage limits. The model uses estimates of average freight rates, average
cargo volumes, and average vessel costs for voyages from fiscal years 2001
to 2003, and includes probability distributions that reflect certain
assumptions about carrier options and behavior.40 Table 3 provides the
annual average simulation estimates for MSF containership voyages that
carried more than 2,500 tons of food aid.41

40See appendix II for a more detailed discussion of the model's
probability distributions, assumptions that are used to assign the
probabilities, and the limitations in financial data that heighten the
model's uncertainties.

41Simulation results for a tonnage limit at 5,000 tons are included in
appendix II. We did not perform the simulation on a tonnage limit at 7,500
tons, given that so few containership voyages carried cargo at that level
and that only one LASH vessel remains in the MSF.

Table 3: GAO Simulation of MSF Voyages with Food Aid above a 2,500 Tonnage
Limit, 3-Year Annual Average for Fiscal Years 2001-2003

Dollars in millions

                      Estimated range of valuesa Low value

Average value High value

  Total food aid tonnage on voyages affected by the limit b 160,000 Simulation
                      estimates with factors 1-3 combined

Factor 1: MSF vessels may carry some food aid up to                
            the limit, depending on number and                        
               terms of food aid shipments:                           
      Food aid tonnage effectively above the limitc     61,000 92,000 138,000 
    Food aid net revenues associated with this tonnage   $6.7  $10.4    $15.6 
     Factor 2: MSF vessels may replace some food aid                  
            above the limit with other cargo:                         
       Estimated tonnage of additional other cargo      3,000  24,000  49,000 
     Net revenues earned from additional other cargo     $0.2   $1.8     $4.0 
     Factor 3: MSF vessels may continue to carry some                 
         food aid above the limit and forfeit the                     
           subsidy if sufficiently profitable:                        
     Food aid tonnage above the limit that is carried   24,000 53,000 102,000 
    Net Revenues earned from food aid above the limit    $2.6   $5.1     $9.4 
             minus forfeited subsidy payments                         

             Decline in MSF food aid tonnaged 17,000 39,000 63,000

                  Decline in MSF net revenuese $2.2 $3.5 $4.8

Source: GAO analysis using USDA data and MSF vessel data.

aRanges provided are based on probability distributions and the range of
probable results given the assumptions at a 90 percent confidence
interval. See appendix II for a more detailed discussion.

bThis tonnage does not include LASH vessels.

cWe assume carriers are not able to carry food aid tonnage exactly up to
the limit in every case, so that the food aid tonnage effectively above
the limit is greater than the difference between the tonnage and the
limit.

dThis tonnage is calculated by subtracting the food aid tonnage above the
limit that is carried from the total food aid tonnage effectively above
the limit.

eThis decline includes the revenue loss from carrying less food aid, the
revenue gain from carrying additional other cargo, the forfeited subsidy
payments, and cost savings from altering the cargo tonnages carried of
food aid and other goods.

The estimates illustrate that, under the assumption that carriers could
respond to a tonnage limit in the ways we have discussed, impacts on MSF
vessels could be reduced. While the total food aid tonnage on voyages
affected by the limit is around 160,000 tons, to the degree that carriers
can keep some food aid on voyages where the total food aid tonnage has
been above the limit, the amount of food aid that they could lose due to
the limit would, under certain assumptions, range from around 61,000 to
138,000

tons. This food aid tonnage that is effectively above the limit would
correspond to estimated net revenues of around $7 million to $16 million.
Based on our assumptions about how much other cargo MSF carriers are able
to secure to replace the food aid, MSF net revenues from additional cargo
might range from an estimated $200 thousand to an estimated $4 million.
Based on our assumptions about net revenues for food aid and other cargo,
the food aid tonnage above the limit that MSF vessels continue to carry
might range from an estimated 24,000 to an estimated 102,000 tons. The net
revenues from this food aid tonnage minus the forfeited subsidy payments
would then range from around $3 million to $9 million. Taking all three
factors into account, the total decline in MSF net revenues under a
limitation of 2,500 tons of food aid might range from around $2 million to
$5 million a year. On a per vessel basis, this amounts to roughly $120,000
to $270,000. By estimating the food aid tonnage effectively above the
limit and subtracting the tonnage that MSF vessels might continue to carry
while forfeiting their subsidy payments, the annual food aid tonnage
available to non-MSF carriers might range from around 17,000 to 63,000
tons.

Impacts on carriers could fall toward the ends of the simulation ranges
reported in table 3 or, in some cases, outside those ranges if carrier
options and responses differ from those simulated. An important
consideration is that certain key assumptions in the simulation are based
on information from fiscal years 2001 to 2003. To the extent that future
market conditions differ from those reflected in recent years, or carriers
respond in different ways than we have considered, the impacts of a
tonnage limitation could be affected. For example, if future food aid
program levels decline, then the overall tonnage and revenue changes from
a shift in MSF's food aid market share would also likely decline. If,
however, future nonfood preference cargo levels decline, then MSF may be
able to replace a smaller share of the food aid tonnage above a limit with
other cargo, and the revenue impacts from a tonnage limit would be
greater. If MSF carriers decide never to carry food aid above a limit-even
when it is profitable to do so, net of a forfeited subsidy payment-then
the total decline in food aid tonnage they carry and the revenue loss to
MSF vessels would increase.

Carriers Reported that MSF carriers told us they face certain logistical
constraints that challenge Structural Challenges Would them in being able
to effectively respond to a tonnage limit at any level.

Impede Their Ability to  o  One challenge is the difficulty in planning
vessel tonnages around aAdapt to Tonnage Limits limit. The MSF carriers
cited their lack of control over when they receive food aid cargoes from
suppliers, which makes it difficult to

distribute the food aid tonnage onto vessel sailings to stay under the
limit and meet delivery deadlines. They stated they could face additional
expenses for cargo storage at ports as well as loading penalties and
charges for delayed delivery. MSF carriers also cited the fact that they
may bid on multiple food aid shipments concurrently as a complication in
planning vessel tonnages around a limit.

o 	A second challenge cited by two MSF carriers was that they have
agreements with other carriers to share space on their vessels that could
be at risk if a carrier is concerned with shipping food aid above a
certain tonnage. USDA and MARAD corroborated this view and expressed
concern that eliminating vessel-sharing agreements would increase
inefficiency in the market.

o 	A third challenge noted by some MSF carriers was that the costs they
incur to maintain an infrastructure to support food aid cargo might become
too high if their food aid tonnage should be reduced. Such infrastructure
might include a U.S.-flag vessel stationed abroad to transfer food aid
from major ports to more remote destinations or the container loading
operations some MSF carriers have set up in the Great Lakes region.

The ability of the non-MSF carriers to benefit from a tonnage limitation
would depend on their market niche, according to our interviews with 10
non-MSF carriers. The simulation we discussed above suggests that, under
certain assumptions, the additional bagged food aid available to all
non-MSF carriers might range from less than 1 percent to 8 percent of this
segment's current bagged tonnage for tonnage limits at 5,000 and 2,500
tons, respectively.42 However, each non-MSF carrier's ability to bid for
and win that cargo would be differentially affected by (1) whether it
carries bagged food aid; (2) whether it services food aid destinations
where cargo has become available; and (3) the tonnage of cargo available,
compared with its vessel capacity. For example, while seven of the non-MSF
carriers we interviewed said they would benefit from a tonnage limitation,
two non-MSF carriers said they would be unaffected because they do not
carry bagged food aid. Three non-MSF carriers supported a lower tonnage
limitation limit, but two of them mentioned being constrained by the

42This percentage only reflects tonnage impacts forecasted for MSF
containerships. If the one remaining LASH vessel in the MSF were to also
carry less food aid, then these percentages would rise.

geographic routes they service in their ability to pick up new business.
Moreover, two other non-MSF carriers responded that a lower tonnage
limitation would actually hurt them because it would encourage MSF
carriers to more intensely compete in their market niche that services
smaller shipments. Three other non-MSF carriers with larger vessels were
satisfied with a higher tonnage limitation because it would reduce MSF
competition in the market niche for large shipments.

Effect on Program Agencies Will Depend on MSF and Non-MSF Carrier
Responses

The impact of a bagged tonnage limitation on program agencies is hard to
predict and will ultimately depend on the degree to which both MSF and
non-MSF carriers alter the terms in which they participate in the Maritime
Security Program and cargo preference. According to DOD, a tonnage limit
could cause some MSF carriers to withdraw from the Maritime Security
Program, though DOD officials indicated that they expect to receive more
applications for the next program than available slots. USDA, USAID, and
MARAD also reported several concerns about a tonnage limit at any level.
These concerns include:

o 	Decreased food aid timeliness: USDA and USAID noted concern that food
aid shipments could be delayed if the non-MSF vessels do not have
sufficient capacity to quickly carry the additional food aid shipments
above a limit or if MSF carriers responded to the limitation by spreading
food aid tonnage over several sailings to stay under the limit.43

43When we asked the non-MSF carriers about excess capacity, only 3 of 10
reported they could do more sailings with their existing vessels while the
others would either have to charter or buy new vessels. Additionally, most
MSF and non-MSF carriers reported that food aid agencies tend to bunch up
contracts in the last 3 months of the fiscal year, such that vessel
availability could become a problem under a tonnage limitation.

o 	Increased administrative burdens: USDA and USAID noted concern about
additional administrative burdens if MSF carriers responded to the limit
by submitting partial bids or dividing up shipments and if non-MSF
carriers increasingly submitted bids for bagged cargo that were contingent
on getting numerous contracts in order to fill larger vessels. MARAD noted
concern that a tonnage limit would negatively affect their initiative to
implement service contracts.44 Additionally, both food aid agencies and
MARAD will face the administrative burden of having to track volumes on a
voyage basis-something they do not currently do.

o 	Increased shipping costs: USDA, USAID, and MARAD noted concern over the
possibility of increased freight rates if (1) non-MSF carriers raised
prices in response to decreased competition, (2) freight rates bid by
non-MSF vessels for contracts that would have otherwise been carried by
MSF vessels are higher because of charter service rather than liner
service or because the non-MSF carrier may not regularly sail to that
location, (3) MSF carriers raised prices in response to losses associated
with carrying less food aid, receiving fewer subsidy payments, or
incurring costs for delayed delivery charges, and/or (4) freight rates bid
by MSF carriers for food aid shipments below a tonnage limitation are
higher than their bid otherwise would have been for a larger shipment due
to the loss of economies of scale.45

Tonnage Limitation Generally Would Have Limited Overall Impact on Ports,
with Greater Potential Effects on Certain Terminals

A bagged food aid tonnage limitation on MSF carriers generally would have
limited impact on the overall activity of the major food aid ports, based
on our analysis of the food aid shipment data and the interviews we
conducted with port representatives. Some ports may experience some shift
in the type of food aid handled, which could affect participating
terminals within these ports. For example, if bagged cargo shipments by
MSF containerships were seriously affected, this would likely have a
greater impact on terminals that predominantly stuff bagged food aid cargo
into

44According to MARAD, a service contract is a long-term contract arranged
for a regular volume of cargo to be delivered over a period of time for a
negotiated freight rate. If a tonnage limit were imposed, MARAD is
concerned that MSF carriers would not be able to participate in service
contracts because they would fear exceeding a tonnage limitation.

45We did not analyze freight rate differences between MSF and non-MSF
carriers. Both groups of carriers emphasized that competitive pressures in
the market limit their ability to raise prices.

containers. However, any impact would depend on the extent of the
limitation imposed and the MSF carriers' responses to it.

The major Gulf ports, from which most food aid cargo is shipped, would
likely experience little impact if a bagged food aid limitation were
imposed on MSF carriers because they service both MSF and non-MSF vessels
that carry bagged cargo. The result would be a shift among their
customers, according to port officials. The ports of Lake Charles and
Jacintoport are specialized agricultural commodity ports that handle only
bagged food aid cargo, and they anticipate that any loss of bagged cargo
by MSF carriers would likely be picked up by the non-MSF carriers who are
their biggest customers. The ports of New Orleans and Houston service both
MSF and non-MSF customers at different terminals. These large ports handle
all kinds of cargo in addition to food aid shipments, and port officials
anticipate that there would be no net loss in overall business for the
port. While officials said there could be some impact on individual
terminals, they estimated that the terminals could likely replace any lost
food aid container cargo with other container cargo.

The large coastal container ports like Norfolk, Charleston, and Seattle
would also likely experience little impact since the volumes of food aid
cargo involved comprise a very small percentage of their total business.
For instance, Norfolk handled 12 million tons of cargo in 2003, of which
70,000 tons were food aid that they stuffed into containers at the port,
according to a port official. Norfolk did not track how much food aid it
handled that was already in containers. The port official in Charleston
estimated that food aid was about 1 percent of its cargo, while the
official in Seattle estimated about 3 to 4 percent. These ports would
experience little impact from a bagged tonnage limitation on MSF carriers,
according to these officials. However, they said that the terminals at the
ports that stuff the bagged food aid into the containers might be
affected.

A port of Chicago representative expressed the greatest concern about the
potential impact from a tonnage limitation, although food aid cargo is a
small portion of Chicago's total cargo. MSF carriers transport most of the
section 17 cargo reserved for Great Lakes ports, and the majority of it
goes through the port of Chicago. The port official said that the port
valued food aid cargo because it was relatively more labor intensive and
generated more jobs than other types of cargo. The preponderance of food
aid cargo going through the port of Chicago is handled at a single
terminal, whose business could be damaged, depending on the nature and
extent of the impact. If MSF carrier participation in food aid shipments
is severely

curtailed and other carriers do not step in to carry section 17 cargo, the
terminal could be seriously affected. In addition, the port official said
that one benefit of the section 17 provision was that it helped make
Midwestern commodity suppliers more competitive when their cargo could be
loaded in nearby Great Lakes ports instead of being transported to Gulf
ports. The official said that the regional effort to encourage more food
aid commodity sourcing from Midwestern suppliers could also be affected if
section 17 bagged cargo was curtailed.

Observations	While we make no recommendations in this report, we believe
that our analysis provides important insights into the nature of
competition between MSF and non-MSF carriers for food aid shipments. A
sharp drop in bulk food aid shipments in fiscal years 2000 and 2001
suggests that competition for bagged food aid has become more intense. In
those years, MSF carriers captured a large share of the business, but
market share of bagged shipments shifted toward the non-MSF carriers in
fiscal years 2002 and 2003. The two segments of the industry appear to be
finding ways to respond to the changes in food aid, but this time frame is
too short to determine any clear trends.

We also believe that our analysis of a potential limit on the MSF
carriers' food aid shipments provides some findings that are not obvious
without a close examination of the system. One finding is that, if MSF
carriers have certain options in responding to a tonnage limit that would
mitigate the impacts of that limit, the potential decline in food aid
shipments by this group would be less than the total volume of food aid
carried on voyages over the limit. This result would occur if MSF vessels
carry some food aid up to the limit on affected voyages, and in some cases
choose to forfeit subsidy payments in favor of carrying profitable
shipments above the limit. To the extent that MSF carriers do choose to
carry food aid over the limit and forfeit the subsidies, a tonnage limit
may not lead to a large shift in food aid shipments and financial benefits
to non-MSF vessels. Where any financial effects of food tonnage
limitations would accrue remains uncertain. For example, MARAD subsidy
payments could be lower if MSF carriers continue to carry food aid.
However, food aid agencies could face higher costs if the limits resulted
in fewer and more expensive options for some shipments, and these agencies
have emphasized their concerns that additional constraints on food aid
shipments could impede their ability to provide food aid to meet critical
humanitarian needs.

Finally, it is important to recognize the limits of any effort to predict
the future course of events in an area in which key factors are so
volatile. For example, the volume of food aid shipments has varied greatly
over recent years, and the relationship between food aid and export
subsidies is also under discussion in the WTO negotiations. The outcome of
MARAD's efforts to support the two key maritime sectors is clearly
influenced by the level and composition of food aid, so long-term trends
and even fluctuations in food aid shipments will affect the program.
Second, the importance and profitability of food aid, compared with other
commercial or preference cargo, has a large influence on the health of the
various firms and components of the industry, and the volume and prices
for these alternative cargoes can also change significantly. In these
cases, firms may decide to move vessels into or out of the program, which
will have an effect on the existing operators.

Agency Comments and Our Evaluation

USAID, USDA, and DOD provided written comments on a draft of this report,
which are reproduced in appendixes IV, V, and VI. USAID stated that we
used sound and logical methodologies to analyze the data and accurately
identified trends pertaining to MSF and non-MSF carriers that carried food
aid over a 5-year period. USAID agreed that predicting the impact of a
tonnage limitation is difficult and said it takes a cautious approach to
changes, citing concerns regarding impacts on administrative systems and
the ability to meet foreign assistance objectives. USDA said that the
report adequately summarized USDA's major concerns over the impact on food
aid programs that could result from a bagged cargo tonnage limitation
placed on MSF carriers, including decreased food aid timeliness, increased
administrative burdens, and increased shipping costs. DOD generally
concurred with our findings. It stated that it would oppose any change in
cargo preference that would adversely impact the U.S. merchant marine
because it believed there would be negative impacts on DOD mobilization
capabilities. DOT provided oral comments. DOT said that the draft report
provided a thoughtful analysis of the potential impact of tonnage limits
on food aid shipments and how they might affect the U.S.
flag shipping industry. However, DOT identified issues with some factors,
and the way they are considered, in the simulation model we used to
estimate the range of impacts from different tonnage limits. In addition,
USAID and DOT provided technical comments, which we incorporated in the
report as appropriate.

DOT officials, including the Director of MARAD's Cargo Preference Program,
said that they identified issues with some factors and the way

they are considered in the simulation model used in our analysis of
potential impacts of a tonnage limitation. In particular, these officials
suggested that the draft report and its simulation model could have more
thoroughly explored the effects of three factors: (1) MSF carriers'
ability to replace food aid cargo with commercial cargo, (2) the
industry's reluctance to carry cargo over the limit and forfeit the
subsidy, and (3) the logistical constraints on carriers' ability to
operate under a low tonnage limit. Specifically, with respect to replacing
food aid cargo, DOT officials questioned whether sufficient commercial
cargo is actually available in the marketplace to replace food aid cargo
for MSF vessels. With respect to carrying cargo above the limit and
forfeiting the subsidy, DOT emphasized that all five MSF carriers stated
they would not give up their subsidy to carry food aid. DOT officials
stated their view that logistical limitations, which would further
constrain MSF carriers' ability to carry food aid shipments under low
tonnage limits, may be underestimated in the model. While the DOT
officials recognized these factors are acknowledged in the draft report as
limitations on the model's predictive value, they emphasized their view
that the cumulative effect of more thoroughly exploring them in the model
might have led us to conclude that the imposition of tonnage limits could
be more detrimental to MSF than the results otherwise indicated. As a
result, the officials suggested the model's limitations be more
extensively and prominently recognized in the body of the report. Finally,
the DOT officials emphasized their agreement with the aspect of our
observations that the imposition of any tonnage limit on MSF vessels could
drive up costs for the food aid program and decrease efficiency by
limiting competition and increasing freight charges.

We agree that MSF carriers may face constraints in terms of their options
in responding to a tonnage limitation. Specifically, we agree that
carriers may have restricted flexibility in managing contract amounts to
keep food aid shipments below limits and still carry food aid, and in
replacing lost food aid with other cargo. Our simulation analysis
specifically incorporates uncertainty in these factors, and we have
modified our report language in several places to clarify the range of
assumptions concerning those and other variables, and the implications of
the uncertainty regarding our results. Additional detail about how these
factors are treated in our analysis is presented in the following
paragraphs. With respect to whether carriers would in some cases carry
food aid above a tonnage limit and forgo the subsidy for affected days, we
agree that including that assumption is important to our simulation model
results. Our simulation model represents the outcome when carriers choose
the most profitable option available on each voyage, and vessel data
reported by MSF carriers suggest

that there are times when carriers would have the financial incentive to
carry food aid above the limit and forfeit their subsidy payment for that
voyage. The presentation of our simulation model results makes it clear
that option is an important one in carriers being able to mitigate the
impact of a tonnage limitation. If carriers never forgo the subsidy, the
impacts of a limitation on MSF carriers would be greater. Neither carriers
nor MARAD provided us a reason why they would not ever forgo a subsidy.

The simulation model incorporates the likelihood that MSF carriers would
face logistical constraints in managing food aid contracts to continue
carrying food aid amounts near but under the limit. It reflects
possibilities ranging from carriers being able to carry food aid exactly
up to the limit amount-for example 2,500 tons-to not being able to carry
any food aid on the share of voyages above the limit. We tested the
sensitivity of our simulation results to the particular probability
distribution assumed for this variable; and we found that if carriers are
assumed to have less flexibility in managing food aid tonnage below a
limit, the average values for the impacts would differ somewhat from the
averages we reported. For example, for one alternative distribution
assuming less flexibility, the average value of food aid tonnage that
carriers would have to give up or lose the subsidy for the voyage
increased from about 92,000 tons to about 109,000 tons.46

Similarly, our simulation model allows for the possibility that MSF
carriers would be unable to replace any food aid above a limit with
commercial or nonfood preference cargo. However, most MSF carriers
reported that they are currently sailing near full capacity, with a range
of capacity utilization rates that together average 90 percent. The
simulation model relies on these reported capacity utilization rates to
determine the most likely value for the share of food aid effectively
above the limit that carriers might be able to replace with other cargo.
However, the simulation model reflects a range of probabilities with
respect to carriers being able to replace lost food aid cargo and achieve
their current (based on the fiscal year 2001 to

46As in the simulation model results in our report, this simulation
assumed that the food aid effectively above the limit (or given up to keep
the subsidy on a voyage) equaled the amount above the limit plus an
additional amount that ranged from 0 to 2,500 tons. The most likely value
in this range of incremental amounts was changed from 0 to 1,250. Other
distributions in the simulation model were unchanged. Other impacts of
this changed distribution include a similar percentage increase in net
revenue lost from carrying food aid and a greater likelihood that carriers
would forgo the subsidy to carry larger food aid shipments on some
voyages. The average overall decline in MSF food aid tonnage due to the
limit under this scenario was about 42,000 tons, compared with 39,000 tons
in the simulation we report.

2003 data we analyzed) average capacity utilization, and includes at one
extreme the possibility that no lost food aid tonnage will be replaced.

To the extent that these constraints strongly affect MSF carriers' ability
to respond to tonnage limits, then the high end of the range of possible
results suggested by the simulation model should be considered. For
example, if MSF carriers face significant logistical constraints to
carrying food aid up to the limit, then, under a tonnage limit of 2,500
tons, they are more likely to have an annual 138,000 tons of food aid
effectively above the limit, compared with the annual 61,000 tons of food
aid effectively above the limit as estimated by the simulation's low value
results.47

In addition to the potential impacts of a tonnage limit that are suggested
by the simulation model under certain assumptions, there are potential
structural constraints we were not able to reliably quantify and include
in the model. One example is the potential impact on MSF carriers' total
tonnage and revenues if a tonnage limit were to jeopardize their vessel
sharing agreements. As we stated in the report, these types of structural
constraints could challenge MSF carriers in being able to effectively
respond to a tonnage limit at any level.

We are sending copies of this report to appropriate congressional
committees, the Secretaries of USDA, DOD, and DOT, and the Administrator
of USAID. We will also make copies available to others upon request. In
addition, this report will be available at no charge on the GAO Web site
at http://www.gao.gov.

47We provided the individual simulation results for each of the three
carrier options discussed in the report in table 3, for a limit of 2,500
tons, and in table 5, for a limit of 5,000 tons. While eliminating any one
of the three carrier options would affect the results for the remainder of
the model, the results are provided individually in order to illustrate
the types of impacts associated with each option.

If you or your staff have any questions about this report, please contact
me at (202) 512-4128. Additional contacts and staff acknowledgments are
listed in appendix VII.

Loren Yager Director, International Affairs and Trade

Report to Congressional Committees

The Honorable John McCain Chairman The Honorable Ernest F. Hollings
Ranking Minority Member Committee on Commerce, Science,

and Transportation United States Senate

The Honorable John W. Warner Chairman The Honorable Carl Levin Ranking
Minority Member Committee on Armed Services United States Senate

The Honorable Duncan Hunter Chairman The Honorable Ike Skelton Ranking
Minority Member Committee on Armed Services House of Representatives

Appendix I

                       Objectives, Scope, and Methodology

In a legislative mandate in section 3535 of the National Defense
Authorization Act for Fiscal Year 2004 (P.L. 108-136), Congress directed
us to review the impact of placing a tonnage limitation on transportation
by the Maritime Security Fleet (MSF) of cargo preference food aid and to
report to the Chairman and Ranking Minority Member of the House and Senate
Committees on Armed Services and the Senate Committee on Commerce,
Science, and Transportation. As discussed with Committee representatives,
we have focused on answering the following questions: (1) how the cargo
preference and Maritime Security Programs are designed to meet their
objectives and who participates in them; (2) what the nature and extent
are of MSF and non-MSF carrier participation in the food aid program; (3)
how establishing a bagged cargo preference tonnage limitation on MSF
vessels would be expected to affect the MSF, other U.S.
flag ships, the cargo preference food aid program, and the ports servicing
these ships.

To examine how the cargo preference and Maritime Security Programs are
designed to meet their objectives and who participates in them, we
reviewed documents, relevant legislation, regulations, and data pertaining
to the cargo preference and Maritime Security Programs from the Maritime
Administration (MARAD) and Department of Defense (DOD), as well as our
prior studies and those done by the Congressional Research Service. We
also obtained and analyzed MSF and cargo preference vessel data and food
aid shipment participation data from MARAD and the Department of
Agriculture (USDA) for fiscal years 1999 to 2003. We examined the data for
their reliability and appropriateness for our purposes through electronic
testing of the data, verification of the data against other sources, and
interviews with agency officials that manage the data. We found the data
to be sufficiently reliable to represent participation by MSF and non-MSF
vessels and carriers in transporting food aid shipments. In addition, we
interviewed agency officials at MARAD, DOD, USDA, and the Agency for
International Development (USAID), as well as representatives of three
maritime trade associations. We also conducted structured interviews with
representatives of 15 carriers that transported the majority of cargo
preference food aid, including 5 MSF and 10 non-MSF carriers.

To determine the nature and extent of MSF and non-MSF carrier
participation and competition in the food aid program, we gathered and
analyzed food aid shipment data from USDA and USAID for fiscal years 1999
to 2003. We examined the data for their reliability and appropriateness
for our purposes and found them sufficiently reliable to represent MSF and
non-MSF carrier participation and competition in the food aid program. We

Appendix I
Objectives, Scope, and Methodology

also interviewed USDA, USAID, MARAD, DOD, and maritime trade association
officials, including company representatives from 5 MSF and 10 non-MSF
carriers. To determine whether bagged cargo has accounted for an
increasing share of food aid shipments, we obtained and analyzed USDA food
aid procurement data from fiscal years 1996 to 2003. We examined the data
for their reliability and appropriateness for our purposes through
electronic testing of the data, verification of the data against other
sources, and interviews with agency officials that manage the data. We
found them sufficiently reliable to confirm that an increasing share of
food aid was shipped as bagged cargo from 1999 to 2003. In addition, we
reviewed agency reports that discussed food aid program activities and
trends, and conducted interviews with USDA and USAID officials. To examine
the process by which agencies award food aid shipments to MSF and non-MSF
carriers, we obtained and reviewed USDA, USAID, and MARAD directives and
regulations governing the ocean transportation of food aid cargo and also
reviewed applicable legislation. We also conducted interviews with USDA
and USAID officials responsible for awarding food aid shipments in
accordance with cargo preference requirements. To identify the U.S. ports
that handled the largest tonnages of food aid cargo shipped by MSF and
non-MSF carriers, we analyzed USDA food aid shipment data. To gain
additional perspectives on how MSF and non-MSF carriers handled and
transported this cargo in preparation for export shipment, we interviewed
port officials from 8 major food aid ports,1 as well as 15 MSF and non-MSF
carrier representatives.

To examine how establishing a bagged cargo preference tonnage limitation
on MSF vessels would potentially affect MSF and other U.S.-flag ships, we
obtained and analyzed USDA food aid shipment data for fiscal years 1999 to
2003. We analyzed the tonnage carried and revenues earned for each MSF
vessel voyage that carried food aid above potential limits of 2,500,
5,000, and 7,500 tons. To illustrate how carriers might respond to a
tonnage limit, we obtained operating and revenue information from the five
MSF carriers on each of their vessels from fiscal years 2001 to 2003. To
account for variation in the values of our estimates, we performed a Monte
Carlo simulation2 that varied the impact model approximately 20,000 times
from

1Port selection criteria included the tonnage of bagged cargo handled by
the port as well as factors relating to geographic location and whether
the port services MSA-17 cargo.

2A Monte Carlo simulation is a widely used computational method for
generating probability distributions of variables that depend on other
variables or parameters represented as probability distributions.

Appendix I
Objectives, Scope, and Methodology

probability distributions characterizing possible values for variables,
such as the percent of food aid above the limit that carriers replace with
other cargo, the freight rate for other cargo, and the cost differential
between food aid and other cargo. This simulation resulted in a range of
estimates, under certain assumptions, for the likely total decline in MSF
food aid tonnage and net revenues on an annual basis. A technical
discussion of the simulation model and the results at a 5,000-ton
limitation is provided in appendix II.3 We examined USDA's food aid
shipment data and carrier's vessel estimates for their reliability and
appropriateness for our purposes. For USDA's data, we performed electronic
testing of the data, verification of the data against other sources, and
interviews with agency officials that manage the data. Although we were
able to do only limited verification of the self-reported data from
carriers, we found both sources to be sufficiently reliable to inform our
simulation model. In addition, we supplemented our simulation results with
information that both MSF and non-MSF carriers provided in interviews
pertaining to any structural constraints they may face in responding to a
tonnage limitation.

To examine how establishing a bagged cargo preference tonnage limitation
on MSF vessels would potentially affect the program agencies, we reviewed
the current extent of data collection and procedures for tracking food aid
shipments to see if additional administrative burdens would be entailed.
We also interviewed agency officials at USDA, USAID, MARAD, and DOD.

To examine how establishing a bagged cargo preference tonnage limitation
on MSF vessels would potentially affect the ports that service food aid
shipments by MSF and non-MSF carriers, we analyzed food aid shipment data
from USDA that identified the ports used for each shipment for fiscal
years 1999 to 2003. We also conducted telephone interviews with
representatives of eight major food aid ports (Charleston, South Carolina;
Chicago, Illinois; Houston and Jacintoport, Texas; Lake Charles and New
Orleans, Louisiana; Norfolk, Virginia; and Seattle, Washington) to obtain
additional information, including their assessment of the potential impact
of a limitation on their port.

3We did not perform the simulation on a tonnage limit at 7,500 tons, given
that so few containership voyages carried cargo at that level and that
only one LASH vessel remains in MSF.

Appendix I
Objectives, Scope, and Methodology

We performed our work from February through August 2004 in accordance with
generally accepted government auditing standards.

Appendix II

Simulating Potential Impacts of a Bagged Tonnage Limitation

This appendix describes the data and methodology that we used to analyze
the impact of a bagged tonnage limitation on MSF and presents some
additional estimates not contained in the letter portion of this report.
This simulation analysis is based on certain assumptions regarding carrier
options and responses and makes use of food aid data from agencies,
reported vessel revenue and cost estimates for recent years, and
information from interviews about the food aid industry. The three
potential carrier responses incorporated into our model include an MSF
vessel's potential ability to continue carrying some food aid on affected
voyages, replace some food aid with other cargo, and forfeit its subsidy
for food aid contracts that are sufficiently profitable. Our methodology
illustrates that, depending on the degree to which these options exist for
MSF, carriers may reduce the overall tonnage and net revenue1 impacts of a
limit. These estimates reflect some probability that carriers will face
constraints in how they respond to limits, however, there is uncertainty
associated with some of the assumptions of the model. Carriers may face
additional logistical or structural constraints relating to program
requirements or company characteristics that would limit their responses
to a greater degree than our simulation reflects. Moreover, future market
conditions may differ from those reflected in recent data, such that our
analysis could not be used as a forecast. Thus, while our simulation can
help decision-makers understand important factors that should be taken
into account when considering tonnage limits-and develops a range of
impact estimates based on recent years that reflect those factors-actual
impacts could be near the outer limits of or fall outside our estimated
ranges.

1We use the term net revenue to reflect revenue minus costs.

    Appendix II Simulating Potential Impacts of a Bagged Tonnage Limitation

Data and Methodology	To analyze the impacts of a tonnage limit on MSF
vessels, we collected data on key tonnage and revenue variables from a
variety of sources for fiscal years 2001 to 2003. To create a list of MSF
vessel voyages that carried food aid tonnage above a potential limit, we
examined USDA's food aid shipment data and identified 123 vessel voyages.
We paired this voyage list with estimates we collected from the MSF
carriers on each vessel's annual costs and annual tonnage and freight
rates for commercial cargo, food aid cargo, and nonfood aid preference
cargo.2 We also calculated the subsidy per voyage each MSF vessel earns,
based on the number of days per voyage in that vessel's regularly
scheduled outbound service.3

To estimate a range of impacts for a tonnage limitation under certain
assumptions, we explicitly consider three options that MSF carriers may
potentially have in responding to such a limit. For affected voyages, an
MSF carrier may be able to (1) continue carrying some food aid up to the
limit, (2) replace some food aid above a limit with other cargo, and (3)
continue carrying food aid above a limit if it were more profitable than
the subsidy payment for that voyage plus any net revenue from replacing
the food aid with other cargo. As discussed below, we rely on assumptions
about the degree to which carriers may be able to respond in these three
ways to assign probabilities to a probability distribution. Table 4 shows
that we use the following five probability distributions to calculate a
range of impacts for MSF carriers under a tonnage limitation. Each of
these distributions is discussed further in the text following table 4.

2Our cost data reflect total costs incurred by the MSF carrier for each
vessel's annual operations. They include vessel expenses such as wages and
insurance, as well as cargo handling expenses. They do not include
administrative and general expenses incurred by the carrier company.

3We used the days of outbound service because a vessel carrying food aid
cargo above a tonnage limit would have to forfeit its subsidy only for the
days the food aid is on the vessel, which occur on the outbound portion of
its voyage.

    Appendix II Simulating Potential Impacts of a Bagged Tonnage Limitation

 Table 4: Assumptions in the Simulation Model and the Probability Distributions
                          Reflecting Those Assumptions

                                                        Probability at 90     
                                                        percent confidence    
                        Type of                              interval         
       Variable      distributiona       Basis       Mean Minimum Likeliest   
                                                     Maximum                  
                                                     Varies by voyage Same as 
                                    USDA and USAID       Total food aid Total 
Food aid tonnage   Triangular                                     food aid 
effectively above                                    likeliest tonnage per 
          the                       report food aid               tonnage per 
        limit b                    voyage tonnage is      voyage minus voyage 
                                     comprised of               limit tonnage 
                                       multiple,               (e.g. 2,500 or 
                                       divisible     
                                      contracts.           5,000 tons)        
    Percent of food                    Capacity         Likeliest value 0     
          aid         Triangular      utilization    percent Reported Same as 
        tonnage                    rates as reported     times 67 percent     
      effectively                                 by capacity usage likeliest 
    above the limit                  each carrier             for each vessel 
         that                                        
is replaced with                                           (ranges from 75 
      other cargo                                              to 95 percent) 

 Freight               Standard Calculated                                          
 rate per  Triangular deviation mean          Calculated   Calculated   Calculated
ton                       of                                        
for other               freight freight rate                                        
  cargo                rates on times        mean freight mean freight mean freight
                          other                                        
                          cargo  1.0, with   rate times   rate times     rate times 
                       reported   certain    0.6,         1.0,                 1.4, 
                         by all                                        
                         MSF    constraintsc with certain with certain with certain 
                      carriers                                         
                                             constraintsc constraintsc constraintsc 
 Per-ton              Estimates    $21.67         $0          $30          $35      
cost    Triangular reported                                         
difference              from                                           
 for food             carriers                                         
                         and                                           
 aid over             industry                                         
  other                experts                                         
  cargo                                                                
Percent of  Uniform   Estimate                            Not                       
  total                 from     35 percent   0 percent   applicable    70 percent
  vessel              MARAD on                            due to       
costs that              fixed                             uniform      
vary with              vessel                             distribution 
the                  costs                                          
 tonnage                                                               
  level                                                                

Source: GAO analysis.

aGiven the information available pertaining to each of our impact
variables, we selected probability distributions that are widely used by
researchers.

bThis term is used to indicate the food aid tonnage that may not be
carried when an MSF vessel voyage is affected by the limit.

cWe considered the freight rates that all five MSF carriers reported for
all of their vessel voyages-both those above and below a tonnage limit-to
calculate a standard deviation ($58.18). We then used this standard
deviation to create a distribution for each voyage ranging from 1.65
standard deviations below the reported freight rate to 1.65 standard
deviations above the reported freight rate. This range was constrained to
fall between the lowest ($39.45) and highest ($300.94) freight rates
reported by the group of carriers, and was further constrained to not
exceed the food aid rate for the voyage. From that distribution (without
the food aid rate constraint), we determined the mean freight rate for
each voyage. When the constraints created by the highest and lowest
reported rates are not binding, the mean equals the reported rate. But
when the lower (higher) constraint is binding, the mean will be greater
(lesser) than the reported rate.

1.	USDA and USAID reported that the food aid tonnage on a voyage often
comprises multiple food aid contracts such that carriers may be able to
continue to bid only on those shipments providing tonnage under the limit.
However, since food aid contract terms vary, the degree to which MSF
carriers can maximize carrying food aid up to the limit will also

Appendix II Simulating Potential Impacts of a Bagged Tonnage Limitation

vary. As a result, we include in our simulation an assumption that
carriers will most likely be able to carry tonnage up to the level of the
limit (based on profit maximization principles), but we use a probability
distribution that includes a range of values for the amount of food aid
that the vessel could potentially lose-otherwise stated as the amount of
food aid effectively above the limit. For example, at a limit of 5,000
tons, for an MSF voyage with 6,000 tons of food aid, only 1,000 tons of
food aid could be effectively above the limit. However, if MSF carriers
had less flexibility in managing food aid tonnage, up to the entire 6,000
tons could be effectively above the tonnage limit.

2.	We asked carriers to provide information about their current capacity
utilization as an indication of the most likely value for the share of
food aid they may be able to replace. Reported capacity utilization rates
were high for all carriers with a range of values averaging 90 percent.
However, we note the uncertainty regarding how close to the reported
capacity utilization rates carriers would be able to come through
replacing lost food aid tonnage with other cargo. We use a probability
distribution to incorporate this uncertainty that allows for the
possibility that carriers would not be able to replace any lost food aid
with other cargo.

3.	We asked carriers to provide their average freight rates for commercial
cargo and nonfood aid preference cargo as an indication of the most likely
freight rate they may receive on replacement cargo. Using annually
weighted information from the five MSF carriers on all of their vessel
voyages, we calculated a standard deviation and used this variation to
apply a range of values for each voyage to reflect likely freight rates
for other cargo, subject to certain constraints.

4.	If MSF carriers replace food aid above a limit with other cargo, they
are also likely to experience a change in costs. We found that it is
generally more costly for the MSF to carry a ton of food aid than it is to
carry a ton of commercial cargo.4 Based on interviews with carriers and
industry experts, we incorporate, across the model, a range of values for
this additional food aid cost differential around a most likely estimate
of $30 per ton.

4The primary reasons for this cost difference include container stuffing
and stripping costs, as well as fumigation costs that are required for
food aid and not for commercial cargo. This estimate does not include cost
differences pertaining to particular voyage contract terms.

    Appendix II Simulating Potential Impacts of a Bagged Tonnage Limitation

If carriers alter the total tonnage on a vessel voyage, their costs will
also vary. We do not have data pertaining to the percentage of MSF total
vessel costs that vary with tonnage levels. Based on broad estimates from
MARAD that around 40 percent of vessel costs are for overhead or fixed
items, we consider a wide range of values around the remaining 60 percent
of total costs.

To incorporate these five assumptions into our impact estimates, we
performed a Monte Carlo simulation.5 In this simulation, values were
randomly drawn 20,000 times from probability distributions characterizing
possible values for impact variables discussed above and listed in table
4. Under assumptions described by probability distributions selected for
these impact variables, the simulation yields estimates for the total
decline in both MSF food aid tonnage and net revenues on an annual basis.

Summary of Results	Using our simulation model, we analyzed the tonnage and
net revenue impacts on MSF of a food aid limit at 5,000 and 2,500 tons.6
Results for a 2,500-ton limit are presented in the letter portion of this
report while table 5 provides the results for a 5,000-ton limit. As shown
in table 5, the estimated decline in MSF food aid tonnage under this
limitation ranges from around 3,000 to 13,000 tons, a decline
significantly less than the total tonnage on voyages affected by the
limit-46,000 tons. In this analysis, carriers are estimated to replace
food aid above the limit with 1,000 to 11,000 tons of other cargo and
continue to carry 5,000 tons to 31,000 tons of food aid above the tonnage
limit. The total decline in net revenues for this group would range from
roughly $500,000 to $1 million.

5Monte Carlo simulation is a widely used computational method for
generating probability distributions of variables that depend on other
variables or parameters represented as probability distributions. Monte
Carlo methods are to be contrasted with the deterministic methods used to
generate specific single number or point estimates.

6We did not analyze a food aid limit at 7,500 tons in our model due to the
fact that recent data show that most of these voyages occurred on LASH
vessels that are being phased out of MSF.

    Appendix II Simulating Potential Impacts of a Bagged Tonnage Limitation

Table 5: GAO Simulation of MSF Voyages with Food Aid above a 5,000 Ton
Limit, 3-Year Annual Average for Fiscal Years 20012003

Dollars in millions

                      Estimated range of valuesa Low value

Average value High value

Total food aid tonnage on voyages affected by the limitb 46,000 tons Simulation
                      estimates with factors 1-3 combined

    Factor 1: MSF vessels may carry some food aid up to                
              the limit, depending on number                           
             and terms of food aid shipments:                          
       Food aid tonnage effectively above the limitc     11,000 22,000 38,000 
    Food aid net revenues associated with this tonnage    $1.3   $2.6    $4.5 
Factor 2: MSF vessels may replace some food aid above               
                the limit with other cargo:                            
        Estimated tonnage of additional other cargo      1,000  5,000  11,000 
      Net revenues earned from additional other cargo     $0.0   $0.4    $0.9 
Factor 3: MSF vessels may continue to carry some food               
            aid above the limit and forfeit the                        
            subsidy if sufficiently profitable:                        
     Food aid tonnage above the limit that is carried    5,000  15,000 31,000 
     Net revenues earned from food aid above the limit    $0.5   $1.5    $3.3 
             minus forfeited subsidy payments                          

              Decline in MSF food aid tonnaged 3,000 7,000 13,000

                  Decline in MSF net revenuese $0.5 $0.7 $1.0

Source: GAO analysis using USDA data and MSF vessel data.

aRanges provided are at a 90 percent confidence interval.

bThis tonnage does not include LASH vessels.

cWe assume carriers are not able to carry food aid tonnage exactly up to
the limit in every case, so that the food aid tonnage effectively above
the limit is greater than the difference between the tonnage and the
limit.

dThis tonnage is calculated by subtracting the food aid tonnage above the
limit that is carried from the total food aid tonnage effectively above
the limit.

eThis decline includes the revenue loss from carrying less food aid, the
revenue gain from carrying additional other cargo, the forfeited subsidy
payments, and cost savings from altering the cargo tonnages carried in
food aid and other goods.

According to this analysis, the impact estimates for a limit at both 2,500
and 5,000 tons are influenced most by variations in assumptions pertaining
to the amount of food aid effectively above the limit for each voyage, and
the share of food aid above the limit that carriers may be able to replace
with other cargo. A higher value for the amount of food aid effectively
above the limit tends to increase the estimate for the total decline in
MSF net revenues because MSF carriers are less able to maximize carrying
food aid up to the limit. A higher value for the share of food aid above
the limit that

Appendix II Simulating Potential Impacts of a Bagged Tonnage Limitation

carriers might replace with other cargo tends to lower the estimate for
the total decline in MSF net revenues because carriers are earning more
money from replacement cargo. However, this assumption tends to raise the
estimate for the total decline in MSF food aid tonnage carried because it
makes the option of forfeiting the subsidy payment to carry food aid above
the limit less profitable.

This simulation model has certain limitations resulting from two broad
areas of uncertainty not incorporated into the estimates. First, MSF
carriers may face logistical or structural constraints as imposed by
program requirements or company characteristics that would alter their
response to a tonnage limit in ways our simulation does not reflect. For
example, if an MSF carrier decides never to carry food aid above a limit-
even if it is profitable to do so, net of a forfeited subsidy payment-then
the total food aid tonnage available to the non-MSF carriers would also
increase. In addition, vessel financial data are based on estimates of
annual averages and may not incorporate the entire range of variation for
every variable.7 One example might include a higher food aid cost
differential associated with an emergency food aid shipment to a remote
area with particularly expensive contract terms.

Second, our model relies on data from fiscal years 2001 to 2003, which may
not be an accurate indicator of future food aid program levels, future
food aid program requirements, or the future number of U.S.-flag vessels
participating in cargo preference. For example, if future food aid program
levels decline, then the overall tonnage and revenue changes from a shift
in the MSF's food aid market share would also likely decline. Therefore,
to the extent that our model's assumptions do not adequately reflect these
two broad areas of uncertainty, the impacts of a tonnage limit could lie
outside our estimated ranges.

7Vessel financial data are also self-reported, and we were only able to do
limited verification of these data.

Appendix III

                       Profile of Maritime Security Fleet

The Maritime Security Fleet currently comprises 47 vessels operated by 12
companies. Table 6 provides a profile of the vessels participating in the
Maritime Security Fleet, as of December 2, 2003.

            Table 6: Maritime Security Fleet, as of December 2, 2003

                  Page 56 GAO-04-1065 Maritime Security Fleet
                (TEUs,   MSP                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          Central                                          Central                                          Central                                                                                       Car                                                                   Car                                                                   Car                            First                                                           First                                                                                                                                        
     Volume     square contract Vessel Company Vessel metric           APL                                 Management,                                                       Management,                                                     Management,              APL                                   Management,            President                        3,600   Management,            President                        3,600   Management,             APL                         3,900   Management,            President                        3,600   Management,            President                        3,600   Management,   MA/MSP-10 Green Gulf    Roll-on/roll-off 131,998 MA/MSP-11 Green Gulf    Roll-on/roll-off 128,328 MA/MSP-12 Green Gulf    Roll-on/roll-off 150,828                   American                                 Carrier,             Patriot American                                 Carrier,             Freedom American                                 Carrier,                        American                         Carrier                        American               2,409     Carrier     MA/MSP-18 Endeavor E-SHIPS, Containership 1,834           Endurance                              
    carried       feet  number   name   name    type   tons)          Korea American               3,900       LLC                    APL     American               3,900       LLC                   APL    American               3,900       LLC                Thailand American               3,900       LLC                  Adams   American                TEUs       LLC                 Jackson  American                TEUs       LLC                China American                TEUs       LLC                  Polk    American                TEUs       LLC                 Truman   American                TEUs       LLC                 Cove  Lines,                   sq. ft.           Point Lines,                   sq. ft.           Lake  Lines,                   sq. ft.                   International                  135,324     Inc.                       International                  155,947     Inc.                       International                  215,709     Inc.               Chesapeake   Bulk                 2,409   Corporation             Delaware   Bulk                  TEUs   Corporation                        Inc.                  TEUs                     E-SHIPS,               1,834
                    or                                       MA/MSP-1         Ship   Containership  TEUs                 MA/MSP-2 Philippines   Ship   Containership  TEUs                 MA/MSP-3 Singapore   Ship   Containership  TEUs                 MA/MSP-4            Ship   Containership  TEUs                 MA/MSP-5             Ship   Containership                       MA/MSP-6             Ship   Containership                       MA/MSP-7         Ship   Containership                       MA/MSP-8             Ship   Containership                       MA/MSP-9             Ship   Containership                                       Inc.                                             Inc.                                             Inc.                             MA/MSP-13 Liberty               Roll-on/roll-off sq. ft.              MA/MSP-14                       Roll-on/roll-off sq. ft.              MA/MSP-15                       Roll-on/roll-off sq. ft.              MA/MSP-16    Bay              Containership  TEUs                 MA/MSP-17   Bay             Containership                                                                       MA/MSP-19             Inc.   Containership  TEUs

                Appendix III Profile of Maritime Security Fleet

                         (Continued From Previous Page)

                  Page 57 GAO-04-1065 Maritime Security Fleet
                (TEUs,   MSP                                                                                                          First                                                     First                                                    First                                                 Maersk                                        Maersk                                        Maersk                                         Maersk                                         OSG Car                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       
     Volume     square contract Vessel Company Vessel metric MA/MSP-20 Enterprise E-SHIPS, Containership 1,834                        Ocean                       I,                            Ocean                       II,                          Ocean                       III,   MA/MSP-24  Maersk   Line,  Containership 3,084 MA/MSP-25 Maersk   Line,  Containership 3,084 MA/MSP-26 Maersk   Line,  Containership 3,084 MA/MSP-27  Maersk   Line,  Containership 3,084 MA/MSP-28 Overseas Carriers,    Roll     100,965           Sealand                            3,606   Inc.             Sealand                                   Inc.             Sealand                                   Inc.              Sealand                                    Inc.             Sealand                                      Inc.             Sealand                                    Inc.             Sealand                                    Inc.              Sealand                  Inc.                                      Inc.              Sealand                  Inc.                                       Inc.              Sealand                  Inc.             Sealand                 Inc.             Sealand                           
    carried       feet  number   name   name    type   tons)                        Inc.                  TEUs             Lykes       Bulk               2,698   LLC               Lykes        Bulk               2,698   LLC               Lykes       Bulk               2,698   LLC              Missouri Limited                TEUs           Virgina Limited                TEUs           Georgia Limited                TEUs           Carolina Limited                TEUs            Joyce   Inc.      on/roll-off sq. ft.           Achiever U.S. Ship                  TEUs                    Florida U.S. Ship                 3,606                     Pride  U.S. Ship                 2,890                    Motivator U.S. Ship                 2,890                    Commitment U.S. Ship                 3,606                    Atlantic U.S. Ship                 3,606                    Defender U.S. Ship                 2,306                    Endurance U.S. Ship                        Sealand  U.S. Ship                        Innovator U.S. Ship                         Sealand  U.S. Ship                        Liberator U.S. Ship                        Patriot U.S. Ship                        Performance U.S. Ship             
                    or                                                                                         MA/MSP-21 Navigator Carrier- Containership  TEUs         MA/MSP-22 Discoverer Carrier- Containership  TEUs         MA/MSP-23 Liberator Carrier- Containership  TEUs                                                                                                                                                                                                                                                     MA/MSP-29          Management, Containership                MA/MSP-30         Management, Containership  TEUs          MA/MSP-31         Management, Containership  TEUs          MA/MSP-32           Management, Containership  TEUs          MA/MSP-33            Management, Containership  TEUs          MA/MSP-34          Management, Containership  TEUs          MA/MSP-35          Management, Containership  TEUs          MA/MSP-36           Management,            MA/MSP-37 Explorer Management,            MA/MSP-38           Management,            MA/MSP-39 Integrity Management,            MA/MSP-40           Management,            MA/MSP-41         Management,            MA/MSP-42             Management,     Inc.  

Containership 2,306 TEUs Containership 2,306 TEUs Containership 2,306 TEUs
Containership 3,606 TEUs Containership 2,306 TEUs Containership 2,306 TEUs
Containership 3,606 TEUs

                Appendix III Profile of Maritime Security Fleet

                         (Continued From Previous Page)

                                                                       Volume 
                                                                      carried 
                                                                       (TEUs, 
                                                                  square feet 
                                                                           or 
MSP contract Vessel name     Company name       Vessel type         metric 
      number                                                            tons) 
                  Sealand              U.S. Ship                              
    MA/MSP-43     Quality            Management,  Containership    3,606 TEUs
                                    Inc.                          
                   Lykes                                                      
    MA/MSP-44    Motivator   Waterman Steamship   Containership    2,500 TEUs
                                Corporation                       
                 Atlantic                             LASHa            40,795 
    MA/MSP-45     Forest     Waterman Steamship                   metric tons 
                                Corporation                       
                Green Dale                                        131,998 sq. 
    MA/MSP-46                Waterman Steamship  Roll on/roll off         ft. 
                                Corporation                       
                   Lykes                                           2,698 TEUs 
    MA/MSP-47    Explorer    Waterman Steamship   Containership   
                                Corporation                       

Source: MARAD and DOD.

aLASH or Lighter Aboard Ships are barge carrying vessels that use barges
like containers.

Appendix IV

Comments from the U.S. Agency for International Development

Appendix IV
Comments from the U.S. Agency for
International Development

Appendix IV
Comments from the U.S. Agency for
International Development

Appendix IV
Comments from the U.S. Agency for
International Development

Appendix IV
Comments from the U.S. Agency for
International Development

Appendix V

Comments from the Department of Agriculture

Appendix V
Comments from the Department of
Agriculture

                                  Appendix VI

                    Comments from the Department of Defense

Appendix VII

                     GAO Contacts and Staff Acknowledgments

GAO Contacts	Celia Thomas (202) 512-8987 Leyla Kazaz (202) 512-9638

Acknowledgments	In addition to those named above, Jay Cherlow, Martin De
Alteriis, Jamie McDonald, Eric Petersen, Kendall Schaefer, Richard Seldin,
and Daniel Williams made key contributions to this report.

GAO's Mission	The Government Accountability Office, the audit, evaluation
and investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of The fastest and easiest way to obtain copies of GAO
documents at no cost

is through GAO's Web site (www.gao.gov). Each weekday, GAO postsGAO
Reports and newly released reports, testimony, and correspondence on its
Web site. To Testimony have GAO e-mail you a list of newly posted products
every afternoon, go to

www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone	The first copy of each printed report is free.
Additional copies are $2 each. A check or money order should be made out
to the Superintendent of Documents. GAO also accepts VISA and Mastercard.
Orders for 100 or more copies mailed to a single address are discounted 25
percent. Orders should be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone:	Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

To Report Fraud, Contact:
Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470

Congressional	Gloria Jarmon, Managing Director, [email protected] (202)
512-4400 U.S. Government Accountability Office, 441 G Street NW, Room 7125

Relations Washington, D.C. 20548

Public Affairs	Jeff Nelligan, Managing Director, [email protected] (202)
512-4800 U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548

                               Presorted Standard
                              Postage & Fees Paid
                                      GAO
                                Permit No. GI00

United States
Government Accountability Office
Washington, D.C. 20548-0001

Official Business
Penalty for Private Use $300

Address Service Requested
*** End of document. ***