Dairy Industry: Estimated Economic Impacts of Dairy Compacts	 
(14-SEP-01, GAO-01-866).					 
								 
U.S. dairy farmers produced 167.7 billion pounds of unprocessed, 
raw farm milk in 2000. Federal and state dairy programs influence
the minimum prices to farmers for raw milk.  These prices are	 
based on how the raw milk is to be used.  Minimum prices set for 
raw milk to be used for making drinking milk (fluid milk) are	 
higher than those for milk used for manufacturing cheese, butter,
and other dairy products. About 70 percent of the raw milk	 
produced in the United States is regulated under the U.S.	 
Department of Agriculture (USDA) federal milk marketing order	 
program. The 1996 farm bill established another pricing program  
-- the Northeast Interstate Dairy Compact (NEDC) -- which is	 
administered by a commission that sets a minimum price for raw	 
milk sold as fluid milk in six New England states. The NEDC works
in conjunction with federal and state dairy programs to establish
an alternative minimum price for raw milk in the Compact states. 
When the monthly NEDC minimum price exceeds the federal marketing
order or state minimum price, the NEDC price becomes the minimum 
price. Congress is currently considering legislation that would  
reauthorize and expand the NEDC and also establish additional	 
interstate dairy compacts. This report reviews the potential	 
economic impacts of different compact alternatives. It is likely 
that the NEDC has affected dairy sector indicators within the six
New England states, but determining the extent to which it has	 
done so is difficult. Although there is general agreement that	 
the Compact was a factor in dairy sector indicators, there is	 
disagreement over what portions of the increases can be 	 
attributed to the Compact because other factors, such as consumer
demand and the costs of processing and marketing fluid milk, also
affect retail prices. According to USDA, while the NEDC has not  
increased the federal government's net costs for its milk price  
support program, it may have increased the government's costs for
one of its nutrition assistance programs--the Food Stamp Program.
The 1996 farm bill requires that the Compact commission 	 
compensate USDA for any resulting increase in costs to the milk  
price support program. As of the end of fiscal year 2000, the	 
Compact had compensated USDA a total of  $3.2 million. GAO	 
estimates that for 1999 the impacts of different dairy compact	 
alternatives on farm-level prices, production, and revenue in	 
other regions of the country would have increased as compacts	 
increased in size and accounted for a greater portion of the	 
nation's milk supply. The NEDC, which accounts for about three	 
percent of the nation's milk supply, had a minimal impact on	 
farm-level prices, production, and revenue in 1999. An expanded  
NEDC would have accounted for approximately 18 percent of the	 
nation's milk supply in 1999, and the interregional impacts of	 
this Compact would have been a little larger but still small in  
that year. An expanded NEDC in conjunction with a southern	 
compact would have accounted for about 27 percent of the nation's
milk supply, and its impacts would have been somewhat larger. The
estimated impacts vary from region to region, and greater impacts
are estimated to occur in those regions with higher levels of	 
milk production and a larger share of raw milk being used to	 
manufacture products such as butter and cheese. In 2000, national
average farm-level milk prices were lower than they were in 1999,
resulting in a larger difference between the NEDC's and USDA's	 
minimum prices. Consequently, the NEDC would have greater	 
economic impacts than in years when USDA's minimum prices were	 
high.								 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-01-866 					        
    ACCNO:   A01786						        
    TITLE:   Dairy Industry: Estimated Economic Impacts of Dairy      
             Compacts                                                         
     DATE:   09/14/2001 
  SUBJECT:   Agricultural programs				 
	     Dairy products					 
	     Farm income stabilization programs 		 
	     Price regulation					 
	     Price supports					 
	     Prices and pricing 				 
	     Food Stamp Program 				 
	     National School Lunch Program			 
	     USDA Northeast Interstate Dairy Compact		 
	     Special Supplemental Nutrition Program		 
	     for Women, Infants and Children			                                                                 
	     USDA New England Milk Marketing Order		 
	     National School Breakfast Program			 
	     USDA Milk Marketing Order Program			 
	     USDA Norheast Milk Marketing Order 		 
	     Program						 
								 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** 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-01-866
     
Report to the Honorable Herbert Kohl, U. S. Senate

United States General Accounting Office

GAO

September 2001 DAIRY INDUSTRY Estimated Economic Impacts of Dairy Compacts

GAO- 01- 866

Page i GAO- 01- 866 Impacts of Dairy Compacts Letter 1

Results in Brief 4 Background 6 Isolating the Intraregional Impacts of the
NEDC Is Difficult 11 The NEDC Has Not Increased Net Federal Costs for the
Milk Price

Support Program, but Its Impact on a Major Nutrition Assistance Program Is
Less Certain 16 Estimated 1999 Interregional Impacts of Various Compact

Alternatives Increased as Compacts Grew in Size 19 Concluding Observations
31 Agency Comments and Our Response 32

Appendix I Objectives, Scope, and Methodology 34 NEDC?s Intraregional
Impacts 34 NEDC?s Impacts on Federal Programs 37 Compacts? Interregional
Impacts 37

Appendix II Methodology for Estimating the Interregional Impacts of Dairy
Compacts 40

IRCM Structure 40 Data Used in the Model 48 1999 Baseline Calibration 49
Compact Scenarios 52 Parameter Values for Our Baseline and Initial Estimates
54 Sensitivity Analyses 56 IRCM Limitations 60

Appendix III NEDC?s Intraregional Impacts 61 Retail Milk Prices Increased in
the NEDC States, but It Is Difficult

to Determine the Amount Attributable to the Compact 61 Limited Data Are
Available to Estimate the Potential Impacts of the

Compact on Producer Income 69 The NEDC?s Impacts on Farm Structure, Milk
Production, and Milk

Consumption Are Difficult to Determine 76

Appendix IV NEDC?s Impacts on Federal Program Costs 85 Federal Net Costs for
the Milk Price Support Program Have Not

Increased 85 Contents

Page ii GAO- 01- 866 Impacts of Dairy Compacts

It Is Unclear Whether Federal Net Costs for Nutrition Assistance Programs
Have Increased 86 Two Studies of the NEDC?s Impact on Nutrition Assistance

Programs Provide Inconclusive Results 92

Appendix V Interregional Impacts of Three Compact Alternatives in 1999 94

Range of Estimates Obtained Across Different Compact Scenarios 95 The
Economic Impacts of the NEDC in 1999 98 The Economic Impacts of an Expanded
NEDC in 1999 105 The Economic Impacts of an Expanded NEDC in Conjunction
With

a Southern Compact in 1999 112 The Economic Impacts of an Expanded NEDC in
Conjunction With

a Southern Compact Using a More Restrictive Fluid Milk Trade Assumption 119

Appendix VI Interregional Impacts of Three Compact Alternatives in 2000 127

Appendix VII Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives 132

USDA (1999)- A Study Using an Annual, Time- Series Dairy Sector Model 132
Cox et al. (1999)- A Study Using a Spatial Market Equilibrium

Model 135 Bailey (2000)- A Study Using a Regional Economic Simulation

Model 138 Balagtas and Sumner (2001)- A Study Using a Price Discrimination

Framework 140 Rosenfeld (2001)- A Study Using an Economic Model of
Classified

Pricing 142 Comparison of Studies Reviewed 144

Appendix VIII Comments From the Executive Director, NEDC 146

Appendix IX GAO Contact and Staff Acknowledgments 155

Page iii GAO- 01- 866 Impacts of Dairy Compacts Related GAO Products 156

Tables

Table 1: Reduction in 1999 Farm- Level Revenue in the Upper Midwest Under
Three Compact Alternatives 5 Table 2: USDA?s Milk Classes Used for Setting
Milk Prices 8 Table 3: Estimated Reduction in 1999 Farm- Level Milk Revenue
in

Noncompact Regions as a Result of the NEDC 22 Table 4: Estimated Reduction
in 1999 Farm- Level Milk Revenue in

Noncompact Regions as a Result of an Expanded NEDC 23 Table 5: Estimated
Reduction in 1999 Farm- Level Milk Revenue in

Noncompact Regions as a Result of an Expanded NEDC Combined With a Southern
Compact 25 Table 6: Estimated Reduction in 1999 Farm- Level Milk Revenue in

Noncompact Regions as a Result of an Expanded NEDC Combined With a Southern
Compact Under a More Restrictive Fluid Milk Trade Assumption 27 Table 7:
Estimated Reduction in 2000 Farm- Level Milk Revenue in

the Upper Midwest and in All Noncompact Regions Under the Compact Scenarios
29 Table 8: USDA's Marketing Orders Before January 2000, States in

the IRCM99 Regions, and States That Had Authorized Entry Into a Compact by
February 2001 42 Table 9: IRCM99 Dairy Product Fixed Component Composition
45 Table 10: IRCM99 Farm- Level Price Calibration 50 Table 11: IRCM99 Farm-
Level Production Calibration 50 Table 12: IRCM99 Aggregate Wholesale
Commodity Price

Calibration 51 Table 13: IRCM99 Aggregate Wholesale Commodity Production

Calibration 52 Table 14: Medium- Term Wholesale Demand Elasticities Used in

Initial Estimates 54 Table 15: Medium- Term Regional Supply Elasticities
Used in Initial

Estimates 55 Table 16: Long- Term Wholesale Demand Elasticities Used in the

Sensitivity Analyses 56 Table 17: Long- Term Regional Supply Elasticities
Used in the

Sensitivity Analyses 56

Page iv GAO- 01- 866 Impacts of Dairy Compacts

Table 18: USDA?s Classified Prices, and the Difference Between These Prices
and the Price for Cheese (USDA's Class III or California?s Class 4b) in 2000
59 Table 19: California's Classified Prices, and the Difference Between

These Prices and the Price for Cheese (USDA's Class III or California?s
Class 4b) in 2000 59 Table 20: IRCM00 Classified Component Price Premiums in

Federal Milk Marketing Orders and California 60 Table 21: Retail Prices for
a Gallon of Milk in NEDC States Before

and Immediately Following the NEDC?s Establishment 62 Table 22: Average
Annual NEDC Over- Order Payment Received by

a Dairy Farmer Located in the Six NEDC States 71 Table 23: Estimated Impact
on Food Stamp, WIC, and School

Feeding Program Participants if Retail Milk Prices Increase 10, 15, or 20
Cents per Gallon Under an Expanded NEDC and a Southern Compact 92 Table 24:
Range of Estimated Impacts on 1999 Farm- Level Prices

Across Different Compact Scenarios 95 Table 25: Range of Estimated Impacts
on 1999 Farm- Level

Production Across Different Compact Scenarios 95 Table 26: Range of
Estimated Impacts on 1999 Farm- Level Revenue

Across Different Compact Scenarios 96 Table 27: Range of Estimated Impacts
on 1999 National Average

Wholesale- Level Prices Across Different Compact Scenarios 97 Table 28:
Range of Estimated Impacts on 1999 Wholesale- Level

Production Across Different Compact Scenarios 97 Table 29: Range of
Estimated Impacts on 1999 Wholesale- Level

Expenditures Across Different Compact Scenarios 98 Table 30: Initial
Estimates of the NEDC?s Impacts on 1999 FarmLevel Indicators 99 Table 31:
Initial Estimates of the NEDC?s Impacts on 1999

Wholesale- Level Indicators 100 Table 32: Estimated Change in 1999 Farm-
Level Prices Using

Different Assumptions Under the NEDC Scenario 101 Table 33: Estimated Change
in 1999 Farm- Level Production Using

Different Assumptions Under the NEDC Scenario 101 Table 34: Estimated Change
in 1999 Farm- Level Revenue Using

Different Assumptions Under the NEDC Scenario 102 Table 35: Estimated Change
in 1999 National Average WholesaleLevel Prices Using Different Assumptions
Under the

NEDC Scenario 102

Page v GAO- 01- 866 Impacts of Dairy Compacts

Table 36: Estimated Change in 1999 Wholesale- Level Production Using
Different Assumptions Under the NEDC Scenario 103 Table 37: Estimated Change
in 1999 Wholesale- Level Expenditures

Using Different Assumptions Under the NEDC Scenario 104 Table 38: Summary of
Estimated Changes in 1999 Farm- Level

Indicators Using Our Initial and Subsequent Sets of Assumptions Under the
NEDC Scenario 104 Table 39: Summary of Estimated Changes in 1999 Wholesale-
Level

Indicators Using Our Initial and Subsequent Sets of Assumptions Under the
NEDC Scenario 105 Table 40: Initial Estimates of an Expanded NEDC?s Impacts
on

1999 Farm- Level Indicators 106 Table 41: Initial Estimates of an Expanded
NEDC?s Impacts on

1999 Wholesale- Level Indicators 106 Table 42: Estimated Change in 1999
Farm- Level Prices Using

Different Assumptions Under an Expanded NEDC Scenario 107 Table 43:
Estimated Change in 1999 Farm- Level Production Using

Different Assumptions Under an Expanded NEDC Scenario 108 Table 44:
Estimated Change in 1999 Farm- Level Revenue Using

Different Assumptions Under an Expanded NEDC Scenario 108 Table 45:
Estimated Change in 1999 National Average WholesaleLevel Prices Using
Different Assumptions Under an

Expanded NEDC Scenario 109 Table 46: Estimated Change in 1999 Wholesale-
Level Production

Using Different Assumptions Under an Expanded NEDC Scenario 109 Table 47:
Estimated Change in 1999 Wholesale- Level Expenditures

Using Different Assumptions Under an Expanded NEDC Scenario 110 Table 48:
Summary of Estimated Changes in Farm- Level Indicators

Using Our Initial and Subsequent Sets of Assumptions Under an Expanded NEDC
Scenario 111 Table 49: Summary of Estimated Changes in 1999 Wholesale- Level

Indicators Using Our Initial and Subsequent Sets of Assumptions Under an
Expanded NEDC Scenario 111 Table 50: Initial Estimates of the Impacts of the
Expanded NEDC in

Conjunction With a Southern Compact on 1999 Farm- Level Indicators 113

Page vi GAO- 01- 866 Impacts of Dairy Compacts

Table 51: Initial Estimates of the Impacts of the Expanded NEDC in
Conjunction With a Southern Compact on 1999 WholesaleLevel Indicators 113
Table 52: Estimated Change in 1999 Farm- Level Prices Using

Different Assumptions Under an Expanded NEDC in Conjunction With a Southern
Compact Scenario 114 Table 53: Estimated Change in 1999 Farm- Level
Production Using

Different Assumptions Under an Expanded NEDC in Conjunction With a Southern
Compact Scenario 115 Table 54: Estimated Change in 1999 Farm- Level Revenue
Using

Different Assumptions Under an Expanded NEDC in Conjunction With a Southern
Compact Scenario 115 Table 55: Estimated Change in 1999 National Average
WholesaleLevel Prices Using Different Assumptions Under an

Expanded NEDC in Conjunction With a Southern Compact Scenario 116 Table 56:
Estimated Change in 1999 Wholesale- Level Production

Using Different Assumptions Under an Expanded NEDC in Conjunction With a
Southern Compact Scenario 117 Table 57: Estimated Change in 1999 Wholesale-
Level Expenditures

Using Different Assumptions Under an Expanded NEDC in Conjunction With a
Southern Compact Scenario 117 Table 58: Summary of Range of Estimated
Changes in Farm- Level

Indicators Using Our Initial and Subsequent Sets of Assumptions Under an
Expanded NEDC in Conjunction With a Southern Compact Scenario 118 Table 59:
Summary of Range of Estimated Changes in 1999

Wholesale- Level Indicators Using Our Initial and Subsequent Sets of
Assumptions Under an Expanded NEDC in Conjunction With a Southern Compact
Scenario 119 Table 60: Initial Estimates of 1999 Regional Farm- Level
Indicators

Using a More Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC
in Conjunction With a Southern Compact Scenario 119 Table 61: Initial
Estimates of 1999 Wholesale- Level Indicators

Using a More Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC
in Conjunction With a Southern Compact Scenario 120 Table 62: Change in 1999
Farm- Level Prices Using Different

Assumptions and a More Restrictive Fluid Milk Trade Assumption Under an
Expanded NEDC in Conjunction With a Southern Compact Scenario 121

Page vii GAO- 01- 866 Impacts of Dairy Compacts

Table 63: Change in 1999 Farm- Level Production Using Different Assumptions
and a More Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC in
Conjunction With a Southern Compact Scenario 121 Table 64: Change in 1999
Farm- Level Revenue Using Different

Assumptions and a More Restrictive Fluid Milk Trade Assumption Under an
Expanded NEDC in Conjunction With a Southern Compact Scenario 122 Table 65:
Change in 1999 National Average Wholesale- Level Prices

Using Different Assumptions and a More Restrictive Fluid Milk Trade
Assumption Under an Expanded NEDC in Conjunction With a Southern Compact
Scenario 123 Table 66: Change in 1999 Wholesale- Level Production Using

Different Assumptions and a More Restrictive Fluid Milk Trade Assumption
Under an Expanded NEDC in Conjunction With a Southern Compact Scenario 123
Table 67: Change in 1999 Wholesale- Level Expenditures Using

Different Assumptions and a More Restrictive Fluid Milk Trade Assumption
Under an Expanded NEDC in Conjunction With a Southern Compact Scenario 124
Table 68: Summary of Estimated Changes in 1999 Farm- Level

Indicators Using Our Initial and Subsequent Sets of Assumptions and a More
Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC in
Conjunction With a Southern Compact Scenario 125 Table 69: Summary of
Estimated Changes in 1999 Wholesale- Level

Indicators Using Our Initial and Subsequent Sets of Assumptions and a More
Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC in
Conjunction With a Southern Compact Scenario 125 Table 70: Range of
Estimated Impacts on 2000 Farm- Level Prices

Across Different Compact Scenarios 128 Table 71: Range of Estimated Impacts
on 2000 Farm- Level

Production Across Different Compact Scenarios 129 Table 72: Range of
Estimated Impacts on 2000 Farm- Level Revenue

Across Different Compact Scenarios 129 Table 73: Range of Estimated Impacts
on 2000 National Average

Wholesale- Level Prices Across Different Compact Scenarios 130 Table 74:
Range of Estimated Impacts on 2000 Wholesale- Level

Production Across Different Compact Scenarios 130

Page viii GAO- 01- 866 Impacts of Dairy Compacts

Table 75: Range of Estimated Impacts on 2000 Wholesale- Level Expenditures
Across Different Compact Scenarios 131 Table 76: USDA Estimates of the
NEDC?s Interregional Impacts on

Average Annual Production Levels and Farm- Level Prices and Revenue, 2000-
05 133 Table 77: Cox et al. Estimates of the Interregional Impacts of

Compacts on Farm- Level Prices and Revenue in 1997 137 Table 78: Cox et al.
Estimates of the Interregional Impacts of

Compacts on Commodity Prices and Expenditures in 1997 137 Table 79: Bailey?s
Estimates of the Interregional Impacts of a

Multiregional Compact on Farm- Level Prices, Revenue, and Production in 2000
139 Table 80: Balagtas and Sumner?s Estimates of the Interregional

Impacts of the NEDC and an Expanded NEDC on FarmLevel Prices and Production
in 1999 141 Table 81: Rosenfeld?s Estimates of the Interregional Impacts of

Compacts on Farm- Level Revenue in 2000 143 Table 82: Summary of
Characteristics of Studies That Estimate the

Interregional Economic Impacts of Dairy Compact Alternatives 144

Figures

Figure 1: USDA?s Milk Marketing Orders 44 Figure 2: Comparison of the
Average Retail Price of a Gallon of

Milk in Boston and in the United States, November 1996- September 2000 64
Figure 3: Comparison of the Average Retail Price of a Gallon of

Milk in Boston and the USDA Class I or NEDC Price, November 1996- September
2000 66 Figure 4: Net Farm Income of a Representative Dairy Farmer in the

Northeastern Region and in Other U. S. Regions, 1991- 99 73 Figure 5:
Comparison of the Percentage of Dairy Farms Having

Favorable Solvency in the Northeastern Region With the Percentage in Other
Regions, 1991- 99 74 Figure 6: Decline in the Number of Licensed Dairy Farms
in the

NEDC States and in the Rest of the United States, 1992- 2000 77 Figure 7:
Average Number of Cows per Farm in the NEDC States

and in the Rest of the United States, 1992- 2000 78 Figure 8: Pounds of Milk
Produced per Cow in the NEDC States

and in the Rest of the United States, 1993- 2000 80

Page ix GAO- 01- 866 Impacts of Dairy Compacts

Figure 9: Per Capita Fluid Milk Consumption in the New England Milk
Marketing Order and in All Other Milk Marketing Orders, 1993- 99 83

Abbreviations

IRCM97 Dairy Sector Interregional Competition Model for 1997 IRCM99 Dairy
Sector Interregional Competition Model for 1999 IRCM00 Dairy Sector
Interregional Competition Model for 2000 NEDC Northeast Interstate Dairy
Compact OMB Office of Management and Budget USDA U. S. Department of
Agriculture WIC Special Supplemental Nutrition Program for Women,

Infants, and Children

Page 1 GAO- 01- 866 Impacts of Dairy Compacts

September 14, 2001 The Honorable Herbert Kohl United States Senate

Dear Senator Kohl: U. S. dairy farmers produced about 167.7 billion pounds
of unprocessed, raw farm milk in 2000. The amount of raw milk produced is
influenced by federal and state dairy programs that affect minimum prices
processors of dairy products must pay farmers for raw milk. These programs
were established because of concern that dairy farmers lack the market power
to substantially influence the prices that they are paid for raw milk,
compared with the market power of processors and manufacturers of dairy
products and retailers that sell these products to consumers. Minimum prices
are based on, and vary according to, how the raw milk is to be used; minimum
prices set for raw milk to be used for making drinking milk (fluid milk) are
higher than those for milk used for manufacturing cheese, butter, and other
dairy products. About 70 percent of the raw milk produced in the United
States is regulated under the U. S. Department of Agriculture?s (USDA)
federal milk marketing order program. Some states, such as California and
Maine, and portions of other states are not covered by the federal program
but instead by state programs.

The 1996 farm bill established another pricing program- the Northeast
Interstate Dairy Compact (NEDC)- which is administered by a commission that
sets a minimum price for raw milk to be used for and sold as fluid milk in
the six New England states- Connecticut, Maine, Massachusetts, New
Hampshire, Rhode Island, and Vermont. 1 The NEDC works in conjunction with
federal and state dairy programs to establish an alternative minimum price
for raw milk to be used for and sold as fluid milk in the Compact states.
When the monthly NEDC minimum price exceeds the federal marketing order or
state minimum price, the NEDC price becomes the minimum price. When the
monthly federal marketing order or state minimum price exceeds the NEDC
price, the federal or state

1 The Federal Agriculture Improvement and Reform Act of 1996 (P. L. 104-
127, Apr. 4, 1996). The act conditioned the six states' implementation of
the Compact on the Secretary of USDA's finding that there was a compelling
public interest in the Compact region to establish the Compact. This finding
was made on August 8, 1996.

United States General Accounting Office Washington, DC 20548

Page 2 GAO- 01- 866 Impacts of Dairy Compacts

price becomes the relevant minimum price. The NEDC's minimum pricing
requirement became effective in July 1997, when the Compact commission set
$16.94 as the minimum price for a hundredweight of raw milk used for and
sold as fluid milk. 2 Since then, the NEDC's price has usually been higher
than the federal price. Proponents of the NEDC assert that the Compact is
necessary to ensure the continued viability of dairy farming in New England
and, thus, an adequate local supply of fluid milk. However, opponents
counter that the NEDC causes retail prices in the Compact states to
increase, harming consumers in those states. They also counter that the NEDC
leads to lower milk prices paid to dairy farmers in other regions, harming
those farmers.

Legislation authorizing the NEDC is scheduled to expire at the end of
September 2001, and the Congress is currently considering proposed
legislation that would not only reauthorize and expand the NEDC but also
establish additional interstate dairy compacts. 3 Consequently, you asked us
to provide information on the potential economic impacts of different
compact alternatives. Specifically, you asked for information on

 the intraregional impacts (that is, the impacts within the six NEDC
states) of the NEDC on dairy sector indicators such as (1) retail milk
prices, (2) milk producer income, (3) dairy farm structure, (4) milk
production, and (5) milk consumption;

 the impact of the NEDC on the costs to the federal government of its milk
price support and nutrition assistance programs; 4 and

 the interregional impacts (that is, impacts on other milk- producing
regions) of the current NEDC, an expanded NEDC, and an expanded

2 A hundredweight of milk is 100 pounds of milk, which equates to
approximately 11. 6 gallons. Thus, the Compact?s price of $16.94 per
hundredweight equates to about $1.46 per gallon of milk.

3 The Dairy Consumers and Producers Protection Act of 2001 (H. R. 1827, May
14, 2001, and S. 1157, June 29, 2001) would authorize 6 additional states to
join the NEDC; establish a southern compact consisting of 17 states; and
establish a Pacific Northwest compact and an Intermountain compact, each
consisting of 3 states. Had these compacts existed in 2000, about 61 percent
of the raw milk produced in the country would have been produced in states
belonging to one of these compacts.

4 The milk price support program, administered by the USDA?s Commodity
Credit Corporation, indirectly ensures a minimum price for milk by
purchasing cheese, nonfat dry milk, and butter from manufacturers at
specified prices. The nutrition assistance programs, administered by USDA?s
Food and Nutrition Service, include the Food Stamp Program; the Special
Supplemental Nutrition Program for Women, Infants, and Children; and the
School Breakfast and Lunch Programs.

Page 3 GAO- 01- 866 Impacts of Dairy Compacts

NEDC in conjunction with a southern compact on selected farm- level and
wholesale- level indicators, such as prices, production, and revenue. 5

To address the first objective, we sought, but did not find, a comprehensive
economic model to estimate the Compact?s intraregional impacts while holding
constant other factors that affect the indicators. Further, time and
resource constraints limited our ability to develop our own model or series
of models to address the Compact's impact on some or all of the indicators.
As a result, we analyzed federal, state, and other data on these indicators,
covering a period of time before and after the NEDC?s minimum pricing
requirement became effective, to determine any changes in historic trends.
We also reviewed available studies regarding the NEDC?s potential
intraregional impacts on these indicators. For the second objective, we
analyzed information from USDA, the NEDC commission, and the six NEDC
states. To address the third objective, we used an interregional dairy
competition model developed by economists at the University of Wisconsin-
Madison. This economic model is useful for estimating the impacts of
different dairy policy options, such as compacts, on such indicators as milk
production, farm revenue, and wholesale commodity prices in various regions
of the country. In employing this model, we used 1999 data on such
indicators as milk production, demand, and prices because they were the most
recent, final data available at the time of our review. Using preliminary
data for 2000, we also conducted some limited modeling to estimate the
impacts of compacts in that year. In addition, we consulted with leading
dairy economists from across the country and USDA officials on our use of
the model and related assumptions, such as how milk production and purchases
of different dairy products respond to price changes. Finally, we reviewed
available studies of the potential interregional impacts of dairy compacts.
Appendix I provides more detail on our scope and methodology. Appendix II
provides information on the University of Wisconsin- Madison dairy model and
the related assumptions that we used.

5 For purposes of our analysis, the expanded NEDC includes the Compact?s
current member states plus Delaware, Maryland, New Jersey, New York, and
Pennsylvania. The southern compact includes Alabama, Arkansas, Georgia,
Kansas, Kentucky, Louisiana, Missouri, Mississippi, North Carolina, South
Carolina, Tennessee, and Virginia. As of the end of February 2001, these
states had enacted laws authorizing their entry into such compacts should
the Congress establish them. While West Virginia has also enacted such a
law, we did not include that state in our analysis because of modeling
difficulties. These difficulties are discussed in app. II.

Page 4 GAO- 01- 866 Impacts of Dairy Compacts

It is likely that the Northeast Compact has affected dairy sector indicators
within the six New England states, but determining the extent to which it
has done so is difficult. Dairy sector indicators changed to varying degrees
after July 1997, when the Compact commission?s price became effective.
Retail milk prices in selected cities and towns within New England increased
by as much as 20 cents per gallon in July 1997. Although there is general
agreement that the Compact was a factor in these increases, there is
disagreement over what portions of the increases can be attributed to the
Compact because other factors, such as consumer demand and the costs of
processing and marketing fluid milk, also affect retail prices. Regarding
milk producer income, according to the Compact commission, farmers supplying
raw milk to be used for and sold as fluid milk in the NEDC states were paid
about $146 million from July 1997 through June 2001. This figure represents
the difference between the Compact price and the minimum federal price. Some
portion of this amount is likely attributable to the Compact, but it is also
likely that dairy farmers would have received some portion even without the
Compact because, depending on market conditions, wholesale milk processors
often pay more than the minimum federal or state prices to purchase raw
milk. Finally, for the remaining indicators that we examined, there was
little difference in historic trends before and after the Compact?s minimum
price was established: The number of farms continued to decline, milk
production continued to increase, and fluid milk consumption continued to
decline. The data showed similar trends for the rest of the United States.

According to USDA, while the Northeast Compact has not increased the federal
government?s net costs for its milk price support program, it may have
increased the government?s costs for one of its nutrition assistance
programs- the Food Stamp Program. The 1996 farm bill requires that the
Compact commission compensate USDA for any increased costs to the milk price
support program resulting from the Compact. As of the end of fiscal year
2000, the Compact had compensated USDA a total of about $3.2 million.
Regarding nutrition assistance programs, USDA is not certain whether the
Compact?s impact on increased retail milk prices in the six Compact states
was sufficiently large to cause national retail milk prices to increase,
which could raise Food Stamp Program benefit levels, thus increasing the
federal government's costs. Concerning USDA?s other major nutrition
assistance programs, the Compact commission directly reimburses the six NEDC
states for the estimated cost of increased milk purchased under some of
these programs. At the end of fiscal year 2000, this compensation totaled
about $4.5 million, including about $3.8 million Results in Brief

Page 5 GAO- 01- 866 Impacts of Dairy Compacts

to cover the estimated increased costs to the Special Supplemental Nutrition
Program for Women, Infants, and Children.

We estimate that for 1999 the impacts of different dairy compact
alternatives on farm- level prices, production, and revenue in other regions
of the country would have increased as compacts increased in size and
accounted for a greater portion of the nation?s milk supply. The Northeast
Compact, which accounts for about 3 percent of the nation?s milk supply, had
a minimal impact on farm- level prices, production, and revenue in 1999. An
expanded Northeast Compact would have accounted for approximately 18 percent
of the nation?s milk supply in 1999, and we estimate that the interregional
impacts of this Compact would have been a little larger but still small in
that year. An expanded Northeast Compact in conjunction with a southern
compact would have accounted for about 27 percent of the nation?s milk
supply, and its impacts would have been somewhat larger. The estimated
impacts vary from region to region, and greater impacts are estimated to
occur in those regions with higher levels of milk production and a larger
share of raw milk being used to manufacture products such as butter and
cheese. This is the case in the Upper Midwest (which includes most of
Minnesota and Wisconsin and portions of five other states), where 80 percent
of the milk produced is used for manufactured dairy products. For example,
we estimate that the three dairy compact alternatives mentioned above would
have reduced 1999 farm- level revenue in the Upper Midwest, which we
estimate would have been about $4.5 billion in the absence of any dairy
compact, by between $4 million and $133 million. Table 1 summarizes our
estimates of how this revenue would have been affected by the three compact
alternatives.

Table 1: Reduction in 1999 Farm- Level Revenue in the Upper Midwest Under
Three Compact Alternatives

Dollars in millions Reduction in 1999 farm- level revenue Compact
alternative Dollars Percentage

Northeast Compact $4 to 9 0. 09 to 0.20 Expanded Northeast Compact 13 to 24
0.29 to 0.53 Expanded Northeast Compact and a southern compact 26 to 133
0.57 to 2.93

Source: GAO's analysis using the University of Wisconsin- Madison dairy
model.

Page 6 GAO- 01- 866 Impacts of Dairy Compacts

In 2000, national average farm- level milk prices were lower than they were
in 1999, resulting in a larger difference between the NEDC's and USDA's
minimum prices. Consequently, one would expect that the NEDC would have
greater economic impacts than in years when USDA's minimum prices were high.
Nevertheless, using preliminary data available for 2000, our estimates of
the impacts of the three compact alternatives are similar to those in 1999.
In addition, our review of the available studies of the interregional impact
of dairy compacts indicated that our estimates are comparable to those
estimated by other agricultural economists.

We provided a draft of this report to USDA officials and the Executive
Director of the Northeast Compact commission to obtain their comments. In
general, these officials stated that we conducted a comprehensive assessment
of the impacts of dairy compacts. However, the Executive Director expressed
some concern about the model we used to estimate interregional impacts.

Of the approximately 167.7 billion pounds of raw milk produced in the United
States in 2000, about 55.5 billion pounds were processed into fluid milk
products- such as whole, 2- percent, 1- percent, and skim milk; flavored
milks; and buttermilk- that yielded approximately $22 billion in retail
sales. The rest of the raw milk was used to produce manufactured products,
such as butter, cheese, ice cream, powdered milk, and yogurt. In the United
States, a complex pricing system has evolved that affects prices paid for
raw milk used to produce processed milk (fluid drinking milk) and
manufactured dairy products, such as cheese and butter. Various milk
regulators- USDA, some states, and the Northeast Dairy Compact- establish
minimum prices that must be paid for raw milk to help stabilize the milk
supply. In addition to USDA, the states, and the NEDC, other entities affect
milk prices, including cooperatives, which may provide services to farmers
such as collecting farmers? milk; milk processors, which convert raw milk to
fluid milk; manufacturers of dairy products; and retailers, which stock and
sell dairy products to consumers. Each of these groups contributes to the
value of fluid milk and dairy products sold at the retail level, and each
receives a portion of the difference between the prices that farmers receive
and the retail price.

USDA's milk marketing and milk price support programs, as well as some
states' dairy programs, are intended to ensure an adequate supply of milk by
establishing milk prices and other milk marketing rules, which, in turn, are
intended to stabilize milk marketing conditions and thus assist Background

Federal and State Dairy Programs

Page 7 GAO- 01- 866 Impacts of Dairy Compacts

individual farmers as well as consumers. In effect, these programs ensure
that farm prices do not fall below a minimum level and provide a safety net
for individual farmers who lack the market power of other entities, such as
wholesale milk processors.

Currently, about 70 percent of the milk produced in the United States is
regulated under the federal milk marketing order program created in 1937 and
administered by USDA. Under this program, on the basis of national dairy
market information, USDA sets the minimum prices that must be paid by
processors for raw fluid grade milk in specified marketing areas, or orders.
6 These prices vary by the type of dairy product for which the milk is used;
the minimum price for raw milk used for fluid drinking purposes also varies
by location. Even though USDA sets minimum prices for raw milk, buyers of
milk can and sometimes do pay farmers prices in excess of the established
minimums- prices known as ?over- order premiums.? Market forces play a role
in determining any such premiums.

Under the federal milk marketing order program, USDA has a classified
pricing system for setting minimum prices, on a monthly basis, for milk that
is based upon its intended use, as shown in table 2. 7 Federal milk
marketing order class prices are determined by using product price formulas
that compute milk component values based on wholesale dairy product prices.
For example, Class III formulas use monthly average butter, cheese, and dry
whey prices to determine values for butterfat, protein, and other solids,
respectively. Class IV formulas use monthly average butter and nonfat dry
milk prices to determine values for butterfat and nonfat solids,
respectively. The Class II price is determined by adding an amount- a Class
II differential- to the Class IV price, while the Class I price is
determined by adding a Class I differential to the higher of the Class III
or IV price. Class I prices can vary from one milk marketing order to
another.

6 The federal milk marketing order program currently in place consists of 11
geographic orders, each covering multiple states or portions of states. For
example, the Northeast order includes the following states or portions
thereof: Connecticut, Delaware, Massachusetts, Maryland, New Hampshire, New
Jersey, New York, Pennsylvania, Rhode Island, Vermont, and Virginia. It also
includes the District of Columbia. Maine is not part of the order. Instead,
Maine has its own commission that regulates milk prices.

7 The 1996 farm bill required USDA to revise its milk marketing order
program. Among other things, USDA collapsed its 31 orders into 11, increased
the number of milk classes from three to four, and changed its formulas for
setting minimum prices. These changes took effect in January 2000.

Page 8 GAO- 01- 866 Impacts of Dairy Compacts

Table 2: USDA?s Milk Classes Used for Setting Milk Prices Class Usage

Class I Fluid milk for drinking purposes Class II Soft manufactured
products, such as cream products, cottage

cheese, ice cream, and yogurt Class III Cream cheese, other spreadable
cheeses, and hard cheese Class IV Butter and dried milk products such as
nonfat dry milk

Source: USDA.

Dairy farmers selling raw milk within a federal milk marketing order receive
an average, or ?blend,? price that is the weighted average of the prices of
Class I through IV milk, with the weights determined by the amount of milk
sold for each class of use in each marketing order. The average price
farmers receive, therefore, depends in part on the extent to which the total
raw milk supply in a specific order is used to make fluid milk as opposed to
the three classes of manufactured products.

Dairy farmers located in one milk marketing order sometimes ship their milk
to another order to obtain a higher price. Depending on the amount of milk
shipped, a producer may qualify for a receiving order?s blend price. If the
producer meets the receiving milk marketing order's blend price
requirements, not only can the milk shipped qualify for the blend price- all
of that producer's milk can qualify for the blend price. However, farmers
must consider whether the cost of transporting a sufficient amount of milk
to qualify for the receiving order's blend price outweighs the benefit of
receiving a higher blend price.

Some states, such as California, Maine, Nevada, New York, Pennsylvania, and
Virginia, have established their own minimum farm- level milk pricing
programs that cover all or portions of their states. These states have
established commissions or boards to perform functions similar to those of
USDA. For example, Virginia?s milk commission, created in 1934, establishes
monthly producer prices to ensure dairy farmers an adequate return on their
investment and to preserve market stability. Similarly, Nevada?s dairy
commission, established in 1955, sets minimum prices for raw milk sold to
processing facilities located within that state.

The federal milk price support program, established in 1949, also influences
farm- level prices. This program supports farm- level prices by providing a
standing offer from USDA to purchase butter, cheese, and nonfat dry milk at
specified prices. The prices offered for these dairy products are intended
to provide sufficient revenue so that dairy product manufacturers can pay
farmers, on average, a legislatively mandated

Page 9 GAO- 01- 866 Impacts of Dairy Compacts

support price. This program is intended to make the support price a floor
price for raw milk used for manufacturing purposes, and it is unlikely that
manufactured product prices will fall below the floor for very long. Because
the price for raw milk used for fluid purposes is based, in part, on the
price of raw milk used for manufacturing purposes, the price support program
influences the price that farmers receive for raw milk used for fluid
purposes as well.

In addition to the federal and state milk marketing order programs that set
minimum milk prices, in 1996, the Congress authorized the Northeast
Interstate Dairy Compact for the six New England states. The Compact
supplements the federal milk marketing order and state programs by setting
the monthly minimum price to be paid for raw milk used for fluid milk
marketed in the six- state area. 8 In July 1997, the Compact set a minimum
price of $16. 94 per hundredweight for raw milk used for Class I, or fluid
milk, and that minimum price has not changed. In months when the federally
set minimum price for Class I milk for the Northeast Milk Marketing Order
falls below the Compact price, the Compact price takes effect. 9 In other
months, when the federally set Class I price is higher than the Compact
Class I price, the federally set Class I price takes effect. Since the
Compact was established, federally set minimum prices for the area of the
Compact that is subject to federal milk marketing regulation have ranged
from $13.50 to $20. 50 per hundredweight but have usually been below the
Compact price of $16.94 per hundredweight.

In those months when the Compact Class I price is higher than the federally
set Class I price, processors having sales of fluid milk in the six NEDC
states are required to pay a monthly over- order obligation per
hundredweight equal to the difference between $16.94 and the federally set
Class I price. Processors multiply the monthly over- order obligation by the
volume of their total fluid milk sales in the six NEDC states in

8 Raw milk sold in the Compact states for manufacturing purposes is not
eligible for the minimum price set by the Compact commission. 9 Prior to
January 2000, USDA?s New England Milk Marketing Order established the
minimum federal milk prices in those New England areas that fell under
federal milk marketing regulations. In January 2000, USDA reorganized its
milk marketing orders, collapsing the then- existing 31 orders into 11. The
new Northeast Milk Marketing Order assumed responsibility for setting
minimum milk prices in those New England areas that fall under federal milk
marketing regulations. Dairy Compacts

Page 10 GAO- 01- 866 Impacts of Dairy Compacts

hundredweight by this difference and pay this amount to the commission that
administers the Compact. After deducting administrative fees and other
expenses, the commission distributes the balance of the proceeds in
accordance with the amount of milk produced that was actually used for fluid
milk, as opposed to cheese or other manufactured products. The commission
makes disbursements to farmer cooperatives and milk handlers, located both
within and outside the NEDC states, who then make individual payments to
farmers based on their production. Thus, dairy farmers from other states,
such as New York, that supply raw milk used to make fluid milk that is sold
in the Compact states also benefit from the Compact?s minimum prices.

The 1996 farm bill provided the Compact with considerable flexibility to
establish regulations to carry out its intended purpose. The legislation
authorized the establishment of a commission composed of delegates from the
six NEDC states to administer the Compact. The state delegates are appointed
by each of their respective states and include farmer, milk processor, and
consumer representatives. In addition to being empowered to establish
Compact prices, the commission may investigate costs associated with
producing and selling milk; examine the economic forces affecting producers,
including trends in production, consumption, and the financial conditions of
dairy farmers; and prepare and provide periodic reports to the states
regarding its efforts. While the commission is required to report annually
to USDA, USDA is not required to investigate or report on the commission?s
efforts.

States in other regions of the country, including some southern states, are
considering the adoption of similar compact arrangements. The proposed Dairy
Consumers and Producers Protection Act, a bill that was introduced in the
Congress in May 2001, if enacted, would reauthorize the NEDC. The bill would
also allow additional states to enter the NEDC. 10 In addition, it would
establish a southern dairy compact consisting of 17 states, as well as a
Pacific Northwest dairy compact and an Intermountain dairy compact, each
consisting of 3 states. 11 The proposed bill, like the 1996 farm

10 The proposed expanded NEDC would also include Delaware, Maryland, New
Jersey, New York, Ohio, and Pennsylvania. 11 The proposed southern compact
would include Alabama, Arkansas, Florida, Georgia, Kansas, Kentucky,
Louisiana, Mississippi, Missouri, Nebraska, North Carolina, Oklahoma, South
Carolina, Tennessee, Texas, Virginia, and West Virginia. The proposed
Pacific Northwest compact would include California, Oregon, and Washington.
The proposed Intermountain compact would include Colorado, Nevada, and Utah.

Page 11 GAO- 01- 866 Impacts of Dairy Compacts

bill, would provide the compact commissions with broad flexibility to carry
out their objective of ensuring the continued viability of the dairy
industry in their states. Therefore, it cannot be known in advance whether
commissions for these new compacts would regulate milk pricing in their
respective states in a manner similar to the way that the NEDC commission
has regulated milk pricing.

In addition to federal and state programs and the NEDC, other entities
affect prices paid for milk at the wholesale and retail levels. For example,
about 83 percent of all raw milk produced in the United States is marketed
through dairy cooperatives that are owned by farmer- members. Cooperatives
perform services for their members and buyers of milk such as (1)
transporting milk among different milk producing areas, (2) scheduling milk
deliveries, (3) testing milk, and (4) paying members for their marketings.
12 Costs for these services are paid by processors and dairy product
manufacturers that purchase milk from the cooperatives at prices above
federally specified minimum prices and then process or manufacture, package,
and distribute fluid milk and manufactured dairy products to retailers. The
costs that processors and manufacturers incur in purchasing raw milk from
farmers or cooperatives and in processing or manufacturing, packaging, and
distributing fluid milk and manufactured dairy products are included in
prices charged to retailers for these products. Finally, the prices that
retailers set for selling milk and dairy products are affected by the
retailers? operating costs, such as labor, rent, and utilities; their
strategies for pricing milk and manufactured dairy products; and the demand
for those products. 13

Although dairy sector indicators that we examined changed after the NEDC?s
milk pricing regulations took effect in July 1997, it is difficult to
determine how much of the change is attributable to the Compact. Such a
determination is difficult because the Compact?s impact on these indicators
cannot be easily isolated from the effects of other factors. For example,
while retail milk prices increased by 15 to 20 cents per gallon in July
1997, and there is general agreement that the Compact contributed to these
increases, the lack of an economic model that fully accounts for the

12 Some cooperatives also process milk and sell the products directly to
retailers or consumers. 13 A detailed discussion of factors affecting milk
prices is included in our report entitled

Dairy Industry: Information on Milk Prices and Changing Market Structure
(GAO- 01- 561, June 15, 2001). Other Factors Affecting

Milk Prices Isolating the Intraregional Impacts of the NEDC Is Difficult

Page 12 GAO- 01- 866 Impacts of Dairy Compacts

influences of other factors, such as costs for processing fluid milk, makes
it difficult to determine how much of that price increase can be attributed
to the NEDC. Similarly, while the Compact has resulted in payments being
made to dairy farmers that reflect the difference between USDA?s marketing
order minimum prices and the NEDC?s minimum price, it is difficult to
determine whether some portion or all of these payments would have been made
to dairy farmers anyway, depending on market conditions. Although economic
reasoning suggests that the Compact would be likely to cause increased milk
production and reduced fluid milk consumption in the six NEDC states,
analyses of relevant data on dairy farm structure, milk production, and milk
consumption show little change in historic trends after the Compact?s
implementation.

Retail milk prices increased by as much as 20 cents per gallon immediately
following the NEDC?s establishment- which is an amount comparable to the
immediate increase in the minimum farm- level price for raw milk to be used
for and sold as fluid milk in the six New England states when the NEDC?s
price regulations became effective. For example, the NEDC minimum price of
$1.46 per gallon was 18 cents higher than USDA?s June 1997 Class I price for
Boston of $1. 28 per gallon and 26 cents higher than USDA?s July 1997 Class
I price of $1.20 per gallon. While this might appear to be a substantial
increase, when compared with USDA?s average Class I price of $1.41 per
gallon during the prior year, the NEDC price did not actually represent such
a large increase. However, without a model of farm- to- retail price
transmission that accounts for how quickly and how fully farm- level price
changes are passed on to wholesale and retail levels, we cannot estimate how
much of the retail price change was due to the Compact.

Furthermore, while retail milk prices in Boston and other selected cities in
the NEDC states remain at the higher levels experienced since the Compact
took effect, national average retail prices have also increased, but at
rates lower than in the NEDC states. Even so, it is not certain what portion
of the retail price increase in the NEDC states is attributable to the
Compact, given that both the retail and farm- level prices for milk have
fluctuated since July 1997. Some portion of the price increase could also be
due to other factors, such as changes in the costs for processing or
retailing milk, marketing strategies, or consumer demand. In addition, it is
difficult to estimate the extent to which fluid milk processors would have
paid more than the minimum farm- level price for milk without the NEDC; that
is, we do not know the extent to which the NEDC price substituted for
market- driven over- order premiums. Several studies analyzing the NEDC?s
impact on retail milk prices concluded that the NEDC has

Page 13 GAO- 01- 866 Impacts of Dairy Compacts

increased prices. 14 However, the amount of the increase attributed to the
NEDC varies from study to study, depending on assumptions made by the
different researchers and the time periods that they examined. For example,
estimates ranged from a low of 2.7 cents to as much as 20 cents per gallon.

Data on farm income are limited, and while dairy farmers have received NEDC
payments, 15 it is unclear to what extent these payments replaced market-
driven over- order premiums that farmers might have been paid in the absence
of the Compact. We estimate that through calendar year 2000, the NEDC
payments made to dairy farmers in the six NEDC states totaled about $99
million, assuming that all dairy farmers located in these states had their
milk processed at fully regulated NEDC plants. 16 The NEDC payments that an
average dairy farmer in one of the six states would have received would have
fluctuated widely from month to month and from year to year, however,
depending on the difference between USDA?s Class I price and the Compact
price of $16.94 per hundredweight and the percentage of milk used for fluid
milk in the NEDC states. For example, in 1998 the average NEDC over- order
producer price payment was 67 cents per hundredweight. This would have
provided dairy farmers supplying raw milk used to produce fluid milk sold in
the NEDC states about 25 cents per hundredweight, based on the percentage of
raw milk used for fluid milk. We estimate that these payments provided the
average dairy farmer in the six NEDC states about $3,892 above the minimum
amount that the farmer would have received in 1998 had USDA?s Class I price
of $16.78 been in

14 Office of Management and Budget, The Economic Effects of the Northeast
Interstate Dairy Compact, Feb. 1998; Daniel A. Lass, Mawunyo Adanu, and P.
Geoffrey Allen,

?Impacts of the Compact on Fluid Milk Retail Prices," The Northeast
Interstate Dairy Compact: Milk Market Impacts, Research Report 73,
Agricultural Experiment Station, University of Vermont, Mar. 2000; Kenneth
Bailey, ?Report on the Operation and Performance of the Northeast Interstate
Dairy Compact,? Department of Agricultural Economics and Rural Sociology,
Pennsylvania State University, July 2000; testimony by Ronald W. Cotterill
on The Impact of the Northeast Dairy Compact and Market Channel Pricing
Strategies on the Performance of the New England Dairy Industry, before the
U. S. Senate Committee on the Judiciary, July 25, 2001; and Ronald W.
Cotterill and Andrew W. Franklin, The Public Interest and Private Economic
Power: A Case Study of the Northeast Dairy Compact, Food Marketing Policy
Center, Department of Agricultural and Resource Economics, University of
Connecticut, May 2, 2001.

15 An NEDC payment is the payment attributable to the difference between
USDA?s minimum Class I price and the Compact price. 16 A fully regulated
NEDC plant is one that is physically located within one of the six NEDC
states and that receives, processes, or packages milk or dairy products.

Page 14 GAO- 01- 866 Impacts of Dairy Compacts

effect. 17 In 2000, the average NEDC over- order producer price payment was
$2.14 per hundredweight. This amount would have provided a farmer supplying
raw milk used to produce fluid milk sold in the NEDC states about 91 cents
per hundredweight, based on the percentage of raw milk used for fluid milk.
These payments provided an average dairy farmer in the six NEDC states about
$15,301 above the minimum amount that the farmer would have received in 2000
had USDA?s Class I price of $14.80 been in effect. These estimates are
comparable to data developed by the Compact commission, which indicate that
dairy farmers in the six NEDC states and New York received over- order
payments totaling about $146 million from July 1997 through June 2001. In
particular, the NEDC data indicate that about 4,200 dairy farmers, including
1,300 in New York, received average annual payments of about $9,800. Whether
these payments were sufficient to alter the financial health of dairy
farmers supplying raw milk used to produce fluid milk sold in the NEDC
states is difficult to determine, however. USDA data are inconclusive as to
whether the Compact had a positive impact on NEDC dairy farmer income, while
NEDC analyses conclude that the Compact stabilized and enhanced farmer
income.

A limited number of studies have been conducted on the Compact?s impact on
farm income. In its 1998 report, the Office of Management and Budget (OMB)
estimated a 6- to 8- percent increase in farm income from July through
December 1997. In addition, an economist at the University of Vermont
modeled the effect on Vermont dairy farmers of establishing a floor for
Class I prices and concluded that stabilizing prices by having a price floor
could have a positive impact on dairy farmer income. 18

The NEDC?s impact on farm structure is unclear. The number of dairy farms
decreased, and the average size of herds increased, both prior to and
following the NEDC?s establishment in both the Compact states and the rest
of the country. For example, the number of licensed dairy farms in the six
NEDC states decreased by 32 percent between 1992 and 2000, from 4,079 to
2,772, while the number of licensed dairy farms in the rest of the

17 USDA?s Class I price was greater than the NEDC price in 4 months of 1998.
In calculating the average annul NEDC over- order payment for 1998, the NEDC
payment for those months was zero. Therefore, the difference between the
average annual NEDC and USDA price s is not equal to the average annual NEDC
over- order payment.

18 Rick W. Wackernagel, ?Potential Economic Impacts of the Northeast
Interstate Dairy Compact on Vermont Dairy Farms,? Agricultural and Resource
Economics Review, Apr. 1998.

Page 15 GAO- 01- 866 Impacts of Dairy Compacts

country decreased by 37 percent during the same period, from 127,456 to
80,253. Regarding herd size, the average herd in the six NEDC states
increased 36 percent, from 58 to 79 milk cows, between 1992 and 2000. In the
rest of the United States, the average herd increased 57 percent during the
same period, from 56 to 88 milk cows. According to USDA, this decline in the
number of farms, along with the increase in herd size, most likely reflects
fundamental changes in dairy farming caused by such factors as technological
and genetic advances.

Although economic reasoning suggests that higher farm- level milk prices
would result in increased raw milk production, we have no basis on which to
estimate the specific impact that the Compact has had on milk production in
the six NEDC states. Data on milk production show an increase in total milk
produced and milk produced per dairy cow, but these trends began prior to
the Compact?s establishment, making it difficult to estimate the specific
impact of the NEDC. Farmers in the NEDC states increased their total milk
production by 2.9 percent, from 4.5 billion pounds in 1993 to 4.7 billion
pounds in 2000, while farmers in the rest of the nation increased production
by 10.3 percent, from 148 billion pounds to 163.3 billion pounds during the
same period. The average amount of milk produced per cow in the NEDC states
increased by about 11.6 percent during the same period, from 15,633 pounds
to 17,440 pounds. Milk production per cow in the rest of the United States
increased by about 15.9 percent during this period, from 15,726 pounds to
18,226 pounds. Studies of the NEDC?s impact on milk production, including
OMB?s study and an analysis by researchers at the University of Vermont,
estimated that the Compact has resulted in a slight increase in milk
production in the NEDC states. 19

We cannot estimate the specific impact that the Compact has had on fluid
milk consumption in the six NEDC states, in part because we cannot estimate
how much of the retail price change since July 1997 has been due to the
Compact. Data on fluid milk consumption show a decrease in per capita milk
consumption, which reflects trends both within the NEDC states and in the
rest of the country that began prior to the NEDC?s establishment. Per capita
consumption of fluid milk was higher in USDA?s New England Milk Marketing
Order than in some other USDA marketing

19 Charles F. Nicholson, Budy P. Resosudarmo, and Rick W. Wackernagel,
?Impacts of the Compact on New England Milk Supply,? The Northeast
Interstate Dairy Compact: Milk Market Impacts, Research Report 73,
Agricultural Experiment Station, University of Vermont, Mar. 2000.

Page 16 GAO- 01- 866 Impacts of Dairy Compacts

orders prior to the Compact. Even so, consumption of fluid milk had been
slowly declining, both in that marketing order and in the rest of the
country, as the consumption of other fluid beverages increased and as the
population aged. For example, annual per capita milk consumption for the New
England Milk Marketing Order declined by 4 percent from 1993 to 1999, or
from about 233 to 223 pounds. Similarly, annual per capita milk consumption
for all the other USDA marketing orders declined 6 percent from 1993 to
1999, or from about 214 to 202 pounds. In its 1998 study, OMB?s analysis of
the NEDC?s impact on fluid milk consumption during the last half of calendar
year 1997 showed a 0.5- percent decline; while in a July 2000 study, an
economist at Pennsylvania State University estimated that the NEDC had no
appreciable impact from mid- 1997 through 1999. 20

Additional details about the intraregional impacts of the NEDC are included
in appendix III.

According to USDA, the NEDC has not resulted in a net increase in the
federal government?s costs for its milk price support program, while it is
not certain whether it has affected federal costs for one of its major
nutrition assistance programs. The Compact commission must, by law,
compensate USDA for any estimated increase in costs to its price support
program that are caused by the Compact, and the NEDC commission has done so.
Regarding its nutrition assistance programs, USDA estimates that federal
costs to its largest nutrition assistance program- the Food Stamp Program-
could have increased but federal costs for its other nutrition assistance
programs have likely not increased.

As required by the 1996 farm bill, when the rate of increase in milk
production in the NEDC states exceeds the rate of increase in national milk
production, the Compact commission must compensate USDA for any additional
costs to the milk price support program that result, and the commission has
done so. 21 According to USDA officials, the NEDC did not result in a rate
of increase in production greater than the national rate of

20 OMB?s Feb. 1998 study; and Kenneth Bailey, Report on the Operation and
Performance of the Northeast Interstate Dairy Compact, Pennsylvania State
University, Department of Agricultural Economics and Rural Sociology, July
2000.

21 Neither the NEDC commission nor USDA must determine if the increase in
the rate is attributable to the NEDC. The NEDC Has Not

Increased Net Federal Costs for the Milk Price Support Program, but Its
Impact on a Major Nutrition Assistance Program Is Less Certain

Page 17 GAO- 01- 866 Impacts of Dairy Compacts

increase in 1997, during the first 6 months of the Compact. USDA calculated
that in 1998, milk production in the NEDC states was 1.8 percent greater
than the average of the prior 2 years, compared with a national increase of
1.3 percent. The NEDC compensated USDA $1.8 million for this higher rate of
increase in production. USDA calculated that in 1999, milk production in the
six states was 3.6 percent greater than the average of the prior 2 years,
compared with a national increase of 3.2 percent. The NEDC compensated USDA
$1.4 million for this higher rate of increase in production. USDA calculated
that milk production in the six NEDC states increased by 0.1 percent in
2000, compared with a national increase of 5.1 percent. Thus, compensation
was not required.

USDA is not certain whether the Compact has affected federal costs for the
Food Stamp Program, which is USDA?s largest nutrition assistance program.
According to USDA, if (1) retail milk prices in the NEDC states increased
sufficiently to increase national average retail milk prices, and (2) the
Compact was the cause of the full amount of the price increases in the NEDC
states, then the Compact might have increased federal Food Stamp Program
costs because program benefits are sensitive to the national average retail
milk price. 22 Benefit levels and federal Food Stamp Program funding have
increased since July 1997, because of, among other things, increased
national average retail milk prices. However, according to USDA, it is
difficult to establish the Compact?s impact on retail milk prices in the six
NEDC states, and thus it is difficult to establish the Compact's role in
affecting national average retail milk prices. If the Compact would have
caused benefit levels to increase to the next dollar, USDA estimates that
the Compact increased annual federal program costs by about $60 million. If
the Compact did not cause benefit levels to increase to the next dollar, any
increased retail milk prices caused by the Compact would have been absorbed
by program participants in the NEDC states.

Regarding USDA?s other major nutrition assistance programs, such as the
Special Supplemental Nutrition Program for Women, Infants, and Children
(WIC), and the National School Breakfast and Lunch Programs, USDA has
concluded that federal costs have not increased as a result of the NEDC.
Federal WIC program costs have not increased because WIC is a

22 Food Stamp Program benefits are indexed to the cost of the Thrifty Food
Plan- a nutritious low- cost model diet plan. Milk represents about 10
percent of the value of the plan. Maximum program benefit levels are rounded
down to the nearest dollar of the plan?s value.

Page 18 GAO- 01- 866 Impacts of Dairy Compacts

discretionary grant program. Federal school breakfast and lunch program
costs have not increased because the level of federal reimbursements is
based on the average price of a large variety of food items, which is
relatively insensitive to changes in the retail price of milk. 23 Because
federal funding for WIC and school breakfast and lunch programs have not
increased, those state or local agencies or organizations that provide
program benefits or program participants have had to absorb any increase in
retail milk prices caused by the NEDC. Although the 1996 farm bill does not
require the Compact to do so, the commission compensates the six states for
the increased milk costs incurred by the WIC and school programs that are
estimated to be attributable to the NEDC. Through December 2000, the NEDC
provided state WIC programs a total of $3. 8 million, and the schools a
total of $662,606.

Should compacts be expanded to include additional states, smaller increases
in retail milk prices within the compact states would be necessary to
increase the national average retail milk price and, hence the level of Food
Stamp Program benefits and federal funding. For example, USDA estimates that
should retail milk prices increase by about 20 cents per gallon- an amount
similar to the immediate increase in the NEDC states when the commission
established an NEDC price- within a compact of states that represents 50
percent of the nation?s fluid milk sales, monthly food stamp allotments
would be forced up by, on average, $1 to $2. USDA estimates that these
increases in monthly food stamp allotments would increase annual Food Stamp
Program costs by $60 to $120 million. Nonfederal costs to its other major
nutrition assistance programs, including WIC and the school programs, could
also increase should compacts be expanded. While two studies analyzed the
NEDC?s impact on USDA?s nutrition assistance programs, the studies relied on
a limited amount of data on retail milk prices in the six NEDC states, and
their results are inconclusive. 24

Additional details about the NEDC?s impact on USDA?s milk price support and
nutrition assistance programs are included in appendix IV.

23 National School Lunch Program and School Breakfast Program reimbursements
are indexed to the consumer price index for food- away- from- home for urban
consumers. 24 OMB?s Feb. 1998 study; and Qingbin Wang, Zooyob Anne,
Catherine Halbrendt, Charles Nicholson, and Jaimie Sung, ?Impacts of the
Compact on the WIC Program: Evidence From Boston and Hartford,? The
Northeast Interstate Dairy Compact: Milk Market Impacts,

Research Report 73, Agricultural Experiment Station, University of Vermont,
Mar. 2000.

Page 19 GAO- 01- 866 Impacts of Dairy Compacts

Our estimates of the interregional impacts of dairy compacts in 1999 on such
measures as farm- level prices, milk production, and farm revenue range from
minimal to somewhat larger, depending on the size of the compact and the
assumptions that we used to run the economic model. The NEDC states account
for only about 3 percent of the milk produced in the nation, and we estimate
that in 1999 it had little to no impact in other regions of the country on
farm- level prices or milk production and, hence, on farm revenue. An
expanded NEDC would account for approximately 18 percent of the milk
produced in the nation, and we estimate that in 1999 it would have had a
larger but still relatively small impact on farm- level prices, milk
production, and farm revenue in other regions of the country. An expanded
NEDC, in conjunction with a southern compact, would account for
approximately 27 percent of the quantity of milk produced in the nation, and
we estimate that in 1999 it would have had a somewhat larger impact on farm-
level prices, milk production, and revenue in other regions of the country.
These estimates are comparable to other economists? estimates of the
interregional impacts of dairy compacts of different sizes.

In general, if dairy farmers located within a compact region received higher
farm- level milk prices than they would otherwise have received, they would
respond by increasing their raw milk production. Moreover, these higher
farm- level milk prices would likely lead to higher fluid milk retail prices
in the compact region- prices that would lower consumer purchases of fluid
milk in that region. These two effects in the compact region- greater raw
milk production and lower consumer fluid milk purchases- would increase the
national supply of raw milk that was available for the manufacture of other
dairy products, such as cheese, butter, and nonfat dry milk. In turn, this
increase in the national supply of milk for manufacturing purposes would
result in lower farm- level prices for raw milk to be used for manufacturing
purposes. Because minimum Class I prices are based on the prices paid for
raw milk to be used for manufacturing purposes, farmers in noncompact
regions would receive lower farm- level prices for all classes of milk and,
thus, lower blend prices. Other things being equal, dairy farmers in
noncompact regions would respond to lower farm- level prices by reducing
their milk production. 25 These two effects in noncompact regions- lower
farm- level prices and reduced production- would cause farm revenue there to
fall. This impact

25 This reduction in production by dairy farmers in noncompact regions would
keep the farm- level price for milk from falling further. Estimated 1999

Interregional Impacts of Various Compact Alternatives Increased as Compacts
Grew in Size

Page 20 GAO- 01- 866 Impacts of Dairy Compacts

would be particularly significant for dairy farmers in regions such as the
Upper Midwest, where most milk is used for manufacturing purposes. Farmers
in noncompact regions who ship their milk to compact regions may be eligible
to receive the compact region?s farm- level price for that milk, which could
offset the loss in revenue associated with lower farmlevel prices for milk.
If the compact price is sufficiently high, any increased transportation or
shipping costs could be offset.

To assess the likely interregional impacts of each compact alternative or
scenario, we derived an initial set of estimates that represents the impact
of that alternative in 1999, given an initial set of assumptions. 26 We then
changed key assumptions to analyze how sensitive our initial estimates were
to such changes. In general, these sensitivity analyses demonstrated that
our initial estimates were not very sensitive to changes in the key
assumptions. (For a more detailed description of our initial and subsequent
sets of assumptions, see app. II.) Accordingly, we present our estimates of
the impacts of the different compact scenarios as ranges that include our
initial set of estimates and the results of our sensitivity analyses. In
addition, we present these estimates as changes from our 1999 baseline
estimates, which represent the estimated values of farm- level and
wholesale- level dairy indicators in that year in the absence of any dairy
compact- our ?no- compact scenario.?

Our estimates apply only to 1999, and they may not represent the
interregional impacts of compacts in all years. In particular, these
estimates are based on data for the period prior to USDA?s milk marketing
order regulatory reforms in January 2000, which have affected some dairy
sector indicators, such as farm- level milk prices. Furthermore, farm- level
milk prices in 1999 were higher than they were in some other years, and we
anticipate that, other things being equal, compacts have less of an impact
in years when farm- level prices are relatively high. 27 In addition,

26 These assumptions pertain to the (1) responsiveness of milk production to
changes in farm- level milk prices, (2) responsiveness of demand to changes
in wholesale- level dairy commodity prices, (3) level of market- driven
over- order premiums, (4) ability of milk to be shipped between regions, (5)
amount of the compact over- order producer price payment, and (6) amount of
transportation costs.

27 We included data from 2000 as one of our sensitivity analyses; however,
we did not use 2000 as a base year for estimating the interregional impacts
because complete data for that year are not yet available. Furthermore,
modeling the impacts of compacts using 2000 data could produce estimates
that are less reliable than estimates derived from 1999 data because the
dairy industry was in the process of adjusting to USDA?s January 2000 milk
marketing order reforms.

Page 21 GAO- 01- 866 Impacts of Dairy Compacts

although our estimated impacts of compacts on noncompact regions for both
1999 and 2000 are relatively small, the impacts on some individual dairy
farmers, such as small producers with marginal profitability, in noncompact
regions could be significant.

Finally, as in any modeling effort, there is some uncertainty about a
model's structure and the data and assumptions used. In addition, the model
that we used was limited in its ability to distinguish between shipments of
bulk raw milk and packaged fluid milk into regions that import milk to meet
their demand because the model is an annual model, and such shipments are
frequently seasonal. (See app. II for a discussion of this as well as other
modeling limitations.) Despite this uncertainty and limitation, we believe
that the process for developing our estimates was rigorous and that the
model is comprehensive and sound. Given these conditions, our estimates
should be interpreted as indicative of the order of magnitude of changes in
farm and wholesale economic values, rather than as precise estimates.

Appendix V provides more detailed information about our estimates of the
impacts of the three compact alternatives on 1999 farm- level prices,
production, and revenue in noncompact regions, as well as on national
average wholesale- level prices and national wholesale- level production and
expenditures.

We estimate that the NEDC resulted in small economic impacts in noncompact
regions in 1999. Specifically, we estimate that the largest reductions in
farm- level revenue under the NEDC compared with the nocompact scenario
occurred in California and the Upper Midwest region: from $4 million to $11
million and from $4 million to $9 million, respectively. 28 Table 3 provides
our estimates of the extent to which the NEDC reduced farm- level revenue-
that is, the value of all milk sold by

28 In the University of Wisconsin- Madison dairy sector model, the Upper
Midwest region is represented as Minnesota, North Dakota, and Wisconsin.
Interregional Impacts of

the NEDC in 1999 Were Small

Page 22 GAO- 01- 866 Impacts of Dairy Compacts

dairy farmers- in these two regions and in all noncompact regions combined.
29

Table 3: Estimated Reduction in 1999 Farm- Level Milk Revenue in Noncompact
Regions as a Result of the NEDC

Dollars in millions Estimated revenue reduction with the NEDC

Region Estimated farm- level

revenue under the nocompact scenario Dollars Percentage

California $3,954 $4 to 11 0.10 to 0.28 Upper Midwest 4, 533 4 to 9 0. 09 to
0.20 All noncompact regions 18,513 11 to 29 0.06 to 0.16

Source: GAO analysis using the University of Wisconsin- Madison dairy model.

These estimated impacts on farm- level revenue were small because dairy
farmers in the NEDC states produced only about 3 percent of the nation?s
milk supply. As a result, any increased supply of milk that was available
for manufacturing purposes in 1999 from NEDC farmers was small compared with
the nation?s total milk supply for manufacturing purposes. Therefore, the
impact on farm- level prices and milk production, and hence on farm revenue,
for producers outside the compact region was also small. For example, we
estimate that, as a result of the NEDC, farm- level prices in all noncompact
regions remained unchanged or fell by no more than 2 cents per
hundredweight, or less than 0.20 percent, while milk production for all
noncompact regions combined fell by less than 0.06 percent.

We estimate that, in 1999, the NEDC?s impact on the national average
wholesale prices of manufactured dairy products was also minimal. For

29 For modeling purposes, the NEDC compact scenario is represented as being
part of the University of Wisconsin- Madison dairy sector model?s northeast
region. This region includes Connecticut, Delaware, the District of
Columbia, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New
York, Pennsylvania, Rhode Island, and Vermont. If the northeast region
states that are not part of the NEDC were modeled as being in a noncompact
region, our estimate of the NEDC?s impact on all noncompact regions combined
would be a little larger than we report, except to the extent that NEDC
payments to New York farmers who sell raw milk used for and sold as fluid
milk to processors in New England affect the lower prices that non- NEDC
producers in the northeast region receive for the remainder of their milk.

Page 23 GAO- 01- 866 Impacts of Dairy Compacts

example, we estimate that the wholesale prices per hundredweight for
American cheese were 3 to 9 cents lower and for butter about 3 to 23 cents
lower than they would have been under the no- compact scenario. 30 These
estimated differences, even at the upper ends of these ranges, represent
about 0.06 and 0.19 percent, respectively, of our estimated 1999 wholesale
American cheese and butter prices under the nocompact scenario.

We estimate that, in 1999, the interregional impacts of an expanded NEDC
that included five additional states would have been a little larger than
the impacts of the NEDC, but still small. Specifically, we estimate that,
compared with our analyses using the no- compact scenario, dairy farm
revenue in 1999 would have been reduced the most in the Upper Midwest
region- by $13 million to $24 million. 31 Table 4 provides our estimates of
the extent to which an expanded Northeast Compact would have reduced farm-
level revenue in the Upper Midwest and in all noncompact regions combined.
32

Table 4: Estimated Reduction in 1999 Farm- Level Milk Revenue in Noncompact
Regions as a Result of an Expanded NEDC

Dollars in millions Estimated revenue reduction in farmlevel revenue with an
expanded NEDC

Region Estimated farm- level

revenue under the no- compact scenario Dollars Percentage

Upper Midwest $4,533 $13 to 24 0.29 to 0.53 All noncompact regions 18,513 30
to 96 0.16 to 0.52

Source: GAO analysis using the University of Wisconsin- Madison dairy model.

30 This equates to 0.03 to 0. 09 cents per pound lower for American cheese
and 0. 03 to 0.23 cents per pound lower for butter. 31 In one of the
sensitivity analyses, farm- level revenue in California decreased by $33
million, or by 0.8 percent. In general, however, our estimated impacts were
smaller in California than in the Upper Midwest region.

32 For modeling purposes, the expanded NEDC is represented as being the
University of Wisconsin- Madison dairy sector model?s northeast region,
which is roughly comparable to USDA?s Northeast Milk Marketing Order.
Interregional Impacts of an

Expanded NEDC in 1999 Would Also Have Been Relatively Small

Page 24 GAO- 01- 866 Impacts of Dairy Compacts

As under the NEDC scenario, we estimate that the impact of an expanded NEDC
would have been relatively small because dairy farmers in the 11 states
included in the expanded Compact produced only about 18 percent of the
nation?s milk supply. As a result, any increased supply of milk that would
have been available for manufacturing in 1999 from those farmers, although a
little larger than with the NEDC, would have still been small compared with
the nation?s total milk supply for manufacturing. Therefore, the impact on
farm- level prices and milk production, and hence on farm revenues, for
producers outside the expanded Compact region would have been small. For
example, under the expanded NEDC scenario we estimate that, compared with
our no- compact scenario, farm- level prices in noncompact regions would
have fallen by no more than 6 cents per hundredweight or less than 0.5
percent, while milk production for all noncompact regions combined would
have been lower by about 0.21 percent or less.

We estimate that, in 1999, the impact of an expanded NEDC on the national
average wholesale prices of manufactured dairy products would have been a
little larger than the impact of the NEDC, but still relatively small. For
example, we estimate that the wholesale prices per hundredweight for
American cheese would have been about 18 to 41 cents lower and for butter 46
to 88 cents lower than under our no- compact scenario. 33 These differences,
even at the upper ends of these ranges, represent less than 0.3 and 0.7
percent, respectively, of our estimated 1999 wholesale American cheese and
butter prices under the no- compact scenario.

We estimate that, in 1999, the interregional impacts of an expanded NEDC in
conjunction with a southern compact- a total of 23 states- would have been
somewhat larger than the impact of our other compact scenarios.
Specifically, under this scenario and using the same assumption about fluid
milk trade between regions as used in the previous scenarios, compared with
our no- compact scenario we estimate that dairy farm revenue in 1999 would
have been reduced the most in California and in the Upper Midwest and
Mideast regions: $26 million to $118 million, $26 million to $63 million,
and $21 to $43 million, respectively. 34 , 35 Table 5

33 This equates to about 0.18 to 0.41 cents per pound lower for American
cheese and 0.46 to 0.88 cents per pound lower for butter. 34 In the
University of Wisconsin- Madison dairy sector model, the Mideast region is
represented as Indiana, Michigan, Ohio, and West Virginia. Interregional
Impacts of an

Expanded NEDC Combined With a Southern Compact in 1999 Would Have Been
Somewhat Larger

Page 25 GAO- 01- 866 Impacts of Dairy Compacts

provides our estimates of the extent to which an expanded NEDC in
conjunction with a southern compact would have reduced farm- level revenue
for milk in these regions and for all noncompact regions combined, in 1999.
36

Table 5: Estimated Reduction in 1999 Farm- Level Milk Revenue in Noncompact
Regions as a Result of an Expanded NEDC Combined With a Southern Compact

Dollars in millions Estimated revenue reduction with an expanded NEDC and

a southern compact Region

Estimated farm- level revenue under the nocompact scenario Dollars
Percentage

California $3,954 $26 to 118 0. 66 to 2.98 Upper Midwest 4, 533 26 to 63
0.57 to 1.39 Mideast 1, 792 21 to 43 1.17 to 2.40 All noncompact regions
14,805 124 to 268 0. 84 to 1.81

Source: GAO analysis using the University of Wisconsin- Madison dairy model.

The estimated impact of the expanded NEDC in conjunction with a southern
compact is relatively larger because dairy farmers in the states included in
these compacts produced about 27 percent of the nation?s milk supply. As a
result, any increased supply of milk that would have been available for
manufacturing purposes in 1999 from farmers in these states would have been
somewhat larger than under the previous scenarios. Therefore, the impact on
farm- level prices and milk production, and hence on farm revenues, for
producers outside the compact regions would have

35 For one sensitivity analysis, in which we increased transportation costs
for milk by 25 percent, our model produced an anomalous result: an increase
in dairy farm revenue in California. Because our initial estimate and the
results of all the other sensitivity analyses show farm revenue losses in
California, we are omitting this result in reporting that California was one
of the regions with the largest estimated declines in farm- level revenue.

36 For modeling purposes, the expanded NEDC in conjunction with a southern
compact is represented as being part of the University of Wisconsin- Madison
dairy sector model?s Northeast, Appalachia, Southeast and Central regions.
The Central region includes seven states, two of which are compact and five
of which are noncompact states. If the five noncompact states in the Central
region were modeled as being part of a noncompact region, our estimate of
the impact of an expanded NEDC in conjunction with a southern compact on
noncompact regions would be a little larger than we report, except to the
extent that producers in these states might receive compact payments from
selling raw milk used for and sold as fluid milk to processors in compact
states that might offset the lower prices that producers in these states
would receive for the rest of their milk. Noncompact regions include the
remaining eight IRCM regions.

Page 26 GAO- 01- 866 Impacts of Dairy Compacts

been somewhat larger. For example, we estimate that farm- level prices in
noncompact regions could have fallen by as much as 36 cents per
hundredweight or about 2.6 percent compared with our no- compact scenario,
while milk production for all noncompact regions combined could have fallen
by as much as 0.75 percent.

We estimate that, in 1999, the impacts of an expanded NEDC in conjunction
with a southern compact on the national average wholesale prices of
manufactured dairy products would have been somewhat larger than the impacts
of our other scenarios. For example, compared with the estimated wholesale
prices under our no- compact scenario, we estimate that the prices per
hundredweight would have been about 62 cents to $1.41 lower for American
cheese and between 21 cents higher and $6.53 lower for butter. 37 At the
upper end of these ranges, these differences represent about 1.0 percent and
5.5 percent, respectively, of our estimated 1999 wholesale American cheese
and butter prices under our no- compact scenario.

For the expanded NEDC plus a southern compact scenario, we found that our
estimates of interregional impacts were sensitive to our assumption about
how much milk can be shipped between noncompact and compact regions. In
particular, our estimated impacts for 1999 of an expanded NEDC in
conjunction with a southern compact on noncompact regions would have been
greater if we had used a more restrictive assumption that limits the amount
of milk that can be shipped from noncompact into compact regions. 38
Specifically, we estimate that using a more restrictive

37 This equates to about 0.62 cents to 1. 41 cents per pound lower for
American cheese and between 6.53 cents per pound lower and 0. 21 cents per
pound higher for butter. 38 In our initial estimates, we assumed that USDA
milk marketing order regulations apply to milk movements between noncompact
and compact regions. This assumption allows milk to move between compact and
noncompact regions with relatively few restrictions, subject to USDA and
compact requirements regarding when, and under what circumstances,
processors must pay exporting as opposed to importing marketing order or
compact minimum blend prices. In this sensitivity analysis, we used a more
restrictive trade assumption. Under this assumption, we restricted the
amount of milk that can move into a compact region to the amount of milk
produced within a 100- mile radius surrounding a compact region. We
performed this analysis because our data are aggregated at the regional
level as opposed to being at the milk plant or dairy farm level. Therefore,
our data on transportation costs of shipping milk are average cost data and
do not apply to individual shipments. As a result, initial results may allow
for more movement of milk between noncompact and compact regions than would
actually occur. This sensitivity analysis shows the extent to which our
estimates would be different if the movement of milk was more restricted.

Page 27 GAO- 01- 866 Impacts of Dairy Compacts

assumption increases our estimate of how much farm- level prices, milk
production, and farm revenues in noncompact regions would have fallen in
1999 under this scenario compared with under our no- compact scenario. Table
6 shows our estimated reductions in farm revenues for raw milk in California
and the Upper Midwest and Mideast regions, and all noncompact regions
combined under our restricted fluid milk trade assumption compared with
under our no- compact scenario. 39

Table 6: Estimated Reduction in 1999 Farm- Level Milk Revenue in Noncompact
Regions as a Result of an Expanded NEDC Combined With a Southern Compact
Under a More Restrictive Fluid Milk Trade Assumption

Dollars in millions Reduction in revenue as a result of an expanded NEDC and
a southern

compact using a more restrictive fluid milk trade assumption

Region Estimated farm- level

revenue under a nocompact scenario Dollars Percentage

California $3,954 $70 to 145 1. 77 to 3.67 Upper Midwest 4, 533 103 to 133
2. 27 to 2.93 Mideast 1, 792 34 to 45 1.90 to 2.51 All noncompact regions
14,805 257 to 374 1. 74 to 2.53

Source: GAO analysis using the University of Wisconsin- Madison dairy model.

We also estimate that the impact in 1999 of an expanded NEDC in conjunction
with a southern compact on the national average wholesale prices of some
manufactured dairy products would have been greater under the more
restrictive fluid milk trade assumption than without that restrictive trade
assumption. For example, under the more restrictive trade assumption, we
estimate that the price per hundredweight for American cheese would have
been about $1.27 to $1.86 lower than under our no- compact scenario. 40
However, for butter we estimate that the impact with the restrictive trade
assumption would have been smaller than the estimated impact without the
restrictive trade assumption. Under the

39 We present our estimates of the interregional impacts under a more
restrictive trade assumption only for the expanded NEDC in conjunction with
a southern compact scenario because this more restrictive trade assumption
had no effect on our estimates under our other compact scenarios.

40 This equates to about 1.27 to 1.86 cents per pound lower for American
cheese.

Page 28 GAO- 01- 866 Impacts of Dairy Compacts

restrictive trade assumption, we estimate that the price per hundredweight
for butter would have changed from 7 cents higher to $2.80 lower than under
our no- compact scenario. 41 At the upper end of these ranges, these
differences represent about 1.3 and 2.3 percent, respectively, of our
estimated 1999 wholesale prices for American cheese and butter under our no-
compact scenario.

As noted previously, the farm- level prices that we used in our model can
affect our estimates of the impacts of compacts on dairy sector indicators
such as farm- level revenue. In 1999, the national average blend price was
$14.09 per hundredweight of milk; in 2000, the national average blend price
was $12.11 per hundredweight of milk. With lower farm- level prices in 2000
than in 1999, the difference between a compact price in our model and the
federal milk marketing order Class I minimum price was larger in 2000 than
in 1999. 42 As a result, the increase in milk production and decrease in
fluid milk purchases that would have likely occurred within a compact region
in 2000 would be expected to be greater than when farmlevel prices were
higher, as they were in 1999. This situation, in turn, would imply a greater
increase in the supply of milk available for manufacturing dairy products in
2000, which, other things being equal, would lead to lower farm- level
prices, reduced milk production, and lower farm- level revenue in noncompact
regions.

However, on the basis of preliminary data for 2000, we estimate that the
impacts of our three compact scenarios, which are based on our initial set
of assumptions, are generally similar to our initial estimates for each
scenario in 1999. Even though our estimates are generally similar, when we
impose our more restrictive fluid milk trade assumption on our scenario of
an expanded NEDC in conjunction with a southern compact, our estimates of
the impact on the Upper Midwest are slightly greater for 2000 than for 1999.
The similarities between our 2000 and 1999 estimates suggest that other
factors may be affecting our estimates for 2000.

41 This equates to about 0.07 cents per pound higher to about 2. 8 cents per
pound lower for butter. 42 That is, the difference will be larger as long as
the compact price is expressed as a fixed amount per hundredweight, such as
$16. 94 per hundredweight of milk as used by the NEDC, rather than as a
fixed amount above any USDA milk marketing order Class I price. Using Farm-
Level Prices

for 2000 as Opposed to 1999 Has a Limited Influence on the Estimated Impacts
of Compacts

Page 29 GAO- 01- 866 Impacts of Dairy Compacts

In those years when noncompact farm- level prices are lower than compact
farm- level prices, a factor offsetting the potentially larger interregional
impacts of compacts is the ability to market noncompact region milk in
compact regions for use as fluid milk. Farmers in noncompact regions whose
milk is marketed in compact regions for use as fluid milk may be eligible to
receive the compact regions? farm- level blend price for that milk. When
noncompact region farm- level blend prices are low, the gain to farmers from
shipping milk to compact regions is greater than when noncompact region
prices are high. This gain can partially offset the larger negative impact
that compacts can have on revenue in noncompact regions when farm- level
prices are low because of the increased supply of milk available for
manufacturing purposes. Table 7 provides our estimates of the impacts of the
compact scenarios on 2000 farm- level revenue in the Upper Midwest region
and all noncompact regions. 43

Table 7: Estimated Reduction in 2000 Farm- Level Milk Revenue in the Upper
Midwest and in All Noncompact Regions Under the Compact Scenarios

Dollars in millions Estimated reduction in 2000 farm- level revenue Upper
Midwest All noncompact regions Compact scenario Dollars Percentage Dollars
Percentage

NEDC $5 0.13 $15 0.09 Expanded NEDC 17 0.44 67 0.41 Expanded NEDC and a
southern compact 18 0. 46 111 0.84 Expanded NEDC and a southern compact
using a more restrictive fluid milk trade assumption 185 4.74 307 2.33

Source: GAO analysis using the University of Wisconsin- Madison dairy model.

Appendix VI contains our estimates of the interregional impacts of compacts
on 2000 farm- level and wholesale- level dairy sector indicators.

43 We did not conduct sensitivity analyses on our 2000 estimates because of
the preliminary nature of the 2000 data used to develop the estimates and
because of potential data unreliability caused by dairy industry reactions
to USDA?s milk marketing order regulatory reforms.

Page 30 GAO- 01- 866 Impacts of Dairy Compacts

We reviewed other studies of the interregional impacts of the NEDC and
larger dairy compacts and found that the results are comparable with ours,
even though they used different methodologies. In a 1999 analysis, USDA
estimated that the impact of the NEDC on farm- level prices and dairy farm
revenue in noncompact regions during the years 2000 through 2005 would have
been minimal. 44 For example, USDA estimated that in 2000 the NEDC would
either have no impact on producer prices in noncompact regions or reduce
producer prices by about 1 cent per hundredweight of milk, or by about 0.07
percent, depending on the noncompact region of the country. An analysis
conducted by researchers at the University of California, Davis, also
estimated that the NEDC reduced producer prices in noncompact regions by
about 2 cents per hundredweight of milk, or by about 0.15 percent, on the
basis of 1999 data. 45 The researchers concluded that the NEDC had such a
small impact because the NEDC states produced such a small portion of the
nation?s milk supply. They also estimated that if the Compact had been
expanded to include additional states that produced 9 percent of the
nation?s milk supply, producer prices in noncompact regions would have
fallen by about 5 cents per hundredweight, or by about 0.35 percent. An
analysis conducted by a researcher at Pennsylvania State University of an
expanded NEDC in conjunction with a southern compact that produced 27
percent of the U. S. milk supply also concluded that compacts have a
relatively small impact. 46 Using a range of assumptions about milk prices
and data for 1997, the researcher projected that in 2000, the compact would
decrease producer prices in noncompact regions by 4 to 14 cents per
hundredweight, or by about 0.3 to 1 percent. Researchers at the University
of WisconsinMadison, using a 1997 version of the Interregional Dairy
Competition Model that we used in our analysis, also estimated that the NEDC
had a

44 USDA Agricultural Marketing Service analysis provided to the Subcommittee
on Livestock and Horticulture, Committee on Agriculture, U. S. House of
Representatives, June 14, 1999.

45 Joseph V. Balagtas and Daniel A. Sumner, ?The Effect of the Northeast
Dairy Compact on Producers and Consumers, with Implications of Compact
Contagion,? Department of Agricultural and Resource Economics, University of
California, Davis, 200l.

46 Kenneth W. Bailey, ?Evaluating the Economic Impacts of Regional Milk
Pricing Authorities: The Case of Dairy Compacts,? Agricultural and Resource
Economics Review,

Vol. 29, No. 2, Oct. 2000, pp. 208- 219. Other Economic Analyses

of Interregional Dairy Compacts Have Produced Similar Estimates

Page 31 GAO- 01- 866 Impacts of Dairy Compacts

small impact on producer prices. 47 This analysis estimated that farm- level
prices would fall from 5 to 10 cents per hundredweight under an expanded
NEDC scenario; 13 to 15 cents per hundredweight under a southern compact
scenario; and 14 to 28 cents per hundredweight under a combined expanded
NEDC and southern compact scenario. In an analysis of the impact of compacts
prepared for the International Dairy Foods Association, one researcher
estimated that the NEDC reduced farm- level revenue in noncompact regions in
2000 by about $29 million, while an expanded 29- state compact would reduce
farm- level revenue in noncompact regions by about $374 million. 48

A more detailed discussion of these studies is included in appendix VII. By
affecting the minimum prices that dairy farmers within the Compact region
receive for their raw milk, the NEDC may have enhanced dairy farmer income
in the six NEDC states, and other states such as New York, that supply raw
milk used for and sold as fluid milk in the NEDC states. It is not certain,
however, whether the NEDC will help ensure the continued vitality of dairy
farming in the New England dairy region. Data indicate that the number of
dairy farms in the six states continued to decrease following the NEDC's
establishment in July 1997. With regard to retail prices, the NEDC
contributed to increased retail fluid milk prices within the six states,
although the extent of its contribution is uncertain. Even so, available
evidence and analyses indicate that the NEDC has had little impact on dairy
farmers or consumers in noncompact regions.

Proposals are pending before the Congress for larger compacts. Our analysis
shows that as the share of the U. S. milk supply covered by compacts
increases, the estimated interregional impacts on farm- level prices and
revenue increase as well. Furthermore, these estimated impacts could be
different under new marketing conditions. Our estimates of the interregional
impacts of compacts are based primarily on data from before January 2000,
when USDA?s regulatory reforms took effect. Data since

47 Tom Cox, Bob Cropp, and Will Hughes, ?Interregional Analysis of
Interstate Dairy Compacts,? Department of Agricultural and Applied
Economics, College of Agricultural and Life Sciences, University of
Wisconsin- Madison, Marketing and Policy Briefing Paper No. 69, July 1999.

48 Allen Rosenfeld, The Impacts of the Proposed Expansion of Dairy Compacts
on Dairy Farm Revenue in Noncompact States, M& R Strategic Services,
Washington, D. C., May 2001. The International Dairy Foods Association
represents dairy food manufacturers. Concluding

Observations

Page 32 GAO- 01- 866 Impacts of Dairy Compacts

January 2000 indicate that the dairy industry is in the process of adjusting
to these substantial changes. Equally as important, our estimates of the
interregional impacts are based on three compact scenarios, the largest of
which includes fewer than the number of states currently being considered
for inclusion in dairy compacts. A thorough understanding of the impacts of
these other potential compacts on dairy sector indicators cannot be
developed until sufficient data become available following the dairy
industry?s adjustment to regulatory reform.

We provided USDA and the Executive Director of the NEDC with a draft of this
report for review and comment. On September 5, 2001, we met with USDA's
Chief Economist, Dairy Programs, Agricultural Marketing Service, and other
officials from USDA's Agricultural Marketing Service, Economic Research
Service, Farm Service Agency, Food and Nutrition Service, National
Agricultural Statistics Service, and the Department's Office of the Chief
Economist to obtain their oral comments. USDA officials stated that they
recognized the difficulty of undertaking a study of this nature and said
that our work represents a reasonable effort to estimate the intraregional
and interregional impacts of dairy compacts. They provided a number of
technical corrections and suggestions, which we incorporated as appropriate.

We also discussed the draft report with the NEDC Executive Director, who
stated that we had dealt with the issues in a constructive and comprehensive
manner. The NEDC Executive Director also provided us with written comments.
While concurring with our estimate of the interregional impacts of the NEDC,
the Executive Director expressed concern that the University of Wisconsin-
Madison dairy model did not measure the benefits that New York dairy farmers
receive when they supply milk to the NEDC states. We concur that the model
does not measure the impacts of compacts on noncompact states that are
within the same region as compact states. As the model is designed, New York
and the NEDC states, as well as several other states, are included in the
same (Northeast) region. The NEDC Executive Director's written comments and
our detailed responses appear in appendix VIII.

We performed our work between September 2000 and September 2001 in
accordance with generally accepted government auditing standards. Appendix I
contains a detailed description of our scope and methodology. Agency
Comments

and Our Response

Page 33 GAO- 01- 866 Impacts of Dairy Compacts

We are sending copies of this report to the Senate Committee on Agriculture,
Nutrition, and Forestry; the House Committee on Agriculture; other
appropriate congressional committees; the Secretary of Agriculture; the
Executive Director of the NEDC; the Director, OMB; and other interested
parties. We will also make copies available to others upon request.

Please contact me at (202) 512- 3841 if you or your staff have any questions
about this report. Another GAO contact and key contributors to this report
are listed in appendix IX.

Sincerely yours, Lawrence J. Dyckman Director, Natural Resources

and Environment

Appendix I: Objectives, Scope, and Methodology

Page 34 GAO- 01- 866 Impacts of Dairy Compacts

In May 2000, Senator Herbert Kohl requested that we examine the economic
impacts of the Northeast Interstate Dairy Compact (NEDC) and other potential
compacts on a variety of dairy sector indicators. Specifically, because
legislation authorizing the Compact is to expire on September 30, 2001, and
the Congress is considering legislative alternatives for reauthorizing the
NEDC and authorizing other states to enter into such compact arrangements,
Senator Kohl asked us to provide information on

 the intraregional impacts of the NEDC (that is, within the six NEDC states
of Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and
Vermont) on dairy sector indicators such as (1) retail milk prices, (2) milk
producer income, (3) dairy farm structure, (4) milk production, and (5) milk
consumption;

 the impact of the NEDC on the costs to the federal government of its milk
price support and nutrition assistance programs; 1 and

 the interregional impacts of the NEDC, an expanded NEDC, and an expanded
NEDC in conjunction with a southern compact (that is, on noncompact milk-
producing regions) on selected indicators such as farmlevel and wholesale-
level indicators such as prices, production, and revenue. 2

To determine the intraregional impacts of the NEDC, we sought, but did not
find, a readily usable economic model that comprehensively estimates these
impacts while holding constant other factors that also affect the selected
dairy sector indicators. Further, due to time and resource constraints, we
were not able to develop a model or series of models to estimate these
impacts. As a result, we analyzed federal, state, and other

1 The U. S. Department of Agriculture?s (USDA) milk price support program
indirectly ensures a minimum price for milk. Under the program, USDA offers
to purchase cheddar cheese, nonfat dry milk, and butter at specified prices.
USDA?s nutrition assistance programs include the Special Supplemental
Nutrition Program for Women, Infants, and Children; the School Breakfast and
Lunch Programs; and the Food Stamp Program.

2 For purposes of our analysis, the expanded NEDC includes the Compact?s
current member states plus Delaware, Maryland, New Jersey, New York, and
Pennsylvania. The southern compact would include Alabama, Arkansas, Georgia,
Kansas, Kentucky, Louisiana, Missouri, Mississippi, North Carolina, South
Carolina, Tennessee, and Virginia. As of the end of February 2001, these
states had enacted laws authorizing their entry into such compacts should
the Congress establish them. While West Virginia has also enacted such a
law, we did not include that state in our analysis because of modeling
difficulties. App. II provides a discussion of our model. Appendix I:
Objectives, Scope, and

Methodology NEDC?s Intraregional Impacts

Appendix I: Objectives, Scope, and Methodology

Page 35 GAO- 01- 866 Impacts of Dairy Compacts

data on these indicators, for a period of time before and after the NEDC?s
minimum pricing regulations became effective, to determine any changes in
historic trends in the NEDC states. In each case, we also obtained these
data for the rest of the United States so that we could compare trends in
New England with those in the rest of the country. We also reviewed
available studies on the NEDC?s potential impacts on the indicators.

Specifically, to determine the impacts on retail milk prices, we obtained
and analyzed retail milk price data from (1) A. C. Nielsen, a private data
collection and analysis company, for the Boston market as well as for other
major U. S. cities for November 1996 through September 2000; 3 (2) the
departments of agriculture in Connecticut, Maine, and New Hampshire for
November 1996 through October 2000; and (3) the International Association of
Milk Control Agencies for those states that have independent milk pricing
agencies for January 1994 through November 2000. We also reviewed available
economic analyses of the NEDC?s impact on retail milk prices and interviewed
USDA?s Agricultural Marketing Service and NEDC officials to obtain their
views on the NEDC?s impact on retail milk prices.

To examine the intraregional impacts of the NEDC on milk producer income, we
compared USDA?s Economic Research Service balance sheet and income statement
data from 1991 through 1999 for a representative composite dairy farmer in
the Service?s northeastern region with data for a farmer located outside the
northeastern region. The Economic Research Service?s northeastern region
includes Connecticut, Delaware, Massachusetts, Maryland, Maine, New
Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont.
States outside the northeastern region include all states with the exception
of these 11 states and Alaska and Hawaii. The Service was not able to
provide data for a representative composite farmer in the six NEDC states
alone because the sample size was not sufficiently large to produce reliable
balance sheet and income data. The Economic Research Service develops these
data through surveys of sampled farm operations. It collects data on
operating costs- such as feed, equipment purchases, and product
distribution- as well as data on returns, such as income received from sales
of field crops and livestock. The Service uses information obtained from
sampled farms to estimate the

3 We obtained these data in the course of our work on Dairy Industry:
Information on Prices for Fluid Milk and the Factors That Influence Them
(GAO/ RCED- 99- 4, Oct. 8, 1998) and Dairy Industry: Information on Milk
Prices and Changing Market Structure (GAO01- 561, June 15, 2001).

Appendix I: Objectives, Scope, and Methodology

Page 36 GAO- 01- 866 Impacts of Dairy Compacts

average costs of milk production in the United States and in various regions
in the country. Costs can vary significantly from farm to farm because of
differences in farm location, size, and production practices. As a result,
the costs and returns for an individual farm can vary considerably from the
average. In addition to using Economic Research Service data to estimate the
impact of the NEDC on farm income, we also estimated the average payment a
licensed dairy farmer in one of the six NEDC states may have received
between July 1997 and the end of calendar year 2000 as a result of the NEDC.
To do this, we used (1) monthly NEDC balance sheets that reflect the total
amount of milk eligible for the NEDC milk price and NEDC over- order
producer price payment amounts available for dairy farmers, (2) USDA?s
National Agricultural Statistics Service milk production data for the six
NEDC states, and (3) American Farm Bureau Federation data on the number of
licensed dairies in the six states. To determine the average payment, we
estimated what proportion of the milk eligible for the NEDC milk price could
be attributed to a licensed NEDC dairy farmer?s milk. We also reviewed
available economic analyses of the potential impacts of the NEDC on dairy
farmer income. Last, we obtained data developed by the NEDC commission on
amounts distributed to farmers as a result of the Compact, and its
assessment of the Compact's impact on farmer income.

To determine the intraregional impacts of the NEDC on dairy farm structure,
we obtained National Agricultural Statistics Service data on the total
number of cows in the NEDC and the rest of the United States, as well as
state- by- state data on the number of farms having at least one milk cow
between 1992 and 2000. We obtained data from the American Farm Bureau
Federation on the number of licensed dairies in the United States, by state,
between 1992 and 2000. We also reviewed information on factors that affect
the structure of dairy farms, and interviewed officials from the
Agricultural Marketing Service, the Economic Research Service, the National
Agricultural Statistics Service, and the NEDC commission to obtain their
views of the Compact?s impacts on farm structure.

To determine the intraregional impacts of the NEDC on milk production, we
reviewed National Agricultural Statistics Service data on the average amount
of milk produced by state and the average amount of milk produced per dairy
cow between 1993 and 2000. We also reviewed available economic analyses of
the impacts that the NEDC may have had on milk production. In addition, we
interviewed officials from the Agricultural Marketing Service, the National
Agricultural Statistics Service, and the NEDC to obtain their views on the
NEDC?s impact on milk production.

Appendix I: Objectives, Scope, and Methodology

Page 37 GAO- 01- 866 Impacts of Dairy Compacts

To examine the intraregional impacts of the NEDC on milk consumption, we
reviewed Agricultural Marketing Service data on the total amount of sales of
packaged fluid milk products in federal milk marketing orders and California
between 1996 and 1999. Such sales are representative of the consumption of
fluid milk products and account for about 93 percent of fluid milk sales in
the United States. In addition, we reviewed data on factors affecting milk
consumption and available economic studies of the NEDC?s impact on milk
consumption.

To examine the impacts of the NEDC on the costs of the federal government?s
milk price support program, we reviewed USDA Farm Service Agency analyses of
estimated amounts of milk production in the six NEDC states compared with
the rest of the United States. We also reviewed USDA and Compact data on
payments made to USDA by the NEDC. In addition, we interviewed Farm Service
Agency and NEDC officials to obtain information on payments made by the
NEDC. To examine the intraregional impacts on nutrition assistance programs,
we interviewed USDA Food and Nutrition Service officials and obtained that
agency?s analyses of the potential impact of the NEDC on its programs. We
also interviewed officials responsible for each of the six states? Special
Supplemental Nutrition Program for Women, Infants and Children and school
nutrition programs. Finally, we reviewed available economic analyses of the
estimated impact of the NEDC on nutrition assistance programs.

To examine the interregional impacts- that is, the economic impacts in other
regions of the country- of the NEDC, an expanded NEDC, and an expanded NEDC
in conjunction with a southern compact, we conducted policy simulations
using the University of Wisconsin- Madison?s Dairy Sector Interregional
Competition Model calibrated to reflect the dairy industry in 1999 (IRCM99).
We contracted with the University to have Dr. Thomas L. Cox, Professor of
Agricultural and Applied Economics and a primary developer of the model,
conduct the policy simulations. Working with Dr. Cox and consulting with
other prominent dairy economists from different regions of the country, we
developed a set of parameters for use in simulating different compacts?
impacts on dairy sector indicators.

We modeled three different compacts- the NEDC, an expanded NEDC, and an
expanded NEDC in conjunction with a southern compact- consisting of an
increasing number of states. The states in the NEDC are Connecticut, Maine,
Massachusetts, New Hampshire, Rhode Island, and NEDC?s Impacts on

Federal Programs Compacts? Interregional Impacts

Appendix I: Objectives, Scope, and Methodology

Page 38 GAO- 01- 866 Impacts of Dairy Compacts

Vermont. The states that we assumed to be included in the expanded NEDC are
these six states and Delaware, Maryland, New Jersey, New York, and
Pennsylvania. The states that we assumed to be included in the expanded NEDC
in conjunction with a southern compact are the above 11 states and Alabama,
Arkansas, Georgia, Kansas, Kentucky, Louisiana, Missouri, Mississippi, North
Carolina, South Carolina, Tennessee, and Virginia. The above states included
in an expanded NEDC and a southern compact were selected because they had
enacted legislation, as of the end of February 2001, that authorized their
entry into a dairy compact should the Congress establish one. While West
Virginia had also enacted such legislation as of the end of February 2001,
we did not include that state in a southern compact for the purposes of our
analysis because of the difficulties associated with accounting for that
state?s milk production in compact versus noncompact regions of the country.
Furthermore, because West Virginia produces a relatively small amount of
milk in comparison with other states included in compact regions, the effect
of excluding West Virginia is negligible.

The agricultural economists and other dairy experts with whom we consulted
concerning model specifications and assumptions include the following:

 Kenneth W. Bailey, Associate Professor, Department of Agricultural
Economics and Rural Sociology, the Pennsylvania State University;

 Joseph V. Balagtas, Research Assistant, Department of Agricultural and
Resource Economics, University of California, Davis;

 Scott Brown, Research Assistant Professor, Food and Agricultural Policy
Research Institute, the University of Missouri;

 Harold M. Harris, Jr., Professor, Department of Agricultural and Applied
Economics, Clemson University;

 Harry Kaiser, Professor, Department of Applied Economics and Management,
Cornell University;

 Richard L. Kilmer, Professor, Food and Resource Economics, the University
of Florida;

 Leigh Maynard, Assistant Professor, Department of Agricultural Economics,
the University of Kentucky;

 Neil Pelsue, Extension Associate Professor, Department of Community
Development and Applied Economics, the University of Vermont;

 William A. Schiek, Economist, Dairy Institute of California;

 Mark Stephenson, Senior Extension Associate, Department of Applied
Economics and Management, Cornell University;

 Daniel Sumner, Professor, Agricultural and Resource Economics Department,
University of California at Davis;

Appendix I: Objectives, Scope, and Methodology

Page 39 GAO- 01- 866 Impacts of Dairy Compacts

 Cameron S. Thraen, Associate Professor, Agricultural, Environmental, and
Development Economics, the Ohio State University; and

 Christopher Wolf, Assistant Professor, Department of Agricultural
Economics, Michigan State University.

In addition, we consulted with several agricultural economists at USDA,
including economists in the Office of the Chief Economist and the Economic
Research Service.

Our process for developing the assumptions that we used to model the
interregional impacts of dairy compacts included reviewing economic
literature to identify estimates of (1) regional supply elasticities, (2)
wholesale demand elasticities, (3) and transportation costs. We also
obtained data on market over- order premiums and compact over- order
producer prices. Finally, we interviewed USDA and other officials to obtain
information on regulations governing milk shipments among federal marketing
orders and noncompact and compact regions.

After identifying assumptions for modeling the three different compact
scenarios, we developed an initial estimate of the economic impacts of the
different compacts. We then conducted sensitivity analyses by varying the
values of our key assumptions. We provided our preliminary estimates to
several agricultural economists to obtain their views, and incorporated many
of their comments in subsequent modeling before developing our final range
of estimates. Our final estimates of the impacts of the different compact
scenarios are presented as ranges that include our initial estimates as well
as estimates from our sensitivity analyses. (A detailed discussion of the
model and assumptions, data, and data sources used is included in app. II.)

In addition to modeling the interregional impacts of the different compact
scenarios, we reviewed economic analyses that have been conducted on the
potential interregional impacts of dairy compacts. We present these reviews
in appendix VII.

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 40 GAO- 01- 866 Impacts of Dairy Compacts

This appendix describes our methodology for estimating the interregional
impacts of three compact scenarios: the six- state NEDC, an expanded NEDC,
and an expanded NEDC in conjunction with a southern compact. To estimate the
interregional impacts, we contracted with the University of Wisconsin-
Madison to use the Dairy Sector Interregional Competition Model (IRCM),
which is an interregional spatial market equilibrium model of the U. S.
dairy sector. This model is useful in estimating the impacts of different
dairy policy options, such as dairy compacts. Dr. Thomas L. Cox, Professor
of Agricultural and Applied Economics at the university and a primary
developer of the model, conducted the policy simulations for us. This
appendix describes

 the structure of the IRCM and how it estimates the interregional impacts
of dairy compacts,

 data and data sources used for conducting policy simulations of different
compact scenarios,

 how we calibrated a baseline for 1999,

 details of each scenario that we modeled,

 parameter values for our baseline and initial estimates,

 how we varied key assumptions to test the sensitivity of our initial
estimates, and

 the limitations of the model. The results of our different policy
simulations and sensitivity analyses are presented in appendixes V and VI.

The IRCM is a hedonic spatial equilibrium model of the U. S. dairy sector
that can be used to estimate the impacts of policy or program changes, such
as the establishment of compacts. 1 The model allocates the production and
consumption of raw milk and nine other different dairy commodities among 12
regions of the country and solves for the trade flows of these commodities
among those regions to achieve a spatial equilibrium. Using nonlinear
programming techniques, the model solves to ensure an efficient regional
distribution of the different dairy commodity

1 A hedonic model is one that separates products, such as farm milk, fluid
milk, and manufactured dairy commodities, into their components or
characteristics. Milk and dairy commodities are composed of different
amounts of milk fat, protein, and carbohydrates. The IRCM incorporates these
components using a vertical markets approach that transforms the primary
product- raw milk- into fluid milk and manufactured dairy commodities.
Appendix II: Methodology for Estimating the

Interregional Impacts of Dairy Compacts IRCM Structure

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 41 GAO- 01- 866 Impacts of Dairy Compacts

resources, given the demand for and supply of those resources at various
prices.

On a more technical basis, the model solution maximizes the sum of producer
and consumer welfare minus processing, transportation, and U. S. Department
of Treasury costs. 2 The model defines aggregate wholesale dairy product
demand and farm- level milk supply functions as follows:

(1a) D i = ik z 0

d ik K

k=1 (q) dq p

and (1b) S i = i w

0 (q) dq p s i

where p i s (w i ) is the price- dependent supply function for milk in the
i- th

region, with p i / w i > 0, i = 1, ?, J, and p ik d (z ik ) is the price-
dependent

demand function for the k- th dairy product consumed in the i- th region,
with p ik

d / z ik < 0, i = 1, ?, J, k = 1, ?, K. Equation (1a) is the sum of the
areas under the K demand curves in the i- th region. This can be interpreted
as a measure of consumer benefits generated by the K commodities in the i-
th region. Equation (1b) is the area under the supply curve, a measure of
milk production cost in the i- th region. The term (D i - S i ), consumer
benefits minus total production costs in the i- th region, minus
transportation costs, is a measure of net social benefits to farmers and
consumers in each region. Federal government costs are then subtracted. 3
Two steps are used to create an IRCM that models the impact of compacts in
1999: IRCM99. First, the model is calibrated to 1999 data so that baseline
estimates of key dairy sector measures of prices, production,

2 These last costs are incurred by the federal government through USDA?s
milk price support program. Under this program, USDA purchases manufactured
dairy commodities- such as cheese, butter, and nonfat dry milk- at specified
prices, thereby indirectly supporting farm- level prices of milk used to
manufacture these commodities.

3 For a more complete technical description of the optimization process, see
T. L. Cox and J. P. Chavas, ?An Interregional Analysis of Price
Discrimination and Domestic Policy Reform in the U. S. Dairy Sector,?
American Journal of Agricultural Economics, 83 (1): 89- 106.

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 42 GAO- 01- 866 Impacts of Dairy Compacts

consumption, and trade flows can be obtained on a regional basis. Second,
simulation analyses are performed to estimate the impacts of the different
compact scenarios.

In the IRCM99, milk production and dairy product consumption in the country
are divided into 12 regions that are based on the current 11 USDA federal
milk marketing orders and California. In addition, the IRCM99 accounts for
net private stocks, net government stocks/ removals, and U. S. imports and
exports. 4 Table 8 compares the IRCM99 regions, the corresponding USDA
marketing orders or states before January 2000, states included in the
IRCM99 regions, and states that had enacted legislation as of February 2001
authorizing entry into any congressionally authorized dairy compact. 5

Table 8: USDA's Marketing Orders Before January 2000, States in the IRCM99
Regions, and States That Had Authorized Entry Into a Compact by February
2001

IRCM99 regions (and pricing points) Federal milk marketing

orders before January 2000 States included in the IRCM99 regions

States that had enacted legislation authorizing entry into a compact as of
February 2001

Northeast (Boston) New England, New York- New Jersey, Mid- Atlantic
Connecticut, Delaware, the

District of Columbia, Maine, Maryland, Massachusetts, New Hampshire, New
Jersey, New York, Pennsylvania, Rhode Island, Vermont

Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New
Jersey, New York, Pennsylvania, Rhode Island, Vermont

Appalachia (Charlotte) Carolina, Tennessee Valley, Louisville- Lexington-
Evansville Kentucky, North Carolina,

South Carolina, Tennessee, Virginia

Kentucky, North Carolina, South Carolina, Tennessee, Virginia Florida
(Tampa) Upper Florida, Tampa Bay,

Southeast Florida Florida a Southeast (Atlanta) Southeast Alabama, Arkansas,
Georgia,

Louisiana, Mississippi Alabama, Arkansas, Georgia, Louisiana, Mississippi

4 U. S. dairy exports are assumed to face a perfectly elastic demand
schedule at 1999 world prices; that is, the model assumes that the United
States exhibits ?small country effects? with respect to world dairy exports.
This assumption is made because U. S. exports are a small share of world
dairy trade.

5 In January 2000, USDA reorganized its milk marketing orders and
reconfigured the thenexisting 31 orders into 11. The significance of this is
that much of the dairy sector indicator data compiled before January 2000
are organized into these 31 orders, which do not necessarily correspond to
the 11 orders. USDA also revised its formulas for setting the minimum prices
that processors must pay for milk.

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 43 GAO- 01- 866 Impacts of Dairy Compacts

IRCM99 regions (and pricing points) Federal milk marketing

orders before January 2000 States included in the IRCM99 regions

States that had enacted legislation authorizing entry into a compact as of
February 2001

Mideast (Cleveland) Michigan Upper Peninsula, southern Michigan, East
OhioWest Pennsylvania, Ohio Valley, Indiana

Indiana, Michigan, Ohio, West Virginia West Virginia

Upper Midwest (Chicago) Chicago Regional, Upper Midwest Minnesota, North
Dakota,

Wisconsin a

Central (Kansas City) Iowa, Nebraska- West Iowa, Eastern South Dakota,
Central Illinois, Southern Illinois, East Missouri, Southwest Plains,
Eastern Colorado, Greater Kansas City

Colorado, Illinois, Kansas, Missouri, Nebraska, Oklahoma, South Dakota

Kansas, Missouri Southwest (Dallas) Texas, New Mexico- West

Texas New Mexico, Texas a Western (Salt Lake City) Southwest Idaho- Eastern

Oregon, Great Basin, Western Colorado

Idaho, Montana, Nevada, Utah, Wyoming

a Pacific Northwest (Seattle) Pacific Northwest Oregon, Washington a
Arizona- Las Vegas (Phoenix) Central Arizona Arizona a California (San
Francisco) b California a

a No states in this region had enacted legislation as of Feb. 2001
authorizing entry into a compact. b California?s milk is not regulated under
a USDA milk marketing order. Rather, the state regulates milk processed and
sold in that state. Source: University of Wisconsin- Madison.

Figure 1 shows USDA?s marketing orders, the corresponding IRCM99 regions,
the corresponding states included in the IRCM 99 regions, and states that
had enacted legislation authorizing entry into a compact as of February
2001. With respect to modeling the impact of compacts, the states that have
enacted legislation are part of four different federal milk marketing orders
and their corresponding IRCM99 regions. Because of this, both compact and
noncompact states are included in some IRCM99 regions when modeling some
compact alternatives. This can influence the interpretation of modeled
results. For example, some states that have enacted legislation authorizing
entry into a dairy compact, such as West Virginia, which is part of the
Mideast Marketing Order, produce only a small portion of the milk produced
by dairy farmers in that marketing order and hence, estimating the effect of
West Virginia?s participation in a compact, would be difficult.

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 44 GAO- 01- 866 Impacts of Dairy Compacts

Figure 1: USDA?s Milk Marketing Orders

Source: USDA?s Agricultural Marketing Service.

IRCM99 solves for regional prices and production levels for farm- level raw
milk on the basis of three milk components: milk fat, protein, and
carbohydrates (primarily lactose). The model also solves for regional
wholesale- level price, supply, demand, and trade flows for the following
dairy products: (1) fluid milk, (2) soft dairy products, (3) American
cheese, (4) Italian cheese, (5) other cheese, (6) butter, (7) frozen dairy
products,

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 45 GAO- 01- 866 Impacts of Dairy Compacts

(8) other manufactured dairy products (a residual product category), and (9)
nonfat dry milk. The model uses a fixed component composition that converts
farm milk into fluid milk and different types of manufactured dairy
commodities, as shown in table 9.

Table 9: IRCM99 Dairy Product Fixed Component Composition Milk component
Dairy commodity Milk fat

(percentage) Protein (percentage) Carbohydrates

(percentage)

Fluid milk 2.0 3. 3 4.7 Soft dairy products 11.4 4. 9 4.2 American cheese
33.2 24.6 1. 5 Italian cheese 22.7 20.8 2. 4 Other cheese 28.1 25.7 2. 4
Butter 81.1 0. 9 0.1 Frozen dairy products 9.5 3. 5 4.9 Other manufactured
products 0.5 12.1 52.0 Nonfat dry milk 0.7 35.6 52.1

Source: University of Wisconsin- Madison.

The regional supply of milk components (milk fat, protein, and
carbohydrates) must be greater than or equal to the regional utilization of
these milk components by the processing sector to ensure regional component
supply/ demand balance. The marginal value of this restriction (given by the
corresponding Lagrange multiplier) measures the shadow value of each milk
component in each region. 6 The model subsequently generates empirical
estimates of regional shadow prices for each milk

6 One approach for solving a constrained optimization problem is the
Lagrange- multiplier method. This technique involves finding the maximum or
minimum value of an objective function (such as utility maximization)
subject to another expression that defines a constraint (i. e., a budget
constraint). In general, given the objective function z = f( x, y) subject
to the constraint, g( x, y) = c, the Lagrangean function is defined as Z =
f( x, y) + [c -

g( x, y)]. This augmented expression includes a variable called the Lagrange
multiplier, . This method has many comparative static properties, including
the fact that the Lagrange multiplier, , provides a measure of the
sensitivity of Z, the objective function, to changes or shifts in the
magnitude of the constraint. In a linear programming context, the Lagrange
multiplier can be interpreted as a shadow price, or imputed price, the
maximum amount that someone would be willing to pay for the next unit of an
input. In the current analysis, the model solution is characterized as the
saddle point of a Lagrangean function, in which the Lagrange multipliers
correspond to trade flow constraints across regions (See footnote 3 in this
appendix).

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 46 GAO- 01- 866 Impacts of Dairy Compacts

component. The model also generates market prices that are consistent with
milk component pricing for each commodity in each region.

The basic structure of the IRCM99 is consistent with a competitive market
equilibrium, where, at the optimum, the market price equals the marginal
cost of each commodity. However, USDA and California use a system of
classified pricing for milk that influences pricing in ways that differ from
the competitive outcome. Therefore, to incorporate the classified pricing
system, price wedges are used in the model to represent the difference
between the minimum price of milk in a particular class and the minimum
price of milk in a reference class. In the model, the reference class for
USDA?s milk marketing orders is Class III, and the reference class for
California is 4b. 7 , 8 For example, the price wedge for raw milk used for
fluid milk is the difference between the price for milk used for Class I
(fluid) and the price for milk used for Class III (cheese), plus potential
over- order premiums. 9 For USDA?s milk marketing orders in 1999, we
calculated the price of raw milk used for manufacturing (Classes II, III,
and IV) as the implicit price of milk used for Class II and Class III in the
Upper Midwest Milk Marketing Order (portions of Illinois, Iowa, Michigan,
Minnesota, North Dakota, South Dakota, and Wisconsin). USDA milk marketing
order prices for nonfat dry milk and butterfat differentials are computed by
using USDA formulas and wholesale commodity prices. California milk prices
are calculated for milk used in fluid products (Class 1); milk used for
heavy cream, cottage cheese, yogurt, and sterilized products (Class 2); milk
used in ice cream and other frozen dairy products (Class 3); milk used in
butter and nonfat dry milk (Class 4a); and milk used in cheese

7 The reference class 4b is comparable to USDA?s Class III, which is the
classification for milk used for manufacturing hard cheese. 8 California?s
Class 4a is for milk used to manufacture butter and nonfat dry milk and
corresponds to USDA?s Class IV. 9 When reviewing and commenting on our draft
report, USDA officials suggested that a different method of calculating the
Class I mover that took into account lags would be more consistent with the
method that USDA uses. We did not use that method originally because the
IRCM99 is an annual, not monthly, model. However, as a result of USDA
officials' comments, we re- ran some specifications of the model using a
method for calculating the Class I mover that more closely resembles USDA's
method. The estimates of the impacts of the different compact scenarios that
we obtained using 1999 and 2000 data were very similar to the estimates in
our draft report and, therefore, we did not change the estimates in
preparing our final report. In both sets of model runs, we used a consistent
method for measuring the Class I mover in our baseline and in our compact
scenarios. The choice of method for measuring the Class I mover, therefore,
did not have a large impact because we are estimating the impacts of
compacts by estimating deviations from a baseline. Modeling Classified

Pricing

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 47 GAO- 01- 866 Impacts of Dairy Compacts

other than cottage cheese (Class 4b). We computed California price wedges by
using administered formulas for the fat and solids- not- fat component
prices by class. These component prices are computed from wholesale
commodity prices for butter, nonfat dry milk, and cheese that are endogenous
to the model. Because Class 4a prices were lower than Class 4b prices in
1999, this method implies a negative price wedge in California for Class 4a.
Finally, because dairy farmers are paid blend prices based on USDA and
California?s classified pricing systems, we incorporated price wedges that
represent these blend prices in each of the respective modeled regions.

We took into account pooling regulations for both USDA and the NEDC in
modeling ?producer settlement pools.? 10 The first step in this calculation
is to compute total revenues from regional milk production under the
assumption that all raw milk is pooled in the region where it is produced.
This is done in each region by multiplying the price wedge for each
commodity by the quantity of raw milk used to produce that commodity.
However, by shipping some of their milk to another order, producers can
sometimes become eligible to receive the minimum prices in the destination
order for their milk, which might be higher than the minimum prices in their
"home" order. As a result, in estimating how the proceeds from raw milk
sales are distributed to producers, through producer settlement pools, the
model can adjust its initial estimate to take into account that not all raw
milk is pooled in the order where it is produced.

When we estimated the impacts of the compact scenarios, no such adjustments
were necessary because the model's solutions did not yield

10 "Producer settlement pool" is the amount of money available for payment
to producers. Processors pay into the pool amounts based on the class of
product for which they use the milk. Producers draw from the pool an amount
that is based on the blend price for milk within each marketing order. The
blend price is the monthly weighted average of the prices of milk used in
the different classes in that marketing order, with the weights determined
by the amount of milk sold for each class of use. A producer's milk is
qualified to be pooled on a particular federal milk marketing order by its
association with a pool plant that is regulated by the order. Once a
producer's milk is qualified to be pooled on the order, it is not necessary
for all of that producer's milk to be delivered to a plant regulated by that
specific order in order to qualify for the blend price. Each marketing order
has its own regulations establishing the amount of milk that must be
delivered to a regulated pool plant in order to qualify for the marketing
order's blend price. The NEDC also regulates the amount of milk that must be
delivered to plants within the six states and to plants outside the six
states that ship milk to the NEDC states in order for the processors to
qualify for the NEDC price. Modeling Producer

Settlement Pools

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 48 GAO- 01- 866 Impacts of Dairy Compacts

any pooling of raw milk outside any home order. As a result, before taking
into account further adjustments described below, the revenue received from
milk sales from farmers in any order were the same as they would have been
had no adjustment been made and all producers' receipts from milk sales were
based on the prices in their home order. We recognize that some milk is
pooled outside the order in which it is produced. However, we did not have
data on shipments of raw milk by individual farmers or shipments of fluid
milk or other dairy commodities by individual processing plants, which
limited the model's ability to estimate raw milk movements across orders. 11
Instead, the model solution includes substantial movement of packaged fluid
milk to balance supply and demand across orders. If enough packaged fluid
milk moves out of an order, the model's structure allows for further
adjustments to be made in the producer settlement pools, but there was no
region affected in this way. 12

However, the producer settlement pools were adjusted for the movement of
packaged fluid milk into compact regions under the various scenarios that we
analyzed. These adjustments were made because processors in exporting
regions shipping packaged fluid milk into compact regions are required to
pay a compact premium on this milk. Similarly, these processors can return
to their producers an amount equal to the compact premium multiplied by the
percentage of the compact region's raw milk that is sold for Class I use.

We obtained data for our analysis from several sources, including USDA, the
state of California, academia, and research institutes. We obtained most of
the price and production data from USDA and the California Department of
Food and Agriculture. In particular, we obtained farm- level milk data from
USDA?s Milk Production, Disposition and Income- 1999 Summary, and commodity
production data from USDA?s Dairy Products- 1999 Summary. We obtained
commodity price, stock, import, export, and

11 Since we are estimating the impacts of compacts by estimating deviations
from a baseline that include the same limitation, it is not clear that this
limitation has much of an effect on our estimates.

12 The model requires an adjustment to be made if more than 40 percent of
the milk packaged in an order is exported to another region. This threshold
was chosen as a proxy for the requirement that an adjustment be made if a
plurality of an order's packaged milk was sold in another region. Data Used
in the

Model

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 49 GAO- 01- 866 Impacts of Dairy Compacts

government utilization data from monthly USDA Economic Research Service
?Livestock, Dairy, and Poultry? reports. We obtained federal milk marketing
order data from USDA?s Agricultural Marketing Service Federal Market Order
Milk Statistics- 1999 Annual Summary. In addition, we obtained data on price
support levels from USDA?s Commodity Credit Corporation.

We obtained regional projections of wholesale dairy product demand by using
aggregate wholesale demand functions for the United States and regional
population data. We obtained component yields- the amount of milk fat,
protein, and carbohydrates per unit of milk and wholesale dairy product-
from a component accounting exercise that fully allocates 1999 aggregate
milk and dairy product production. 13 We obtained regional farmlevel milk
supply elasticities from the Food and Agricultural Policy Research Institute
and product demand elasticities from research conducted by Cox et al. 14 and
USDA?s Economic Research Service. 15 These demand elasticities were
estimated using USDA aggregate national time series data on commercial
disappearance and wholesale prices. The estimates are consistent with a
complete demand system specification covering all major food groups and
products. We obtained data on refrigerated and nonrefrigerated
transportation and assembly costs for farm milk from dairy researchers at
Cornell University. 16 We used USDA estimates of dairy manufacturing costs
to incorporate processing costs into our analysis.

We calibrated the model to yield solutions that are close to the observed
1999 data for farm- level and wholesale- level measures of prices and
production and to link key regional prices to commodity reference prices
used by USDA and California, such as those reported by the Chicago

13 R. Selinsky, T. L. Cox, and E. V. Jesse, ?Estimation of U. S. Dairy
Product Component Yields.? University of Wisconsin- Madison, Agricultural
Economics Staff Paper No. 355, Sept. 1992.

14 T. L. Cox, D. Lewis, and R. Selenski, Estimation of Wholesale Demand for
U. S. Dairy Products, University of Wiscons- Madison, Department of
Agricultural Economics Mimeo, 1992.

15 K. S. Huang, A Complete System of U. S. Demand for Food, USDA, Economic
Research Service, Technical Bulletin No. 1821, Sept. 1993. 16 The
methodology used to generate these cost estimates is documented at http://
cpdmp. cornell. edu/ CPDMP/ Pages/ Publications/ Pubs/ rb9709. pdf. 1999
Baseline

Calibration

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 50 GAO- 01- 866 Impacts of Dairy Compacts

Mercantile Exchange and USDA?s National Agricultural Statistics Service.
Tables 10 through 13 compare the 1999 model solutions with actual 1999 data
for farm- level prices, farm- level production, wholesale commodity prices,
and wholesale commodity production, respectively. At the farm level, the
simulated values calibrate closely with the actual data- all discrepancies
are 0.5 percent or less, except for the farm- level price in California, for
which the discrepancy is about 1 percent. At the wholesale level, the price
discrepancies for major products (fluid milk, American and Italian cheeses,
butter, and nonfat dry milk) are less than about 3 percent, while production
discrepancies are 3.5 percent or less.

Table 10: IRCM99 Farm- Level Price Calibration Farm- level prices in 1999
(dollars per hundredweight) Difference between actual and

modeled 1999 farm- level prices IRCM99 region Actual Modeled Dollars per

hundredweight Percentage

Northeast $15.21 $15.15 $( 0.06) (0.4) Appalachia 16.02 15.96 (0.06) (0.4)
Florida 17.20 17.13 (0.07) (0.4) Southeast 16.07 16.00 (0.07) (0.4) Mideast
14.70 14.64 (0.06) (0.4) Upper Midwest 13.88 13.81 (0.07) (0.5) Central
14.27 14.21 (0.06) (0.4) Southwest 14.54 14.48 (0.06) (0.4) West 13.23 13.17
(0.06) (0.5) Northwest 14.90 14.84 (0.06) (0.4) California 13.12 12.97
(0.15) (1.1) Arizona 13.80 13.75 (0.05) (0.4) United States 14.27 14.23
(0.04) (0.3)

Note: Numbers in parentheses are negative. Source: University of Wisconsin-
Madison.

Table 11: IRCM99 Farm- Level Production Calibration Farm- level production
in 1999

(pounds in millions) Difference between actual and

modeled 1999 farm- level production IRCM99 region Actual Modeled Pounds in

millions Percentage

Northeast 29,205 29,178 (27) (0.1) Appalachia 6, 450 6,435 (15) (0.2)
Florida 2, 393 2,389 (4) (0.2) Southeast 3, 553 3,547 (6) (0.2)

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 51 GAO- 01- 866 Impacts of Dairy Compacts

Farm- level production in 1999 (pounds in millions)

Difference between actual and modeled 1999 farm- level

production IRCM99 region Actual Modeled Pounds in

millions Percentage

Mideast 12,253 12,237 (16) (0.1) Upper Midwest 32,848 32,798 (50) (0.2)
Central 14,879 14,852 (27) (0.2) Southwest 10,292 10,276 (16) (0.2) West
8,869 8,853 (16) (0.2) Northwest 7, 145 7,133 (12) (0.2) California 30,408
30,451 (43) (0.1) Arizona 2, 918 2,914 (4) (0.1) United States 161,213
161,063 (150) (0.1)

Note: Numbers in parentheses are negative. Source: University of Wisconsin-
Madison.

Table 12: IRCM99 Aggregate Wholesale Commodity Price Calibration Wholesale
commodity price

in 1999 (dollars per hundredweight) Difference between actual and

modeled 1999 commodity prices Dairy commodity Actual Modeled Dollars per

hundredweight Percentage

Fluid milk $15.07 $15.09 $0.02 0.1 Soft dairy products 27.30 27.29 (0.01) 0
American cheese 139.84 139.20 (0.64) (0.5) Italian cheese 93.48 93.20 (0.28)
(0.3) Other cheese 115.68 115.49 (0.19) (0.2) Butter 122.94 119.38 (3.56)
(2.9) Frozen dairy products 21.41 21.38 (0.03) (0.1) Other manufactured
products 27.36 27.68 0.32 1.2

Nonfat dry milk 101.78 102.63 0.85 0.8 Note: Numbers in parentheses are
negative. Source: University of Wisconsin- Madison.

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 52 GAO- 01- 866 Impacts of Dairy Compacts

Table 13: IRCM99 Aggregate Wholesale Commodity Production Calibration Level
of production in 1999

(pounds in millions) Difference between actual and

modeled 1999 levels of production

Dairy commodity Actual Modeled Pounds in millions Percentage

Fluid milk 55,039 55,020 (19) 0 Soft dairy products 6,068 5,986 (82) (1.4)
American cheese 3,577 3,539 (38) (1.1) Italian cheese 3,143 3,176 33 1.0
Other cheese 578 598 20 3.5 Butter 1, 275 1,274 (1) (0.1) Frozen dairy
products 12,166 12,134 (32) (0.3) Other manufactured products 4,351 4,270
(81) (1.9)

Nonfat dry milk 1,383 1,399 16 1.1 Note: Numbers in parentheses are
negative. Source: University of Wisconsin- Madison.

Because our baseline was calibrated to actual 1999 data, our original
baseline values represented estimates for a time when the Compact was in
place. As a result, we used the IRCM99 to estimate the impact of the NEDC by
simulating the year 1999 without the NEDC and comparing those results with
those obtained in our original baseline. That is, the estimated impacts of
removing the Compact from our original baseline are interpreted as the
estimated impacts of adding the Compact to a nocompact baseline, with the
signs reversed. For each subsequent compact scenario, we used the IRCM99 to
estimate the interregional impact of compacts. That is, we compared our
estimates of farm- level and wholesale- level prices, production, and
revenue in 1999 under each compact scenario with our estimates of what they
would have been in that year without any compact. In effect, the no- compact
scenario became our new baseline for comparison.

We developed three compact scenarios: the NEDC, which includes Connecticut,
Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont; an expanded
NEDC that also includes Delaware, Maryland, New Jersey, New York, and
Pennsylvania; and an expanded NEDC in conjunction with a southern compact
that includes Alabama, Arkansas, Georgia, Kansas, Kentucky, Louisiana,
Missouri, Mississippi, North Compact Scenarios

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 53 GAO- 01- 866 Impacts of Dairy Compacts

Carolina, South Carolina, Tennessee, and Virginia. These states were
included in the expanded NEDC and the southern compact on the basis that
they had enacted legislation as of the end of February 2001 that would allow
them to enter into a dairy compact, should the Congress enact legislation
allowing them to do so. While West Virginia had also enacted such
legislation as of the end of February 2001, we did not include it in our
analysis of an expanded NEDC in conjunction with a southern compact because
of the difficulty of modeling West Virginia as part of a southern compact,
given that it would have been the only state in USDA?s Mideast Marketing
Order whose milk would fall under compact regulation. Since West Virginia
produced less than 0.2 percent of the nation?s milk supply in 2000, we do
not believe that its omission significantly affects our estimates of the
impact of the southern compact.

The IRCM99, as discussed previously, models different regions of the country
based, in part, on USDA's milk marketing orders and California. Because of
this, the states in the NEDC are incorporated as part of the IRCM99
Northeast region. This region encompasses the following states: Connecticut,
Delaware, Maine, Massachusetts, Maryland, New Hampshire, New Jersey, New
York, Pennsylvania, Rhode Island, and Vermont (and the District of
Columbia). To account for the NEDC, we calculated the amount of milk
produced by the six NEDC states as a percentage of the total amount of milk
produced within the Northeast Milk Marketing Order. Therefore, while we are
able to estimate the impact of the NEDC on other noncompact regions of the
country that do not contain any compact states, we could not estimate the
NEDC's impacts on other noncompact states within the Northeast region.
Further, because we defined the impacts of the NEDC scenario to be based on
the amount of milk produced by the six NEDC states, the impacts of this
scenario do not account for milk that noncompact states, such as New York,
may ship into the NEDC states. Further, reductions in revenue in noncompact
states are also not accounted for. Thus, our estimate of the NEDC's impact
on noncompact regions is a little smaller than would be expected had we been
able to isolate the effects of noncompact and compact states within the same
region.

Similarly, in our scenario of an expanded NEDC in conjunction with a
southern compact, the IRCM99 has the compact region extending into four
regions: the Appalachian, Central, Northeast, and Southeast. While this
compact scenario fully includes the majority or all of the states
encompassed by three of the four regions, one region- the Central- is not
fully encompassed. The Central region includes the following seven states:
Colorado, Illinois, Kansas, Missouri, Nebraska, Oklahoma, and South

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 54 GAO- 01- 866 Impacts of Dairy Compacts

Dakota. However, only two of these states- Kansas and Missouri- are assumed
as part of a compact. As with our modeling of the NEDC, we calculated the
amount of milk produced by the two states as a percentage of the total
amount of milk produced within the Central region. Therefore, while we are
able to estimate the impact of the expanded NEDC in conjunction with a
southern compact on noncompact regions of the country that do not contain
any compact states, we could not estimate this compact scenario's impact on
the Central region. Similarly, under this compact scenario, we do not take
into account the amount of milk that may be shipped into the compact states
by bordering noncompact states in the Central region. Further, reductions in
noncompact producer revenue in these noncompact states was also not
accounted for. Thus, our estimate of the expanded NEDC in conjunction with a
southern compact's impact on all noncompact regions combined is a little
smaller than would be expected had we been able to isolate the effects of
noncompact and compact states within the Central region.

We used specific parameter values to arrive at our baseline, or nocompact,
scenario and initial estimates of the impacts of the different compact
scenarios. We also used key assumptions regarding the amount of market-
driven over- order premiums, compact producer price payment levels, and
transportation costs, and the ability of milk to move from region to region.

Specifically, we used a milk price support payment amount, or price floor,
of $9.90 per hundredweight of milk, which equates to $1.10 per pound for
American cheese, 65 cents per pound for butter, and $1.01 per pound for
nonfat dry milk. In addition, we assumed medium- term, or 5- year, wholesale
demand elasticities for fluid milk and other dairy products, as shown in
table 14.

Table 14: Medium- Term Wholesale Demand Elasticities Used in Initial
Estimates Dairy commodity Medium- term demand elasticity

Fluid milk -0.144 Soft dairy products -0.42 American cheese -0.16 Italian
cheese -0.251 Other cheese -0.477 Butter -0.243 Frozen dairy products -0.327

Parameter Values for Our Baseline and Initial Estimates

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 55 GAO- 01- 866 Impacts of Dairy Compacts

Dairy commodity Medium- term demand elasticity

Other manufactured products -0.276 Nonfat dry milk -0.449

Source: Cox et al., 1992, and USDA?s Economic Research Service.

We also assumed medium- term regional supply elasticities as shown in table
15.

Table 15: Medium- Term Regional Supply Elasticities Used in Initial
Estimates IRCM99 region Medium- term supply elasticity

Northeast 0.258 Appalachia 0. 641 Florida 0.408 Southeast 0.396 Mideast
0.317 Upper Midwest 0. 312 Central 0.430 Southwest 0.388 West 0.424
Northwest 0.430 California 0.433 Arizona 0.389

Source: Food and Agricultural Policy Research Institute.

We also assumed that within a compact region, there was no marketdriven
over- order premium above the compact minimum price. This initial assumption
represents a lower bound, and assumes that compact overorder producer price
payments may have replaced, or become a substitute for, much of the market-
driven over- order premiums in the compact region. We also assumed that the
Class I price in a compact region, or the compact price, was $16.94 per
hundredweight of milk.

To better model the effect of pooling and to account for milk shipments from
region to region and from noncompact to compact regions, we assumed that no
more than 40 percent of any one region?s packaged fluid milk could be
shipped to another region without being pooled in the receiving region. We
also assumed that a processor had to pay the compact over- order producer
price into a compact pool for milk shipped into a compact region in order to
receive a compact blend price for these fluid shipments. With these
assumptions, we did not restrict milk flows from region to region or from
noncompact to compact regions. We also used transportation costs as
developed by researchers at Cornell University, with no adjustment.

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 56 GAO- 01- 866 Impacts of Dairy Compacts

Following our initial analyses, we conducted additional analyses of the
impacts of compacts by varying our initial parameter values or assumptions
to determine how sensitive our initial estimates were to changes in key
values or assumptions. Specifically, with respect to wholesale demand
elasticities, we changed the medium- term elasticities used in our initial
analyses to long- term, as shown in table 16, to determine what impact
higher elasticities would have on our initial estimates.

Table 16: Long- Term Wholesale Demand Elasticities Used in the Sensitivity
Analyses

Dairy commodity Long- term demand elasticity

Fluid milk -0.30 Soft dairy products -0.907 American cheese -0.625 Italian
cheese -0.741 Other cheese -0.477 Butter -0.42 Frozen dairy products -0.42
Other manufactured products -0.45 Nonfat dry milk -0.60

Source: GAO, based on an analysis of pertinent agricultural economics
literature.

In a separate analysis, we changed our regional supply elasticities to
longterm, as shown in table 17, to determine if changes in long- term supply
elasticities would affect our initial estimates.

Table 17: Long- Term Regional Supply Elasticities Used in the Sensitivity
Analyses IRCM99 region Long- term supply elasticity

Northeast 0.431 Appalachia 0. 951 Florida 0.674 Southeast 0.629 Mideast
0.517 Upper Midwest 0. 514 Central 0.666 Southwest 0.614 West 0.727
Northwest 0.748 California 0.793 Arizona 0.605

Source: Food and Agricultural Policy Research Institute.

Sensitivity Analyses

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 57 GAO- 01- 866 Impacts of Dairy Compacts

In a separate analysis, we changed our assumption regarding the lack of any
market- driven over- order premium within compact regions and added a 50-
cent- per- hundredweight over- order premium above the compact's minimum
price to determine what impact this would have on our initial estimates. We
then simultaneously used long- term supply and demand elasticities in
conjunction with the 50- cent over- order premium in compact regions to
determine what combined effect these three changes taken together would have
on our initial estimates.

In a separate analysis, we inflated transportation costs that we obtained
from Cornell University by 25 percent within each of the IRCM99 regions to
determine what effect increased transportation costs would have on our
initial estimates.

In addition, we conducted a separate sensitivity analysis specific to the
expanded Northeast Compact in conjunction with a southern compact scenario.
In this analysis, we varied the assumption regarding the Class I minimum
price, or compact price, in the compact region. We increased the minimum
price from $16. 94 to $18.00 per hundredweight in the southern compact but
retained the $16.94 minimum price in the expanded Northeast Compact. We
conducted this analysis because data on cooperative pay prices in selected
cities in USDA?s Appalachia, Southeast, and Central milk marketing orders
were about a dollar higher in 1999 than in the Northeast Milk Marketing
Order. 17

We also analyzed how sensitive the model was to trade limitations across
regions. Our initial estimates and previous sensitivity analyses assumed
that milk flowed relatively freely between noncompact and compact regions.
We revised this assumption by limiting the amount of milk that could flow
into a compact region from a noncompact region to that amount of milk
produced within 100 miles of a compact region?s border. Although we
recognize that there are no regulations establishing such a limit, we
performed this particular analysis to reflect the fact that the IRCM99 uses
average transportation costs, on a regional basis, because of the lack of
specific data on the location of processing plants for 1999. This limitation
may lead to a model solution with unrealistically high interregional milk
trade. We then compared the estimated impacts of the

17 Each month, cooperative associations announce cooperative pay prices. A
cooperative pay price is based on a blend price- a weighted- average value
based on milk usage- plus the costs of services performed by the cooperative
association and cooperative membership dues, and other costs such as the
national advertising and promotion program.

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 58 GAO- 01- 866 Impacts of Dairy Compacts

different compact scenarios with the impacts of the no- compact scenario
using this revised trade assumption. We also conducted the same sensitivity
analyses discussed above to determine how sensitive these new estimates were
to changes in key values and assumptions.

We also conducted an analysis of how sensitive our initial estimates were to
the magnitude of the difference between USDA and compact prices. We derived
our initial estimates and our sensitivity analyses using 1999 pricing data
because 1999 was the most recent year for which complete data were
available. During 1999, the national average blend price of milk, or the
average weighted minimum farm- level price, was $14.09 per hundredweight,
which was higher than the prices observed in some other years, including
2000. For example, the national average blend price of milk in 2000 was
$12.11 per hundredweight. To determine how sensitive our 1999 estimates
were, we used preliminary milk pricing data for 2000 to determine what
impact this difference may have had. This required a new baseline for
assessing the magnitude of change that compacts made, given the relative
magnitude of change in compact and USDA prices. We did not, however, conduct
a full range of sensitivity analyses regarding this change in assumptions
because of the preliminary nature of the 2000 pricing data. Further, the
2000 pricing data reflected regulatory reform measures implemented by USDA
in January 2000, which posed potential data reliability concerns.

We calibrated the model using dairy sector data for 2000 to simulate the
impact of compacts during that year. The method used to develop the IRCM00
was similar to the method used to develop the IRCM99, with two exceptions:

 We used data on 2000 regional farm production and milk prices; aggregate
commodity supply/ demand balance (commodity production, imports, exports,
and stocks) and prices; and component balance (using 2000 disaggregate
commodity production data) to develop the base model.

 We incorporated revised January 2000 USDA milk marketing order pricing
regulations into the model. These revisions include a methodology for
computing (1) classified prices based on USDA?s ?Final Rule? multiple
component pricing, (2) Class I prices based on the higher of the Class III
or Class IV multiple component price, and (3) minimum Class II prices based
on adding a 70- cent- per- hundredweight premium to Class IV prices.

As in the IRCM99, all classified pricing commodity wedges are computed
relative to Class III prices, so that if Class IV prices are higher than
Class III prices, additional classified pricing premiums are added to Class
I, II,

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 59 GAO- 01- 866 Impacts of Dairy Compacts

and IV prices. Tables 18 and 19 summarize the additional classified pricing
premiums for USDA and California.

Table 18: USDA?s Classified Prices, and the Difference Between These Prices
and the Price for Cheese (USDA's Class III or California?s Class 4b) in 2000

Dollars per hundredweight Class USDA's class price

Difference between USDA's class price and USDA?s Class III

price of $9.74 Difference between

USDA's class price and California?s Class 4b

price of $9.69

I minimum $14.03 $4.29 $4.34 I actual 15.31 5.57 5.62 II 12.53 2.79 2.84 III
9.74 0 0. 05 IV 11.83 2.09 2.14

Source: USDA?s Agricultural Marketing Service.

Table 19: California's Classified Prices, and the Difference Between These
Prices and the Price for Cheese (USDA's Class III or California?s Class 4b)
in 2000

Dollars per hundredweight Class

California's milk marketing program price

Difference between California's milk marketing program price

and USDA?s Class III price of $9.74

Difference between California's milk marketing program price and
California?s Class 4b price of $9. 69

1 $13.38 $3.64 $3.69 2 12.34 2.60 2.65 3 12.20 2.46 2.51 4b 9.69 (0.05) 0 4a
11.82 2.08 2.13

Note: Numbers in parentheses are negative. Source: USDA?s Agricultural
Marketing Service.

USDA?s revised pricing regulations result in larger Class I and II premiums
over Class III milk than occur under California?s pricing regulations, as
shown in table 20.

Appendix II: Methodology for Estimating the Interregional Impacts of Dairy
Compacts

Page 60 GAO- 01- 866 Impacts of Dairy Compacts

Table 20: IRCM00 Classified Component Price Premiums in Federal Milk
Marketing Orders and California

Dollars per hundredweight Commodity premium

Dairy commodity (Class)

California Federal marketing order

Fluid milk (I or 1) $3.92 $4.35 Soft dairy products (II or 2) 2.59 2.54
Frozen dairy products (II or 3) 2.32 2.36 Butter (IV or 4a) 0.71 0.73 Nonfat
dry milk (IV or 4a) 22.37 23.02 Residual (IV or 4a) 17.51 18.02

Source: University of Wisconsin- Madison.

USDA Class I and II premiums are also higher in 2000 than in 1999 because of
lower Class III prices in 2000 relative to 1999, and the impact that Class
IV prices had on Class I and II. Also, in modeling the impacts of compacts
in 2000, we computed commodity premiums in producer price settlement pools
on the basis of fat and skim- not- fat commodity components, using revised
classified fat and skim- not- fat prices.

As with any modeling exercise, the IRCM has certain limitations. These
limitations include: (1) the model cannot identify individual shipments of
raw milk from one milk marketing order to another because we do not have
data on shipments of raw milk and packaged fluid milk by individual farmers
and processors, respectively, (2) the model is static and does not take into
account dynamic adjustments, (3) the model ignores some institutional and/
or historical rigidities and capacity constraints (other than for the milk
marketing orders and component flow constraints), (4) the model assumes the
absence of farmer, processor, and/ or retailer market power, and (5) the
model takes exports and imports as being exogenous to the model. IRCM
Limitations

Appendix III: NEDC?s Intraregional Impacts Page 61 GAO- 01- 866 Impacts of
Dairy Compacts

The intraregional impacts of the NEDC on (1) retail milk prices, (2) milk
producer income, (3) dairy farm structure, (4) milk production, and (5) milk
consumption are difficult to determine. Data indicate that retail milk
prices increased when the NEDC?s alternative minimum pricing requirement
took effect in July 1997, and prices continue to remain relatively high
compared with retail milk prices in the rest of the country. However,
because many factors affect retail milk prices, we were unable to determine
what portion of the retail price increases in the NEDC states was due to the
NEDC as opposed to other factors. With regard to milk producer income, when
the NEDC price has been higher than the Class I price, dairy farmers have
received payments that reflect the NEDC price- that is, the Class I price
plus the difference between the Class I price and the NEDC price. (This
difference is called an over- order payment.) However, it is likely that
farmers would have received some portion of the difference even without the
Compact in the form of market- driven overorder premiums. Data on the
structure of dairy farms, milk production, and fluid milk consumption in the
six NEDC states and data for the rest of the United States show similar
trends, suggesting little or no change in the NEDC states following the
Compact?s establishment.

Retail milk prices increased in the NEDC states in July 1997, when the
Compact?s alternative minimum price for raw milk to be used for and sold as
fluid milk in the six states took effect, but not in the rest of the
country. Data indicate that since July 1997, retail prices in Boston and
other selected cities and towns in New England have remained relatively
high, compared with prices in other major cities in the country. It is
difficult to determine, however, what portion of the retail price increase
can be attributed to the NEDC?s alternative minimum price as opposed to
other factors that affect retail milk prices. Economic analyses of the
NEDC?s impact on retail milk prices have concluded that retail milk prices
increased following the Compact?s establishment, but the estimated amount
attributable to the Compact?s higher alternative minimum price varies among
the different studies.

The retail price of milk in the New England states increased sharply during
July 1997, when the NEDC began setting minimum prices that processors must
pay for raw milk used for and sold as fluid milk- milk used for drinking
purposes- in the NEDC states, but comparable increases did not occur in most
other locations in the United States. In that month, the NEDC price-$ 1.46
per gallon- was 26 cents higher than USDA?s milk marketing order Class I
price of $1.20 per gallon. This 26- cent per gallon Appendix III: NEDC?s
Intraregional Impacts

Retail Milk Prices Increased in the NEDC States, but It Is Difficult to
Determine the Amount Attributable to the Compact

Retail Milk Prices Increased in July 1997 in New England but Not in the Rest
of the Country

Appendix III: NEDC?s Intraregional Impacts Page 62 GAO- 01- 866 Impacts of
Dairy Compacts

(or 22 percent) price increase appears to have been passed on to consumers
at the retail level on fluid milk prices during the same month. Table 21
shows the increase in retail prices according to data collected by the
departments of agriculture in Connecticut, Maine, and New Hampshire; the
International Association of Milk Control Agencies; 1 and A. C. Nielsen. 2

Table 21: Retail Prices for a Gallon of Milk in NEDC States Before and
Immediately Following the NEDC?s Establishment

Organization collecting the data June 1997 July 1997

Connecticut, Maine, and New Hampshire departments of agriculture $2.43 $2.58
International Association of Milk Control Agencies a 2.49 2.66 A. C. Nielsen
b 2.40 2.60 a Association data are an average for five cities or towns in
Maine; Boston, Massachusetts; and three cities or towns in Vermont. b A. C.
Nielsen data for Boston, Massachusetts, which we selected to represent the
NEDC.

Source: GAO analysis of data from Connecticut, Maine, and New Hampshire
departments of agriculture; the International Association of Milk Control
Agencies; and A. C. Nielsen.

Data from the International Association of Milk Control Agencies indicate
that between June and July 1997, the average retail price of a gallon of
milk increased by about an average of 7 percent in nine cities in Maine,
Massachusetts, and Vermont. A. C. Nielsen data for Boston indicate that
retail milk prices increased by an average of about 8 percent during the
same period.

Our review of data for the rest of the United States indicates few increases
in the retail price of milk in July 1997 that were comparable to what
occurred in the New England states. For example, data from the International
Association of Milk Control Agencies for 42 cities and regions outside the
NEDC states indicate that only two of those cities or regions experienced
retail price increases comparable to or larger than those observed in
selected cities in Maine, Massachusetts, and Vermont in July 1997: Eastern
Virginia had a 31- cent increase and Reno, Nevada had a

1 The International Association of Milk Control Agencies is an organization
whose members represent states and territories that regulate milk prices
within their jurisdictions. Among other things, members collect data on
retail milk prices for milk sold in selected cities within their
jurisdictions.

2 A. C. Nielsen is a private data collection and analysis company.

Appendix III: NEDC?s Intraregional Impacts Page 63 GAO- 01- 866 Impacts of
Dairy Compacts

10- cent increase between June and July 1997. Moreover, data from A. C.
Nielsen for 13 major cities outside the NEDC states show an increase in
retail milk prices in only one city between June and July 1997: in Seattle,
the price of milk rose from $2.91 to $3.01 a gallon. Elsewhere, retail milk
prices declined, for example, in Cincinnati, from $1.43 to $1.35 per gallon
between June and July 1997 and in Washington, D. C., from $2.47 to $2.44 per
gallon.

In the longer term, between July 1997 and September 2000, retail milk prices
in most of the United States did not increase as much as they did in the
NEDC states, according to A. C. Nielsen data, as shown in figure 2. 3 The
average retail milk price in Boston increased by 18 percent between June
1997 and September 2000, from $2.40 to $2.84 per gallon. The national
average retail price of milk increased by 8 percent during the same period,
from $2.49 to $2.68 per gallon.

3 We are using A. C. Nielsen data for Boston to represent the NEDC states.
From July 1997 to

September 2000, Retail Milk Prices in the Rest of the Country Have Increased
Less Than in the NEDC

Appendix III: NEDC?s Intraregional Impacts Page 64 GAO- 01- 866 Impacts of
Dairy Compacts

Figure 2: Comparison of the Average Retail Price of a Gallon of Milk in
Boston and in the United States, November 1996- September 2000

Source: GAO analysis of data from A. C. Nielsen.

A. C. Nielsen data for 13 major cities outside the NEDC indicate that retail
milk prices increased between June 1997 and September 2000 in 10 of those
cities as well. For example, retail prices increased by 3 percent in
Phoenix, or from $2.21 to $2.27 per gallon; by 12 percent in Seattle, or
from $2.91 to $3.26 per gallon; by 17 percent in New Orleans, from $2.58 to
$3.03 per gallon. Data for one city- Cincinnati- show a greater price
increase than in Boston: prices increased by 80 percent, from $1.43 to $2.57
per gallon, between June 1997 and September 2000. However, data for
Cincinnati indicate that between March 1997 and May 1998, retail milk prices
were considerably lower than those in other major cities. In contrast to
retail milk prices in cities such as Boston and Cincinnati, prices fell in
Dallas, Denver, and San Diego.

Data from the International Association of Milk Control Agencies also
indicate that retail milk prices increased from June 1997 to November 2000
in the NEDC states and the rest of the country. For example, retail prices
increased by 17 percent, from $2.59 to $3. 04 in Burlington, Vermont; and

Appendix III: NEDC?s Intraregional Impacts Page 65 GAO- 01- 866 Impacts of
Dairy Compacts

by 10 percent, from $2.18 to $2. 39 in Augusta, Maine, from June 1997 to
November 2000. In the rest of the country, of the 29 cities or regions for
which the Association had June 1997 and November 2000 data, retail milk
prices increased in 21 of them. For example, retail milk prices increased by
4 percent, from $2.76 to $2.88 per gallon in Philadelphia, Pennsylvania; and
by 20 percent, from $2.55 to $3.05, in Salem, Oregon.

Even though it is likely that the Compact caused some portion of the retail
milk price increase in the NEDC states, it is difficult to determine the
size of that portion. In part, this is because retail milk prices vary
considerably in relation to the minimum farm- level price for raw milk. For
example, after retail prices increased by about 20 cents per gallon in July
1997, they fell by about 5 to 7 cents per gallon for a period of several
months. In addition, even though the NEDC price remained stable at $1.46 per
gallon between July 1997 and August 1998, data on retail milk prices for
Boston indicate that retail prices fluctuated during that period from a high
of $2.60 to a low of $2.53 per gallon. Figure 3 shows the relationship
between the Class I or NEDC price and the average retail price of milk sold
in Boston.

Appendix III: NEDC?s Intraregional Impacts Page 66 GAO- 01- 866 Impacts of
Dairy Compacts

Figure 3: Comparison of the Average Retail Price of a Gallon of Milk in
Boston and the USDA Class I or NEDC Price, November 1996- September 2000

Note: The NEDC price took effect beginning in July 1997. The NEDC price per
gallon is the higher of either USDA?s Class I price or the NEDC floor price
of $1.46 per gallon.

Source: GAO analysis of data from A. C. Nielsen, USDA?s Agricultural
Marketing Service, and the NEDC.

Between July 1997 and September 2000, the retail price in Boston varied
between $2.53 and $2.90 per gallon. In September 1998 when the NEDC price
increased by 9 cents, from $1.46 to $1.55 per gallon, the retail price
increased by only 1 cent, from $2.55 to $2.56. As figure 3 shows, retail
prices increased in early calendar year 1999 and again in late calendar year
1999. In November 1999, the retail price of milk reached a high of $2.90,
when the NEDC price was $1.68 per gallon.

One of the reasons that there is not a close relationship between the NEDC
price for milk and the retail price is that many factors affect the retail
price of milk, including wholesalers? costs, state regulations,

Appendix III: NEDC?s Intraregional Impacts Page 67 GAO- 01- 866 Impacts of
Dairy Compacts

consumer demand, and retailers? pricing strategies. 4 More specifically, the
retail price of milk is affected by wholesalers? costs of acquiring and
processing raw milk and packaging and distributing processed fluid milk to
retail outlets. The retail price can also be affected by state regulations
that, for example, dictate how and where milk can be distributed. Another
factor is consumers? shrinking demand for milk products, as opposed to other
beverages. This shrinking demand has created a need to advertise and improve
products, which has increased retail costs. Finally, retail milk prices are
affected by retail pricing strategies involving such factors as retail
costs, competitor pricing, if and how milk prices are used to attract
customers, shopping convenience, the image a store may want to project
regarding quality or low prices, and the extent to which retailers exercise
market power.

Four studies have estimated the Compact?s impact on retail milk prices, and
each has concluded that the NEDC has resulted in increased retail prices.
Each study provides a different estimate of the amount that the NEDC has
caused retail prices to increase, however, largely because of the different
methodologies used and the time frames analyzed. The Office of Management
and Budget (OMB) issued a study in 1998 that analyzed retail milk price data
for the first 6 months that the NEDC was in effect. 5 OMB estimated that the
Compact could have had a small impact (an increase of 5 to 10 cents per
gallon), a medium impact (an increase of 10 to 15 cents per gallon), or a
large impact (an increase of 15 to 20 cents per gallon) on retail prices,
depending on the extent to which costs were passed on from the farm to the
retail level, and the extent to which wholesalers and retailers absorbed or
passed on any increased costs. However, OMB cautioned that its study was
completed too soon after the Compact began operating in July 1997 to
determine its economic impacts and implications with confidence or
precision. OMB further cautioned that U. S. dairy industry economics are
complex, and that producer, wholesale, and retail prices are affected by
numerous proprietary, regional, and national factors. OMB concluded that
retail price patterns have fluctuated in recent years and provide no
definitive indication of the retail price levels that would have occurred
without the Compact.

4 A detailed discussion of these various factors is included in our report
entitled Dairy Industry: Information on Milk Prices and Changing Market
Structure (GAO- 01- 561, June 15, 2001).

5 OMB?s The Economic Effects of the Northeast Interstate Dairy Compact, Feb.
1998. Other Studies Estimate

That the NEDC Has Increased Retail Prices

Appendix III: NEDC?s Intraregional Impacts Page 68 GAO- 01- 866 Impacts of
Dairy Compacts

In 2000, economists (Lass et al.) at the University of Massachusetts issued
a study that analyzed the NEDC?s impact on retail milk prices during the
first year of the Compact. 6 For this study, which was conducted for the
NEDC, Lass et al. used data from January 1982 through June 1996 to develop a
model of farm- to- retail price behavior in two markets: Boston and
Hartford. They then used data from an 18- month period- July 1996 through
December 1997- to predict what the retail price for milk would have been
without the Compact and to compare those predicted effects with the effects
that actually occurred in New England. Lass et al. concluded that the NEDC
caused an average retail milk price increase of about 7 cents per gallon in
the Boston market and about 6 cents per gallon in the Hartford market. They
also concluded that because this estimated retail price increase was less
than the NEDC increase in costs to fluid milk processors (the over- order
amount due to the NEDC), an amount less than the NEDC over- order amount was
being passed on to consumers. The authors cautioned, however, that the model
did not capture changes that may have occurred in the farm- to- retail price
relationship from July through December 1997.

A third study, issued in 2000 by an economist (Bailey) at Pennsylvania State
University, analyzed the farm- to- retail markup for fluid milk over the
period January 1996 through December 1999. 7 Using a simple but direct
markup model to evaluate the impact of the NEDC on retail fluid milk prices
in Boston and Hartford, Bailey concluded that the retail price of milk
increased after the Compact established its minimum price in July 1997.
Specifically, Bailey concluded that the retail price of milk rose 24 cents
per gallon from July 1997 to December 1999 over the average price in effect
from January 1996 to June 1997. According to Bailey, the majority of this
24- cent per gallon increase- 17 cents per gallon- was attributable to the
Compact, while the rest was attributable to other factors. 8 Finally,
studies issued in 2001 by economists (Cotterill and Franklin) at the
University of Connecticut examined specific factors that

6 Daniel A. Lass, Mawunyo Adanu, and P. Geoffrey Allen, ?Impacts of the
Compact on Fluid Milk Retail Prices,? The Northeast Interstate Dairy
Compact: Milk Market Impacts,

Research Report 73, Agricultural Experiment Station, University of Vermont,
Mar. 2000. 7 Kenneth Bailey, Report on the Operation and Performance of the
Northeast Interstate Dairy Compact, Department of Agricultural Economics and
Rural Sociology, Pennsylvania State University, July 2000.

8 The increase in the retail price of milk was partially offset by a 3-
cent- per- gallon decline in the retail price of milk due to the reduction
in the cooperative market over- order premium.

Appendix III: NEDC?s Intraregional Impacts Page 69 GAO- 01- 866 Impacts of
Dairy Compacts

may have increased retail milk prices in four New England marketing areas-
Boston, Providence, Hartford/ Springfield, and Northern New England-
following the Compact?s establishment in July 1997. 9 Cotterill and Franklin
compared retail price data from February 1996 through early July 1997 with
retail price data from July 1997 through August 1998. The authors separated
out retail price increases as caused by four factors: (1) the increased farm
price of milk caused by the NEDC; (2) the increased farm price of milk
caused by strong raw milk markets when the farm price spiked above the NEDC
price; (3) nonmilk costs, such as increased processing costs other than
costs for purchasing raw milk and increased distribution costs; and (4)
changes in pricing at the wholesale and retail levels. Using this
methodology, Cotterill and Franklin concluded that of the 29- cent increase
in the retail price of a gallon of milk, on average, 2.7 cents per gallon
were caused by the NEDC and 6.5 cents per gallon were caused by strong milk
markets. 10

While it is likely that the NEDC has stabilized producer income, it is
difficult to determine how large of an impact it has had on producer income
because of uncertainty about what dairy farmers in the six NEDC states would
have been paid for their milk in the absence of the Compact. This
uncertainty arises because NEDC payments made to dairy farmers may be, in
part, substitutes for market- driven over- order premium payments. Even so,
the NEDC commission has concluded that the Compact has had a positive effect
on the financial status of dairy farmers in the six NEDC states and New
York. USDA data on dairy farm income do not clearly indicate whether the
improved financial status of an average dairy farmer in the NEDC states
resulted from the NEDC. Finally, while two studies of the NEDC?s impact on
farm income concluded that the NEDC could have a positive impact on farm
income, they provide no definitive estimate of the size of this impact.

9 Testimony by Ronald W. Cotterill on The Impact of the Northeast Dairy
Compact and Market Channel Pricing Strategies on the Performance of the New
England Dairy Industry before the U. S. Senate Committee on the Judiciary,
July 25, 2001; and Ronald W. Cotterill and Andrew W. Franklin, The Public
Interest and Private Economic Power: A Case Study of the Northeast Dairy
Compact, Food Marketing Policy Center, Department of Agricultural and
Resource Economics, University of Connecticut, May 2, 2001.

10 Cotterill and Franklin, in their May 2001 study, attributed 4.5 cents of
the 29- cent per gallon milk price increase to the NEDC. They revised this
estimate to 2.7 cents in the July 2001 study. The 1. 8 cents- per- gallon
reduction in the amount attributable to the NEDC was caused by a reduction
in cooperative premiums. Limited Data Are

Available to Estimate the Potential Impacts of the Compact on Producer
Income

Appendix III: NEDC?s Intraregional Impacts Page 70 GAO- 01- 866 Impacts of
Dairy Compacts

We cannot determine how much impact the NEDC has had on dairy farmer income
because we do not know what dairy farmers? incomes would have been from July
1997 to the present in the absence of the NEDC. While farmers would have
received at least the minimum federal marketing order or state prices, in
some cases it is likely that they would have received an amount greater than
these prices even in the absence of the NEDC. According to USDA officials,
dairy farmers in the NEDC states had received some market- driven over-
order premiums prior to the NEDC?s establishment. For example, in 1996, the
over- order Class I premium in the New England order averaged 76 cents per
hundredweight. 11 Research conducted by economists at Cornell University
indicates that after the compact began, producers receiving NEDC payments
received no over- order premiums above the amount needed to compensate them
for cooperative services. In one survey conducted in August 2000, dairy
farmers in New York who received NEDC payments were receiving no market-
driven over order premiums, while producers not shipping milk into the NEDC
states received about 60 cents per hundredweight in market- driven over-
order premiums. USDA officials concurred with this assessment. According to
one USDA official, overorder charges decreased about 50 cents per
hundredweight after the NEDC began setting its price for Class I milk, and
were at levels indicative of the costs of services provided by cooperatives
for handlers.

We estimate, however, that dairy farmers supplying raw milk for use as fluid
milk sold in the NEDC states have received revenue as a result of the
Compact. We estimate that an average licensed dairy farmer located in one of
the six NEDC states received annual payments of between $3,892 and $15,301
since the NEDC regulations took effect in July 1997 through December 2000,
as shown in table 22.

11 The difference between the price actually paid for Class I milk and the
Federal order minimum Class I price, called the "over- order premium," is
negotiated between dairy cooperatives (which sell milk produced by their
farmer members) and fluid milk processors. The over- order premium includes
the price for cooperative services and any additional price that fluid milk
processors must pay to obtain needed supplies that are driven by market
forces (that is, supply and demand for Class I milk). NEDC Payments May Be,

in Part, Substitutes for Market- Driven Premium Payments

Appendix III: NEDC?s Intraregional Impacts Page 71 GAO- 01- 866 Impacts of
Dairy Compacts

Table 22: Average Annual NEDC Over- Order Payment Received by a Dairy Farmer
Located in the Six NEDC States Calendar year Number of dairy farms in

the NEDC states Milk produced within the

NEDC states (pounds in millions)

Average NEDC net producer price (dollars

per hundredweight) Average annual dairy

farm payment as a result of the NEDC (dollars per licensee)

Last half of 1997 3,237 2,272.2 $. 93 $6,582 1998 3,049 4,663.0 .25 3,892
1999 2,964 4,691.0 .49 7,931 2000 2,772 4,664.2 .91 15,301

Source: GAO analysis of data from USDA and the NEDC.

As table 22 shows, the dairy farm payment resulting from the difference
between the USDA Class I price and the NEDC price (the over- order producer
price payment) increased significantly in 2000, when the USDA Class I price
was low-$ 14.80 per hundredweightative to the NEDC price of $16.94. 12

Although we are unable to determine how much of an impact the NEDC has had
on dairy farmer income in the six NEDC states, dairy farmers supplying milk
and receiving over- order producer price payments have likely benefited from
the NEDC. At a minimum, the NEDC has had a stabilizing impact on the prices
paid to farmers for milk, irrespective of the amount of additional income it
may have generated, because the $16.94 NEDC price per hundredweight has
protected dairy farmer income when the minimum federal marketing order price
has fallen below the NEDC price. This occurred in 35 of the 46 months
between July 1997 and March 2001.

NEDC commission data indicate that the difference between the Compact Class
I price and the USDA minimum Class I price from July 1997 through June 2001,
minus fees for administering the NEDC, totaled about $146 million. This
amount was provided to 4,217 farms supplying the New England market, of
which 1,300 are estimated to be located in New York. According to the NEDC,
between July 1997 and June 2001, the NEDC resulted in annual payments of
between $3, 900 and $14,700, to these

12 The amounts available in the form of payments to farmers exclude fees
collected by the NEDC commission for paying commission staff salaries, legal
costs, and costs for compensating USDA and the states for estimated
increased costs to milk price support and selected nutrition assistance
programs. NEDC Data Indicate a

Positive Impact on Net Farm Income, but USDA Data Are Less Clear

Appendix III: NEDC?s Intraregional Impacts Page 72 GAO- 01- 866 Impacts of
Dairy Compacts

farms, depending on herd size. The average annual payment was $9, 812 per
farm. According to the NEDC commission, this additional income helped
stabilize and enhance farm- level prices for farmers in the six NEDC states
as well as New York, some of whom have historically been part of the New
England milk shed.

Similarly, NEDC data indicate that net farm earnings improved as a result of
the Compact. For example, the NEDC commission estimated that in 1997, the
Compact increased net farm earnings of those supplying the New England milk
shed by about $6,800, from about $11,000 to about $17,800. The NEDC
commission estimated that in 2000, the Compact increased net earnings by
about $15,200, from about $8,100 to about $23,300. The NEDC commission also
estimated that in 2000, the percentage of dairy farms that experienced
financial stress was 20 percent lower than it would have been in the absence
of the Compact: About 50 percent of farms experienced some degree of
financial stress, compared with 70 percent that would have experienced some
degree of financial stress without the NEDC. These figures led the NEDC
commission to conclude that the overall reduction in financial stress
resulted in a significant reduction in the likely net loss of dairy
operations in the Northeast.

While USDA data indicate that the net farm income of an average dairy farmer
in the northeastern region has increased slightly more than the net income
of an average farmer in other regions between 1991 and 1999, the data also
indicate that net farm income in the northeastern region was lower than that
of farmers in other regions during the same time period. 13 , 14 Figure 4
shows that the net farm income of average dairy farmers in the northeastern
region and other regions in the country has increased between 1991 and 1999,
but that the trend increased at a lower rate in the northeastern region than
in other regions in the country. Regarding the northeastern region, the data
indicate that an average dairy farmer?s net farm income increased by 36
percent, from $34,064 in 1996 to $46,415 in 1999. In the rest of the United
States, an average dairy farmer?s net farm income increased by 61 percent,
from $45,650 in 1996 to $73,486 in 1999. It

13 USDA?s Economic Research Service defines net farm income as total accrual
receipts minus total accrual expenses when cash income is adjusted for
changes in crop and livestock inventories.

14 The northeastern region includes Connecticut, Delaware, Maine, Maryland,
Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode
Island, and Vermont.

Appendix III: NEDC?s Intraregional Impacts Page 73 GAO- 01- 866 Impacts of
Dairy Compacts

is difficult to determine, however, the specific impact of the NEDC, because
many factors influence dairy farm income, and USDA?s northeastern region
includes other states in addition to the six NEDC states.

Figure 4: Net Farm Income of a Representative Dairy Farmer in the
Northeastern Region and in Other U. S. Regions, 1991- 99

Notes: The northeastern region includes Connecticut, Delaware, Maine,
Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania,
Rhode Island, and Vermont. The other U. S. regions exclude these 11 states,
Alaska, and Hawaii.

Net farm income is in 1999 constant dollars. Source: GAO analysis of data
from USDA?s Economic Research Service.

USDA data also indicate that between 1991 and 1999, the percentage of
northeastern dairy farms having favorable solvency 15 grew more than the
percentage of dairy farms with favorable solvency in other regions of the

15 Dairy farms having a favorable solvency have debt- to- asset ratios of
less than 0.40 and positive net farm incomes.

Appendix III: NEDC?s Intraregional Impacts Page 74 GAO- 01- 866 Impacts of
Dairy Compacts

country. 16 Figure 5 shows that the percentage of dairy farms having
favorable solvency in other regions of the country remained relatively
constant.

Figure 5: Comparison of the Percentage of Dairy Farms Having Favorable
Solvency in the Northeastern Region With the Percentage in Other Regions,
1991- 99

Note: The northeastern region includes Connecticut, Delaware, Maine,
Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania,
Rhode Island, and Vermont. The other U. S. regions exclude these 11 states,
Alaska, and Hawaii.

Source: GAO analysis of data from USDA?s Economic Research Service.

About 61 percent of the dairy farms in the northeastern region had favorable
solvency in 1991, in comparison with about 69 percent of the dairy farms in
the rest of the U. S. regions. About 82 percent of the dairy farms in the
northeastern region had favorable solvency in 1999, in

16 USDA?s Economic Research Service uses favorable solvency as an indicator
of the financial position of a farm business based on a combination of
income (net farm income) and solvency (debt/ asset ratio) measures.

Appendix III: NEDC?s Intraregional Impacts Page 75 GAO- 01- 866 Impacts of
Dairy Compacts

comparison with about 73 percent of the dairy farms in the rest of the U. S.
regions. Whether the NEDC caused the percentage of dairy farms in the
northeastern region having a favorable solvency to grow faster is difficult
to determine because, as noted previously, the six NEDC states form only a
portion of the northeastern region.

Two studies have analyzed the NEDC?s impact on dairy farmer income and
concluded that the Compact has the potential to improve farmer finances.
OMB?s 1998 analysis was limited by the fact that the Compact had been in
effect for only about 6 months when OMB conducted its study. 17 To estimate
the NEDC?s impact on farm income, OMB developed two alternative scenarios of
what milk prices would have been had the Compact not been established. Under
the first scenario, OMB assumed that the Class I price would have averaged
$15.92 per hundredweight in 1997; under the second scenario, OMB assumed
that the Class I price would have been $16.10 per hundredweight, taking into
account the decline in the market- driven over- order premium when the NEDC
took effect. OMB then compared these prices with a minimum NEDC price of
$16.94 per hundredweight. Using $16.94 as a basis for calculating a blend
price, OMB estimated that in 1997, the NEDC generated an average increase in
gross farm income of $5,650 under the first scenario and $4,770 under the
second scenario. Regarding the NEDC's overall impact on dairy farmers'
income, OMB concluded that, if other factors affecting dairy farmers were
held constant, higher milk prices would not be expected to greatly alter the
long- term trend toward fewer, but larger and more efficient dairy
operations in New England.

In 1998, an economist (Wackernagel) at the University of Vermont issued a
study on the potential impact of the dairy Compact that used computer models
to simulate characteristics of Vermont dairy farms under different milk
pricing policies. 18 The models varied with respect to factors such as farm
size, farm profitability, productivity growth rate, and milk prices.
Although no one model specifically simulated the impact of the NEDC,
Wackernagel concluded that using his compact scenarios, the impact of

17 OMB, Feb. 1998. 18 Rick W. Wackernagel, ?Potential Economic Impacts of
the Northeast Interstate Dairy Compact on Vermont Dairy Farms,? Agricultural
and Resource Economics Review, Apr. 1998. Two Studies Provide

Limited Information on the NEDC?s Impact on Income

Appendix III: NEDC?s Intraregional Impacts Page 76 GAO- 01- 866 Impacts of
Dairy Compacts

stabilizing prices could increase farmer cash reserves and net worth, but he
estimated that such gains would be more limited than those that would be
achieved by having a higher farm- level price that varied from month to
month. Wackernagel also concluded that gains associated with price
stabilization would be limited unless a Compact?s policies recognized the
impact of inflation as well as the variability and level of milk prices.

Data on the number of farms and herd size show similar trends in the NEDC
and other states before and after the NEDC was established, suggesting
little or no change in the NEDC states due to the Compact. Moreover, the
data suggest that the trends are caused by other factors, such as major
technological advancements in dairy farming.

The number of dairy farms in the NEDC and the rest of the United States has
been decreasing, as measured by both the number of farms having at least one
dairy cow and the number of licensed dairies. In the NEDC states, the number
of farms having at least one milk cow decreased by 33 percent between 1992
and 2000, from 5,050 to 3,370. In the rest of the United States, the number
of farms having at least one milk cow decreased by 39 percent in that same
period, from 166,510 to 101,880. Given that some of the farms with at least
one milk cow may not produce milk for sale to dairy processors, we also
looked at the number of farms licensed to sell milk to dairy processors.
Between 1992 and 2000, the number of licensed dairy farms in the six NEDC
states decreased by 32 percent, or from 4,079 to 2,772, while the number of
licensed dairy farms in the rest of the country decreased by 37 percent
during that same time period, or from 127,456 to 80,253 (see fig. 6). From
1992 to 1998, the percentage decrease in the number of licensed dairy farms
in the NEDC states was greater each year than in the rest of the United
States; in 1998, the percentage decrease in the rest of the United States
became greater each year. With respect to the change from 1997 to 2000, the
number of licensed dairy farms The NEDC?s Impacts

on Farm Structure, Milk Production, and Milk Consumption Are Difficult to
Determine

The Number of Farms Has Continued to Decline as Herd Size Has Increased

Appendix III: NEDC?s Intraregional Impacts Page 77 GAO- 01- 866 Impacts of
Dairy Compacts

decreased by 14 percent in the NEDC states, or from 3,237 to 2,772 farms;
while the number of licensed dairy farms decreased by 17 percent in the rest
of the country, or from 96, 176 to 80,253 farms. It is not known whether
this change in trends was caused by the NEDC or some other factors.

Figure 6: Decline in the Number of Licensed Dairy Farms in the NEDC States
and in the Rest of the United States, 1992- 2000

Source: GAO analysis of data from the American Farm Bureau Federation.

Between 1992 and 2000, the average herd size in the NEDC states increased by
36 percent, from 58 to 79 milk cows, while in the rest of the United States,
average herd size increased by 57 percent, from 56 to 88 milk cows. As shown
in figure 7, although average herd size has increased in the NEDC states,
this increase consistently lagged behind the increases in the rest of the
United States.

Appendix III: NEDC?s Intraregional Impacts Page 78 GAO- 01- 866 Impacts of
Dairy Compacts

Figure 7: Average Number of Cows per Farm in the NEDC States and in the Rest
of the United States, 1992- 2000

Source: GAO analysis of data from USDA?s National Agricultural Statistics
Service.

One of the reasons for the declines in herd size is that over the past 50
years, technological developments have significantly altered both dairy
farming itself and the way farm products are processed and distributed.
Farming has changed from an operation that was historically dependent on
human and animal labor to one in which most operations are mechanized. As a
result, at every level, economies of scale (the lower cost of large- scale
versus small- scale operations) have led to fewer and larger farms. The
number of farms in general, and dairy farms in particular, has been
shrinking since the Depression. While dairy farms with 100 cows were
considered large in 1950, today dairy farms as large as 1,500 to 3, 000 are
emerging in the western, northwestern, and midwestern regions of the
country. This trend, along with the pressure to convert land used for
agriculture into land used for nonagricultural purposes, will likely result
in a continued increase in the size of farms with a commensurate decline in
the total number of farms in the Northeast.

Appendix III: NEDC?s Intraregional Impacts Page 79 GAO- 01- 866 Impacts of
Dairy Compacts

Milk production increased in the six NEDC states and the rest of the country
both before and after the NEDC was established. This trend toward greater
milk production has occurred at the same time as the total number of milk
cows has declined, reflecting a greater amount of milk produced per cow.
These trends make it difficult to determine what impact, if any, the Compact
has had on milk production. Two studies have analyzed the NEDC?s impact on
milk production and concluded that while the NEDC may have caused an
increase in production, any increase was small in relation to the total
amount of milk produced in the NEDC states.

Milk production has increased in the NEDC states and the rest of the
country, as measured by both total milk produced and milk produced per cow.
From 1993 to 2000, total milk produced in the NEDC states increased 2.9
percent, from 4.545 billion pounds to 4.678 billion pounds, while production
in the rest of the country increased 10.3 percent from 146.091 to 163.274
billion pounds. With respect to the change from 1997 to 2000, production
increased by 2.5 percent in the NEDC states, or from about 4.6 billion
pounds to 4.7 billion pounds; while production increased 7.8 percent in the
rest of the country, or from about 151.5 billion pounds to 163.3 billion
pounds. Much of this growth reflected an increase in milk production per
cow. Figure 8 shows that from 1993 to 2000 the average amount of milk
produced per cow in the NEDC states increased by 11.6 percent, from 15,633
pounds to 17,440 pounds, while production per cow in the rest of the United
States increased 15.9 percent, or from 15,726 pounds to 18, 226 pounds.
Production Has Continued

to Increase

Appendix III: NEDC?s Intraregional Impacts Page 80 GAO- 01- 866 Impacts of
Dairy Compacts

Figure 8: Pounds of Milk Produced per Cow in the NEDC States and in the Rest
of the United States, 1993- 2000

Source: GAO analysis of data from USDA?s National Agricultural Statistics
Service.

Whether the increase in milk production per cow was dramatically influenced
by the NEDC, however, is not clear. The data on milk production per cow do
not show a large difference in upward trend between the NEDC states and
states in the rest of the country between 1997 and 2000. Specifically,
during this period production per cow increased by 6.6 percent in the NEDC
states, or from 16,360 pounds to 17,440 pounds; while production per cow
increased by 7.9 percent in the rest of the country, or from 16,887 pounds
to 18,226 pounds. Increases in production per cow, and increased efficiency
in all of the states, has more likely resulted from fundamental changes in
the structure of dairy farming throughout the country, such as the increased
cost of labor; new improved machinery; artificial breeding services; better
feed and forage; adoption of different strains of livestock; and careful use
of fertilizer, irrigation, and chemicals. The overall impact drastically
increased production per cow, which has led to the need for fewer cows to
supply the market.

Appendix III: NEDC?s Intraregional Impacts Page 81 GAO- 01- 866 Impacts of
Dairy Compacts

Both of the studies that analyzed the NEDC?s impact on milk production were
based on a limited amount of data. OMB?s 1998 study found that from July
through December 1997, New England milk production was up 3 percent from the
same period in the previous year, while national milk production was up 2
percent during the same period. 19 Given this increase in New England milk
production and using USDA?s dairy economic model, OMB estimated that the
Compact commission?s price of $16.94 per hundredweight for raw milk used for
and sold as fluid milk would have caused milk marketings to increase from
5.38 billion pounds to 5.40 billion pounds, or by about 0.4 percent ,
between July and December 1998. OMB cautioned, however, that the analysis
addressed only the first 6 months of the NEDC.

At the request of the NEDC, economists (Nicholson et al.) at the University
of Vermont also used a model to estimate the impact of higher Compact prices
on milk production. 20 Their study, issued in 2000, used state quarterly
all- milk prices for each of the six Compact states as a basis to estimate
what milk prices would have been without a Compact. In Vermont, these prices
ranged from $12.85 to $15.09 per hundredweight. On the basis of their
analysis of milk production data for the period July 1997 to June 1998,
Nicholson et al. concluded that the NEDC caused milk production to increase
by 45 million pounds, or about 1 percent. Nicholson et al. estimated that
the Compact resulted in increased production in each of the six NEDC states,
with the largest increase occurring in Vermont. Nicholson et al. postulated
that without the NEDC milk production in that state would have declined.

The impact of the NEDC on milk consumption in the six NEDC states is
difficult to determine because even though the data indicate that per capita
consumption of fluid milk was higher in the NEDC states than in much of the
rest of the United States between 1993 and 1999, consumption slowly declined
during that period, both within the NEDC states and throughout much of the
rest of the country, suggesting little or no change

19 OMB, Feb. 1998. 20 Charles F. Nicholson, Budy P. Resosudarmo, and Rick W.
Wackernagel, ?Impacts of the Compact on New England Milk Supply,? The
Northeast Interstate Dairy Compact: Milk Market Impacts, Research Report 73,
Agricultural Experiment Station, University of Vermont, Mar. 2000. Milk
Consumption Has

Continued to Decrease

Appendix III: NEDC?s Intraregional Impacts Page 82 GAO- 01- 866 Impacts of
Dairy Compacts

as a result of the NEDC. On the other hand, one study that analyzed the
impact of the NEDC on milk consumption concluded that the Compact could have
caused a slight reduction in milk consumption in the NEDC states in the
latter part of 1997.

As figure 9 shows, from 1993 to 1999 annual per capita milk consumption in
the New England Milk Marketing Order declined by about 4 percent, from about
233 pounds to about 223 pounds. 21 This decline is equivalent to a reduction
from about 27 gallons to 26 gallons. Similarly, annual per capita milk
consumption in the rest of the milk marketing orders in the United States
for the same time period declined by about 6 percent, from about 214 pounds
to about 202 pounds. This is equivalent to a reduction from about 25 gallons
to 23 gallons.

21 Because data are not available on milk consumption in the six NEDC
states, we used data for USDA?s New England Milk Marketing Order. The New
England Milk Marketing Order includes Connecticut, Massachusetts, New
Hampshire, Rhode Island, and Vermont, or portions thereof. In January 2000,
this marketing order was dissolved and a larger Northeast Marketing Order,
which incorporates additional states, replaced it.

Appendix III: NEDC?s Intraregional Impacts Page 83 GAO- 01- 866 Impacts of
Dairy Compacts

Figure 9: Per Capita Fluid Milk Consumption in the New England Milk
Marketing Order and in All Other Milk Marketing Orders, 1993- 99

Source: GAO analysis of data from USDA?s National Agricultural Statistics
Service.

This slow decline in milk consumption began in the late 1970s, probably as
the result of such factors as increasing consumption of other beverages,
such as fruit beverages, bottled water, and carbonated soft drinks; growing
dietary concerns about fat and cholesterol; and an aging U. S. population.
USDA estimated that without marketing efforts undertaken by the milk
industry, such as advertising and product innovation, milk consumption would
have been lower than it was by as much as 1.4 percent from 1996 to 1999. 22

OMB?s 1998 study found that from July to December 1997, fluid milk sales in
New England totaled 1.3 billion pounds, which was down 0.7 percent

22 USDA Report to Congress on the Dairy Promotion Programs, 2000.

Appendix III: NEDC?s Intraregional Impacts Page 84 GAO- 01- 866 Impacts of
Dairy Compacts

from the same period 1 year earlier. 23 Nationally, fluid milk sales
increased by 0.2 percent in the last half of calendar year 1997. OMB noted
that studies on the relationship between retail prices and consumption
suggest that a 10- percent increase in retail fluid milk prices reduces
consumption by 1 to 2 percent. Using USDA?s model, OMB estimated that the
average retail price in the Compact states would have been $2.46 per gallon
without the NEDC price regulations, and that with the NEDC price regulations
retail prices increased by 3.7 percent to $2.55 per gallon. Given the retail
price and consumption relationship found in other studies, and the NEDC?s
impact on retail milk prices, OMB concluded that the NEDC?s price
regulations reduced fluid milk consumption in the NEDC states by about 10
million pounds, or about 0.5 percent, between July and December 1997.

Bailey?s July 2000 study also analyzed the NEDC?s impact on retail fluid
milk consumption. 24 Bailey examined Class I sales in the New England
Marketing Milk Marketing Order from 1996 through 1999 and observed a retail
milk price increase of 24 cents per gallon in the Hartford and Boston
markets over that period. He attributed this increase to an average 10-
centper- gallon Compact obligation and a general rise in the farm- to-
retail markup. Even though retail prices increased, Bailey concluded that
total fluid milk consumption did not change appreciably after introduction
of the Compact: He estimated that the amount of milk consumed decreased by
less than 0.3 percent in 1998 and 1999.

23 OMB, Feb. 1998. 24 Kenneth Bailey, July 2000.

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 85 GAO- 01- 866 Impacts of Dairy Compacts

According to USDA, the NEDC has not increased the net federal costs of the
milk price support program, and the agency is not certain whether the
Compact has increased costs of the nutrition assistance programs
administered by USDA. Net costs for USDA's milk price support program have
not increased because the 1996 farm bill requires that the NEDC commission
compensate USDA for any additional dairy commodity purchases to the extent
that the percentage change in milk production in the NEDC states exceeds the
national average. 1 Regarding nutrition assistance programs, according to
USDA, federal costs for its Food Stamp Program may have increased as a
result of the NEDC, but federal costs for its other nutrition assistance
programs have not increased because of how these programs are funded or
because of how federal program benefits are calculated for them. According
to USDA officials, regardless of whether federal costs for its other
nutrition assistance programs have increased, increased retail milk prices
caused by the Compact have been borne by agencies or organizations that
provide program benefits or by program participants in the NEDC states. The
NEDC commission directly reimburses the NEDC states for increased costs
incurred by selected nutrition assistance programs. While two studies
assessed the NEDC?s impacts on selected nutrition assistance programs, study
results are inconclusive as to whether the NEDC has impacted these programs.

Even though USDA estimates that the NEDC states? rates of increase in milk
production have exceeded the national rates of increase in 2 of the 4 years
since the Compact began setting the price of milk used for and sold as fluid
milk in the NEDC states in July 1997, net federal milk price support program
costs have not increased. The price support program, created in 1949,
supports farm- level prices by providing a standing offer from the
government to purchase butter, cheese, and nonfat dry milk at specified
prices. The 1996 farm bill requires that the Compact commission compensate
USDA when the estimated rate of increase in milk production in the six NEDC
states is greater than the estimated rate of increase in national milk
production. The NEDC commission is required to compensate USDA before the
end of any fiscal year in which the NEDC states? estimated rate of increase
in production is greater than the national rate of increase for the
preceding fiscal year. Neither the NEDC nor USDA

1 Federal Agriculture Improvement and Reform Act of 1996 (P. L. 104- 127,
Apr. 4, 1996). Appendix IV: NEDC?s Impacts on Federal

Program Costs Federal Net Costs for the Milk Price Support Program Have Not
Increased

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 86 GAO- 01- 866 Impacts of Dairy Compacts

must determine whether the greater rate of increase, or what portion of the
greater rate of increase, is attributable to the NEDC.

According to USDA officials, milk production in the NEDC states did not
increase at a greater rate than the national rate of increase between July
and September 1997. However, in 1998, USDA estimated that milk production in
the NEDC states increased 1.8 percent, based on its analysis of the average
amount of milk produced in the six states during 1997 and 1998, as compared
to a national average increase in production of 1.3 percent. On the basis of
these two rates, USDA estimated that it purchased about $1.8 million of
nonfat dry milk that was attributable to the NEDC?s greater rate of increase
in production. In 1999, USDA estimated that milk production in the NEDC
states increased 3.6 percent, based on its analysis of the states? average
milk production in 1998 and 1999, compared with a national average rate of
increase of 3.2 percent. On the basis of these two rates, USDA estimated
that it purchased about $1.4 million of nonfat dry milk that was
attributable to the NEDC?s greater rate of increase in production. The NEDC
commission compensated the federal government these amounts for these 2
fiscal years. The Compact commission was not required to compensate USDA for
fiscal year 2000 because USDA concluded that the average rate of increase in
milk production in the NEDC states (0.1 percent) was less than the average
national rate (5.1 percent).

USDA analyzed the Compact?s impact on federal nutrition assistance program
costs and concluded that it is uncertain whether the federal costs of one of
its major programs- the Food Stamp Program- have increased, while federal
costs of its other programs most likely have not. Food Stamp Program costs
may have increased to the extent that the NEDC caused retail milk prices to
increase nationally and to the extent that the agencies or organizations
that provide or receive program benefits are not reimbursed. USDA?s Food and
Nutrition Service is responsible for nutrition assistance programs, which
include the Food Stamp Program; the Special Supplemental Nutrition Program
for Women, Infants and Children (WIC); school programs such as the National
School Breakfast and Lunch Programs and the Special Milk Program; the Child
and Adult Care Food Program; and several small food distribution programs
for Indian reservations, the elderly, pregnant women, and children. These
programs are carried out at the state and local level by state and local
agencies and organizations, such as day care centers or schools. It Is
Unclear Whether

Federal Net Costs for Nutrition Assistance Programs Have Increased

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 87 GAO- 01- 866 Impacts of Dairy Compacts

Regarding the Food Stamp Program, USDA officials said that federal costs may
have increased as a result of the NEDC, but they do not know because it is
difficult to determine (1) how much of the increased retail milk price in
the NEDC states was caused by the Compact, and (2) to what extent that
portion of the increased retail milk price caused an increase in the
national average cost of food items used to determine program benefits. The
Food Stamp Program is operated by state and local welfare offices. Under the
program, food stamp recipients spend their benefits (in the form of paper
coupons or electronic benefits on debit cards), to buy eligible food in
authorized retail food stores. Program benefits under this program are
indexed to the cost of a selected group of foods, which is sensitive to the
changes in the national average retail price of milk. 2 Benefit levels have
increased since July 1997, because of, among other things, increased
national average retail milk prices. USDA officials said that the retail
milk price increases in the six NEDC states caused by the Compact had the
potential to increase the index used to set benefit levels and, thus,
federal costs, but it is not certain whether they did so. Further, it is
difficult to determine what portion of the retail milk price increases was
attributable to the Compact. USDA estimates that if the NEDC caused a 16-
cent per gallon increase in the national average retail milk price, there is
a 50 percent chance that Food Stamp Program benefits increased by $1 per
program participant per month and annual federal Food Stamp Program costs
increased by about $60 million. USDA officials also noted that even if the
increased price of milk attributable to the NEDC was enough to increase
national food stamp benefit levels, some portion of the increased retail
milk costs have been borne by participants located in the six NEDC states.

Regarding WIC, USDA officials said that the Compact has not increased
federal costs to the program because of the discretionary nature of its
funding. Even so, because approximately 30 percent of the funds spent on WIC
foods are used to buy fluid milk, any increase in retail milk prices can
significantly affect the food package cost per participant and these higher
costs, unless offset, would reduce the potential number of participants. WIC
is administered by state agencies, most of whom provide WIC participants
checks or food instruments to purchase specific foods each month at
authorized retailers. While the 1996 farm bill does not require the Compact
commission to reimburse the WIC program, NEDC regulations

2 This selected group of foods is the core of the Thrifty Food Plan- a
nutritious, low- cost model diet plan. Milk makes up about 10 percent of the
value of this plan.

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 88 GAO- 01- 866 Impacts of Dairy Compacts

specify that the Compact commission is to compensate New England state
agencies that administer WIC for any increased costs due to the Compact. The
amount of compensation equals the over- order producer price payment
multiplied by the volume of milk used by program participants in a given
month. 3 For example, in January of 1998 when the Vermont WIC program
purchased 52, 403 gallons of milk, the difference between the NEDC Class I
milk price and the minimum USDA Class I price was 74 cents per
hundredweight. Given that a hundredweight is equivalent to about 11.6
gallons of milk, the Compact commission compensated Vermont about $3,343 for
that month. In total, the Compact commission compensated the six states? WIC
programs $3.8 million from 1997 through 2000 for increased milk costs due to
the NEDC.

According to state WIC officials, the programs used to administer WIC are
being held harmless by the NEDC. These officials said that the NEDC has not
resulted in increased WIC program costs or reduced program participation.
One state WIC official said that the NEDC has done everything it can to
ensure that the WIC programs in the six states remain unharmed by
reimbursing the states the full difference between the NEDC?s minimum Class
I producer price and the milk marketing order minimum Class I price on all
WIC milk purchases.

Regarding school programs, according to USDA officials, it is likely that
federal costs to these programs have not increased as a result of the NEDC
because program benefits are based on a broad index of food prices that is
relatively insensitive to changes in the price of milk in the six NEDC
states. 4 Thus, any increases in milk prices caused by the Compact would
either have to be absorbed by the schools or passed on to paying students.
Programs such as the School Lunch Program are usually administered by state
education agencies, which operate the program through agreements with local
school districts. NEDC regulations specify that the Compact will reimburse
schools for any Compact- related increased costs of fluid milk

3 The memorandum of agreement between the NEDC commission and the state WIC
agencies specifies that reimbursement shall be an amount equal to the over-
order obligation for each month per hundredweight multiplied by the actual
number of hundredweight equivalents of milk provided to WIC participants in
each month.

4 Under the school lunch and breakfast programs, meal reimbursements are
indexed to the consumer price index for food- away- from- home for urban
consumers. Under the Special Milk Program, reimbursements are indexed to the
producer price index for milk.

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 89 GAO- 01- 866 Impacts of Dairy Compacts

sold in 8- ounce containers by schools in the six NEDC states. 5 This
commitment applies to all milk served in 8- ounce containers by schools,
including milk provided under such child nutrition programs as the school
lunch and breakfast programs and the Special Milk Program. The NEDC
commission requires that school food authorities submit claim forms at the
end of each school year that identify the number of 8- oz. cartons of milk
purchased during the school year on these forms. School authorities document
whether part of the price for the 8- ounce milk containers is attributable
to the NEDC over- order premium and, if so, how much on the basis of milk
vendor submissions. Our review of school food authority claim data indicates
that the portion of the contract price that milk vendors have attributed to
the NEDC varies. For example, Connecticut school food authorities attributed
from .2 cents to 1. 37 cents per 8- ounce container to the NEDC over- order
premium during the 1999- 2000 school year.

The NEDC commission then verifies that amounts claimed do not exceed the
NEDC?s average over- order premium for the school year and compensates
school food authorities either the average over- order premium or the amount
vendors attributed to the NEDC over- order premium, whichever is less. Thus,
the amounts paid to school food authorities vary depending on the amount
that they attributed to the NEDC and the amount of milk purchased. For
example, one school food authority in Massachusetts was compensated $308.35
for 146,835 cartons of milk purchased during the 1999- 2000 school year,
while another school food authority in that state was compensated $316.33
for 31,633 cartons of milk purchased for the same year. In total, the NEDC
commission reimbursed the states $662,606 for the 1998- 1999 and 1999- 2000
school years.

Although the amounts paid varied from school to school, officials in the six
states? departments of education generally said that schools claiming and
receiving compensation were not, in the end, spending additional funds for
milk or having to charge higher prices for milk sold to students. However,
these officials also said that many school food authorities have chosen not
to seek compensation because some school food authorities view the claim
process as burdensome and not worth the effort given the relatively small
amounts of money that they would receive.

5 The regulations cover fluid milk sold in 8- ounce containers distributed
by handlers under open competitive bid contracts and sold by schools.

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 90 GAO- 01- 866 Impacts of Dairy Compacts

As is the case with the school programs, according to USDA officials it is
likely that federal costs of the Child and Adult Care Food Program also have
not increased as a result of the NEDC because program benefits are based on
a broad index of retail food prices that is relatively insensitive to
changes in the price of milk. 6 State education or health departments
administer the program, with independent centers and sponsoring
organizations entering into agreements with these departments to operate the
program. Under the program, USDA provides eligible centers and sponsoring
organizations, such as family day care homes and child- care centers,
reimbursements for meals served. According to USDA, it has no data on the
amount of milk purchased under the program, and it would be prohibitively
labor- intensive for the NEDC to establish a method for compensating
thousands of individual homes and centers for any increased retail milk
prices.

Regarding USDA?s food distribution programs, according to USDA officials,
USDA has not estimated the potential NEDC- related increased costs to these
programs. USDA officials said, however, that any increased costs would be
relatively small, given the small size of these programs compared with
programs such as WIC. Furthermore, because most of these programs are not
entitlement programs and thus federal funding is not mandatory, any
increased costs due to the NEDC would have to be borne by program providers
and could result in fewer participants being served.

According to USDA, if the NEDC is expanded to include additional states or
if a southern compact is also created, it is more likely that federal costs
for the Food Stamp Program would increase than with the existing NEDC: The
likely increase in retail milk prices in more states would have a more
direct impact on the index used to set program benefits.

USDA estimated that if retail milk prices in the states of an expanded NEDC
and a southern compact increased by about 20 cents per gallon- an amount
that USDA noted is possible given the NEDC experience- food stamp
participants in compact states would spend about $93 million a year more to
purchase milk. If this price increase did not cause the national

6 Under the Child and Adult Care Food Program, meal reimbursements are
indexed to the cost of the consumer price index for food- away- from- home
for urban consumers. If Compacts Are

Expanded, Federal Food Stamp Program Costs Could Increase, as Could
Nonfederal Costs for Other Nutrition Assistance Programs

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 91 GAO- 01- 866 Impacts of Dairy Compacts

price increase to rise sufficiently to increase program benefits, Food Stamp
Program participants would have to absorb this cost. However, if the 20-
cent- per- gallon price increase resulted in a sufficiently large national
average retail milk price increase to cause a $1- to $2- perparticipant
increase in Food Stamp Program benefits, USDA estimated that federal Food
Stamp Program costs could increase by as much as $60 to $120 million per
year- an amount that would have to be federally funded. 7 Moreover, if the
NEDC is expanded and a southern compact is established, and if the NEDC does
not provide reimbursements, increased retail milk costs could result in
fewer participants being served by state WIC programs. Given these
assumptions, it is also likely that costs to school programs would have to
be absorbed by school food authorities and program participants because of
the index used to establish benefits under these programs. Table 23
summarizes the additional costs that USDA estimated could be incurred under
the above assumptions in fiscal year 2000 by food stamp, WIC, and school
program providers or participants in compact states if the NEDC is expanded
and a southern compact is established, and if retail milk prices increase in
the compact states. 8

7 USDA used 2000 as a base year for developing its estimate of increased
program costs. 8 For purposes of its analysis, USDA included Connecticut,
Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New
York, Ohio, Pennsylvania, Rhode Island, and Vermont in an expanded NEDC, and
Alabama, Arkansas, Florida, Georgia, Kansas, Kentucky, Louisiana, Missouri,
Mississippi, North Carolina, South Carolina, Tennessee, Oklahoma, Texas,
Virginia, and West Virginia in a southern compact.

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 92 GAO- 01- 866 Impacts of Dairy Compacts

Table 23: Estimated Impact on Food Stamp, WIC, and School Feeding Program
Participants if Retail Milk Prices Increase 10, 15, or 20 Cents per Gallon
Under an Expanded NEDC and a Southern Compact

Increase in the retail price of a gallon of milk in an expanded NEDC and
southern compact Annual impact on the program 10 cents 15 cents 20 cents
Increased Food Stamp Program costs a $46,676,933 $70,015,399 $93,353,866
Increased cost to maintain level of WIC participation that existed before
the NEDC expanded (number of participants that would leave WIC if funding
remained constant) 15,586,611

(42,498) 23,379,916 (63,429) 31,173,222

(84,153) Increased cost to National School Breakfast and Lunch Programs
20,315,004 30,472,505 41,630,007

Total increased cost $82,578,547 $123,867,821 $165,157,094

a Increased costs are borne by Food Stamp Program participants if federal
benefits do not increase. If federal benefits increase, increased costs
would be borne by participants in the first year of the compact until
federal program benefits are reindexed in the year following the compact?s
establishment. According to USDA officials, even if federal benefits
increase, the increase in program benefits may not be sufficient to fully
offset increased retail milk prices in compact states. Further, households
in noncompact states would receive higher benefits, even though retail milk
price increases would be limited to compact states.

Source: USDA?s Food and Nutrition Service.

Two studies- one prepared by the Office of Management and Budget (OMB) and
the other by University of Vermont researchers- offer inconclusive results
on the NEDC?s potential impacts on USDA?s nutrition assistance programs.
Both studies were conducted early in the Compact?s existence and relied on
limited retail milk price data. For example, OMB?s 1998 study began before
the NEDC commission and the states entered into agreements for compensating
states? WIC and school programs and relied on only 6 months of retail price
data- from July through December 1997. 9 Because of this, the study
projected either a low (5- to 10- cent), medium (10- to 15- cent), or high
(15- to 20- cent) impact on retail milk prices, as reflected in an increased
price for a gallon of milk. Assuming no NEDC reimbursements and a medium
impact on retail milk prices, OMB estimated that a 15- cent increase in the
retail price of milk in the first 6 months of the NEDC would increase state
WIC program costs by about $721,300, which would require a reduction in
program participation of about 3,000 people if the states did not spend the
additional money. OMB also estimated that school lunch and breakfast program
costs would increase by $1.2 million during the first 6 months of the NEDC-
an increase that would have to be absorbed by schools or passed on to

9 OMB?s The Economic Effects of the Northeast Interstate Dairy Compact, Feb.
1998. Two Studies of the

NEDC?s Impact on Nutrition Assistance Programs Provide Inconclusive Results

Appendix IV: NEDC?s Impacts on Federal Program Costs

Page 93 GAO- 01- 866 Impacts of Dairy Compacts

families who pay for meals and snacks. For the same period, OMB also
estimated that participants in USDA?s Food Stamp Program who reside in the
NEDC states would pay an additional estimated $2.4 million because of
increased retail milk prices- an amount that the federal government would
not be required to pay because it is likely that national average milk
prices would not have increased sufficiently to warrant an increase in
benefits. Even if the price increase were large enough to trigger an
increase in the index used to establish program benefits, only a small
portion of the additional program benefits would go to food stamp recipients
in the NEDC states, because all recipients nationwide would receive the
increase.

Researchers at the University of Vermont, who were asked by the NEDC
commission to conduct the study, also relied on limited data. 10 Wang et al.
focused on the NEDC?s potential impact on the WIC program and analyzed
retail milk price and program participation data for the period between June
1997 and February 1998. They accounted for NEDC reimbursements to the
states? WIC programs and examined retail milk prices in Boston and Hartford.
Their analysis concluded that WIC program participation had not been
significantly affected by the NEDC during the time frame analyzed. The study
also concluded that retail milk prices in Hartford increased significantly
more than in Boston, an increase that might be explained by differences in
market concentration and competition. However, the authors concluded that
their study results had two principal limitations: (1) their analysis was
limited to Boston and Hartford and (2) the NEDC had been in effect only
since July 1997, thus providing a small amount of data for the analysis.

10 Qingbin Wang, Zooyob Anne, Catherine Halbrendt, Charles Nicholson, and
Jaimie Sung,

?Impacts of the Compact on the WIC Program: Evidence From Boston and
Hartford,? The Northeast Interstate Dairy Compact: Milk Market Impacts,
Research Report 73, Agricultural Experiment Station, University of Vermont,
Mar. 2000.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 94 GAO- 01- 866 Impacts of Dairy Compacts

This appendix provides our 1999 estimates of the interregional farm- and
wholesale- level impacts of the NEDC, an expanded NEDC, and an expanded NEDC
combined with a southern compact on various dairy sector indicators. To
develop our estimates, we first estimated the impact of each compact
scenario on the basis of certain assumptions, such as transportation costs
and supply and demand elasticities. We then varied these assumptions to test
the sensitivity of our initial estimates. (See app. II for a detailed
discussion of our methodology; a description of the IRCM; a list of the
states included in the different compact scenarios; values for parameters
used, such as the responsiveness of consumers to changing commodity prices;
and a summary of the data used and sources for these data.)

We present the data in a series of tables that summarize (1) the range of
estimates that we obtained using our initial and subsequent sets of
assumptions across the various compact scenarios for each of the dairy
sector indicators that we analyzed; (2) our initial estimates of farm- level
and wholesale- level impacts of the NEDC scenario, and the results of our
sensitivity analyses for that scenario; (3) our initial estimates of farm-
level and wholesale- level impacts of the expanded NEDC scenario, and the
results of our sensitivity analyses for that scenario; (4) our initial
estimates of farm- level and wholesale- level impacts of the expanded NEDC
scenario combined with a southern compact, and the results of our
sensitivity analyses for that scenario; and (5) the results of our
sensitivity analysis for the expanded NEDC scenario combined with a southern
compact using a more restrictive trade assumption. In all instances, we
present the estimated impacts of the various compact scenarios as changes to
our nocompact baseline values for 1999. Appendix V: Interregional Impacts of
Three

Compact Alternatives in 1999

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 95 GAO- 01- 866 Impacts of Dairy Compacts

Table 24: Range of Estimated Impacts on 1999 Farm- Level Prices Across
Different Compact Scenarios Dollars per hundredweight

Farm- level prices IRCM99 region No- compact

baseline in 1999 Change to

baseline with the NEDC

Change to baseline with

an expanded NEDC

Change to baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Northeast $15.07 $0.07 to 0. 12 $0.30 to 0. 48 $0.22 to 0. 41 $0.26 to 0. 45
Appalachia 15.97 (0.01) to 0 (0.05) to (0.02) 1.10 to 1.85 1.16 to 1.96
Florida 17.14 (0.02) to 0 (0.05) to (0.02) (0.02) to 0.04 (0.07) to 0
Southeast 16.01 (0.02) to 0.02 (0.05) to 0.02 0.79 to 1.66 2.99 to 4.51
Mideast 14.64 (0.02) to 0 (0.05) to (0.03) (0.24) to (0.13) (0.28) to (0.21)
Upper Midwest 13.82 (0.02) to (0.01) (0.05) to (0.03) (0.11) to (0.06)
(0.30) to (0.24) Central 14.22 (0.01) to 0 (0.05) to (0.02) 0.10 to 0.27
0.15 to 0.33 Southwest 14.49 ( 0.01) to 0 (0.04) to (0.02) (0.17) to (0.05)
(0.16) to (0.05) Western 13.18 (0.01) to 0 (0.04) to (0.02) (0.14) to (0.09)
(0.19) to (0.12) Northwest 14.85 (0.01) to 0 (0.04) to (0.03) (0.13) to 0.03
(0.17) to (0.10) California 12.98 (0.02) to (0.01) (0.06) to 0.02 (0.18) to
0.18 (0.27) to (0.16) Arizona 13.76 (0.01) to 0 (0.04) to 0.03 (0.36) to
0.06 (0.09) to 0.06

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Table 25: Range of Estimated Impacts on 1999 Farm- Level Production Across
Different Compact Scenarios Pounds in millions

Farm- level production

IRCM99 region

No- compact baseline in 1999

Change to baseline with

the NEDC Change to

baseline with an expanded

NEDC Change to

baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Northeast 29,138 36 to 90 150 to 383 106 to 313 127 to 255 Appalachia 6, 437
(6) to (1) (19) to (7) 283 to 504 300 to 533 Florida 2, 390 (1) to 0 (4) to
(2) (1) to 9 (4) to 0 Southeast 3, 548 (3) to 0 (7) to 2 70 to 164 261 to
534

Range of Estimates Obtained Across Different Compact Scenarios

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 96 GAO- 01- 866 Impacts of Dairy Compacts

Pounds in millions Farm- level production

IRCM99 region

No- compact baseline in 1999

Change to baseline with

the NEDC Change to

baseline with an expanded

NEDC Change to

baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Mideast 12,238 (7) to 0 (16) to (7) (100) to (35) (101) to (56) Upper
Midwest 32,803 (20) to (4) (54) to (20) (152) to (43) (326) to (180) Central
14,855 (5) to 0 (22) to (12) 45 to 121 103 to 186 Southwest 10,278 (3) to 0
(15) to (5) (45) to (13) (46) to (15) Western 8, 854 (3) to (1) (20 ) to (7)
(64) to (28) (83) to (35) Northwest 7, 134 (2) to 0 (12) to (5) (42) to 5
(51) to (23) California 30,459 (35) to (7) (112) to 21 (400) to 185 (491) to
(168) Arizona 2, 914 (1) to 0 (3) to 3 (29) to 6 (7) to 7

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Table 26: Range of Estimated Impacts on 1999 Farm- Level Revenue Across
Different Compact Scenarios Dollars in millions

Farm- level revenue IRCM99 region No- compact

baseline in 1999 Change to

baseline with the NEDC

Change to baseline with

an expanded NEDC

Change to baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Northeast $4,391 $26 to 46 $110 to 193 $80 to 159 $95 to 165 Appalachia 1,
028 (2) to 0 (6) to (2) 119 to 204 126 to 217 Florida 410 (1) to 0 (2) to
(1) (1) to 5 (2) to 0 Southeast 568 (1) to 1 (3) to 1 40 to 85 165 to 242
Mideast 1, 792 (2) to 0 (8) to (5) (43) to (21) (45) to (34) Upper Midwest
4, 533 (9) to (4) (24) to (13) (63) to (26) (133) to (103) Central 2, 112
(2) to 0 (10) to (5) 21 to 58 37 to 67 Southwest 1, 489 (1) to 0 (6) to (3)
(24) to (7) (23) to (7) Western 1, 167 (1) to 0 (6) to (3) (20) to (13) (26)
to (15) Northwest 1, 059 (1) to 0 (4) to (3) (15) to 3 (17) to (11)
California 3, 954 (11) to (4) (33) to 9 (118) to 80 (145) to (70) Arizona
401 0 (2) to 1 (14) to 2 (4) to 3

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 97 GAO- 01- 866 Impacts of Dairy Compacts

Table 27: Range of Estimated Impacts on 1999 National Average Wholesale-
Level Prices Across Different Compact Scenarios Dollars per hundredweight

Wholesale- level prices Dairy commodity No- compact

baseline in 1999 Change to

baseline with the NEDC

Change to baseline with

an expanded NEDC

Change to baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Fluid milk $15.04 $0.04 to 0. 07 $0.16 to 0. 27 $0.39 to 0. 63 $0.66 to 0.
93 Soft dairy products 27.31 (0.10) to (0.02) (0.19) to (0.08) (0.55) to
(0.17) (0.52) to (0.18) American cheese 139.24 (0.09) to (0.03) (0.41) to
(0.18) (1.41) to (0.62) (1.86) to (1.27) Italian cheese 93.24 (0.08) to
(0.03) (0.31) to (0.13) (1.11) to 0.36 (1.36) to (0.95) Other cheese 115.58
(0.21) to 0.06 (0.44) to (0.05) (1.11) to 0.69 (1.65) to (0.75) Butter
119.52 (0.23) to (0.03) (0.88) to (0.46) (6.53) to 0.21 (2.80) to 0.07
Frozen dairy products 21.40 (0.02) to 0 (0.11) to (0.05) (0.53) to (0.10)
(0.65) to (0.38) Other manufactured products 27.68 (0.03) to 0.04 (0.01) to
0.07 (0.11) to 1.04 (0.18) to 0.08 Nonfat dry milk 102.63 (0.01) to 0 (0.06)
to (0.02) (1.22) to (0.07) (0.05) to 0.01

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Table 28: Range of Estimated Impacts on 1999 Wholesale- Level Production
Across Different Compact Scenarios Pounds in millions

Wholesale- level production Dairy commodity No- compact

baseline in 1999 Change to

baseline with the NEDC

Change to baseline with

an expanded NEDC

Change to baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Fluid Milk 55,043 (69) to (22) (293) to (86) (647) to (233) (778) to (313)
Soft dairy products 5,984 1 to 4 8 to 20 27 to 51 29 to 50 American cheese
3,539 0 to 1 1 to 6 1 to 20 5 to 25 Italian cheese 3,175 0 to 2 1 to 8 (3)
to 21 8 to 33 Other cheese 598 0 to 1 0 to 1 (1) to 4 3 to 5 Butter 1, 274 0
to 1 1 to 3 (1) to 17 0 to 8 Frozen dairy products 12,131 2 to 4 13 to 20 26
to 98 41 to 79 Other manufactured products 4,270 (2) to 3 (2) to 1 (56) to
15 (3) to 9 Nonfat dry milk 1,396 1 to 4 9 to 23 22 to 98 27 to 54

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 98 GAO- 01- 866 Impacts of Dairy Compacts

Table 29: Range of Estimated Impacts on 1999 Wholesale- Level Expenditures
Across Different Compact Scenarios Dollars in millions

Wholesale- level expenditures

Dairy commodity No- compact baseline in 1999

Change to baseline with

the NEDC Change to

baseline with an expanded

NEDC Change to

baseline with an expanded NEC and a southern

compact Change to baseline

with an expanded NEDC, a southern compact, and a 100- mile import constraint

Fluid Milk $8,220 $18 to 33 $69 to 126 $144 to 293 $259 to 438 Soft dairy
products 1,626 (5) to 0 (9) to (1) (19) to (2) (19) to 3 American cheese
4,823 (3) to (1) (11) to (2) (41) to (15) (55) to (16) Italian cheese 3,028
(2) to (1) (8) to (3) (25) to 9 (33) to (12) Other cheese 792 (1) to 0 (3)
to 1 (4) to 4 (7) to (2) Butter 1, 495 (2) to 1 (7) to (2) (64) to 1 (27) to
1 Frozen dairy products 2,570 (2) to 0 (9) to (3) (43) to (6) (62) to (37)
Other manufactured products 1,199 (1) to 1 0 to 3 (1) to 28 (5) to 3 Nonfat
dry milk 1,012 0 (1) to 0 (1) to 1 0

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

To obtain our initial estimates of the effects of the NEDC in 1999, we used
the following assumptions (our no- compact baseline includes the first three
assumptions):

 No more than 40 percent of any one region?s milk may be shipped to another
region without being subject to the receiving region?s pricing requirements.
This assumption is used to simulate USDA milk marketing order regulations
regarding minimum pricing requirements for milk shipped between marketing
orders. This threshold was chosen as a proxy for the requirement that an
adjustment be made if a plurality of an order's packaged milk was sold in
another region.

 Supply elasticities are medium- term (that is, 5 years).

 Demand elasticities are medium- term.

 The Class I minimum price in the Compact region is the higher of the
Compact price of $16.94 per hundredweight of milk, or the USDA milk
marketing order price.

 Market- driven over- order premiums are zero in the Compact region. This
initial assumption represents a lower bound, and assumes that all
marketdriven over- order premiums in NEDC states are replaced by Compact
over- order producer price payments. The Economic

Impacts of the NEDC in 1999

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 99 GAO- 01- 866 Impacts of Dairy Compacts

 A handler must pay the compact over- order producer price into the Compact
pool for milk shipped into the Compact region in order to receive the
compact price.

We then performed a series of sensitivity analyses by varying key
assumptions to test the ?robustness? of these initial estimates- that is,
whether, and if so by how much, our initial estimates would change when we
used different assumptions. Tables 30 through 39 present our initial and
subsequent estimates of the impacts of the NEDC in 1999 compared to a no-
compact scenario on farm- and wholesale- level indicators.

Table 30: Initial Estimates of the NEDC?s Impacts on 1999 Farm- Level
Indicators Farm- level prices (dollars per

hundredweight) Farm- level production (pounds in millions) Farm- level
revenue (dollars in

millions) IRCM99 region No- compact

baseline in 1999 Change to baseline No- compact

baseline in 1999 Change to baseline No- compact

baseline in 1999 Change to baseline

Northeast $15.07 $0.08 29,138 40 $4,391 $29 Appalachia 15.97 (0.01) 6,437
(2) 1, 028 (1) Florida 17.14 (0.01) 2,390 (1) 410 0 Southeast 16.01 (0.01)
3,548 (1) 568 (1) Mideast 14.64 0 12,238 (1) 1,792 0 Upper Midwest 13.82
(0.01) 32,803 (5) 4,533 (4) Central 14.22 (0.01) 14,855 (3) 2,112 (2)
Southwest 14.49 (0.01) 10,278 (2) 1,489 (1) Western 13.18 (0.01) 8,854 (1)
1, 167 (1) Northwest 14.85 (0.01) 7,134 (1) 1, 059 (1) California 12.98
(0.01) 30,459 (8) 3,954 (4) Arizona 13.76 (0.01) 2,914 0 401 0 All
noncompact regions combined

a a 131,910 (25) 18,513 (15) Note: Numbers in parentheses are negative. a
Because farm- level prices are regional, an ?all noncompact regions
combined? price is not

meaningful. Source: GAO analysis using the IRCM99.

Initial Estimates of the NEDC?s Impacts

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 100 GAO- 01- 866 Impacts of Dairy Compacts

Table 31: Initial Estimates of the NEDC?s Impacts on 1999 Wholesale- Level
Indicators Wholesale- level prices (dollars

per hundredweight) Wholesale- level production (pounds in millions)
Wholesale- level expenditures

(dollars in millions) Dairy commodity No- compact

baseline in 1999 Change to baseline No- compact

baseline in 1999 Change to baseline No- compact

baseline in 1999 Change to baseline

Fluid milk $15.04 $0.05 55,043 (23) $8,220 $24 Soft dairy products 27.31
(0.02) 5,984 2 1, 626 (1) American cheese 139.24 (0.04) 3,539 0 4, 823 (1)
Italian cheese 93.24 (0.04) 3,175 1 3, 028 (1) Other cheese 115.58 (0.09)
598 0 792 (1) Butter 119.52 (0.14) 1,274 0 1, 495 (2) Frozen dairy products
21.40 (0.02) 12,131 3 2,570 (2) Other manufactured products 27.68 0 4, 270 0
1, 199 0 Nonfat dry milk 102.63 0 1, 396 3 1, 012 0

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Tables 32 through 37 display the results of our sensitivity analyses for
1999 farm- and wholesale- level indicators. In comparison with our initial
estimates, we used (1) 10- year regional supply elasticities as opposed to
5year; (2) long- term (i. e., higher), as opposed to medium- term, commodity
demand elasticities, (3) higher market- driven over- order premiums as
opposed to zero; (4) a combination of the previous three assumptions; and
(5) an overall 25- percent increase in transportation costs. 1

1 In tables 32 through 37, as well as in comparable tables showing our
estimates of the impacts of other compact scenarios, we include a column
showing our baseline estimates for our no- compact scenario to provide a
context for assessing the estimated changes due to the compacts under the
assumptions used in each sensitivity analysis. In conducting each of the
separate sensitivity analyses, we developed a separate baseline for our
farmand wholesale- level variables. For presentation purposes, however, we
include only our baseline estimates developed for our initial estimates for
each compact scenario because they were very similar to the baselines for
each sensitivity analysis. Sensitivity Analyses for the

NEDC Scenario

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 101 GAO- 01- 866 Impacts of Dairy Compacts

Table 32: Estimated Change in 1999 Farm- Level Prices Using Different
Assumptions Under the NEDC Scenario Dollars per hundredweight

Estimated change to baseline farm- level prices using different assumptions
IRCM99 region

No- compact Baseline farmlevel price in 1999

10- year regional supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Northeast $15.07 $0.08 $0.08 $0.12 $0.11 $0.07 Appalachia 15.97 0 (0.01)
(0.01) (0.01) 0 Florida 17.14 (0.01) (0.01) (0.01) (0.02) 0 Southeast 16.01
0 (0.01) (0.01) (0.02) 0.02 Mideast 14.64 (0.01) (0.01) (0.01) (0.02) 0
Upper Midwest 13.82 (0.01) (0.01) (0.01) (0.02) (0.01) Central 14.22 0 0
(0.01) 0 0 Southwest 14.49 0 0 (0.01) (0.01) (0.01) Western 13.18 0 (0.01)
(0.01) 0 0 Northwest 14.85 0 (0.01) (0.01) (0.01) (0.01) California 12.98
(0.02) (0.01) (0.01) (0.02) (0.01) Arizona 13.76 0 0 (0.01) 0 (0.01)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO analysis using the IRCM99.

Table 33: Estimated Change in 1999 Farm- Level Production Using Different
Assumptions Under the NEDC Scenario Pounds in millions

Estimated change to baseline farm- level production using different
assumptions IRCM99 region

No- compact baseline farm- level production in 1999

10- year regional supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Northeast 29,138 61 37 60 90 36 Appalachia 6, 437 (3) (2) (2) (6) (1)
Florida 2, 390 0 (1) (1) (1) (1) Southeast 3, 548 0 (1) (2) (3) 2 Mideast
12,238 (3) (3) (3) (7) 0 Upper Midwest 32,803 (9) (4) (7) (20) (6) Central
14,855 (1) (3) (5) 0 (2) Southwest 10,278 (2) 0 (3) (2) (2) Western 8, 854
(3) (3) (3) (2) (1) Northwest 7, 134 0 (2) (2) (2) (1) California 30,459
(33) (7) (11) (35) (11) Arizona 2, 914 (1) (1) (1) 0 0

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 102 GAO- 01- 866 Impacts of Dairy Compacts

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO analysis using the IRCM99.

Table 34: Estimated Change in 1999 Farm- Level Revenue Using Different
Assumptions Under the NEDC Scenario Dollars in millions

Estimated change to baseline farm- level revenue using different assumptions
IRCM99 region

No- compact baseline farm- level

revenue in 1999 10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Northeast $4,391 $33 $29 $44 $46 $26 Appalachia 1, 028 0 (1) (1) (2) 0
Florida 410 0 0 0 (1) 0 Southeast 568 0 (1) (1) (1) 1 Mideast 1, 792 (2) (2)
(2) (3) 0 Upper Midwest 4, 533 (5) (4) (4) (9) (4) Central 2, 112 0 0 (2) 0
0 Southwest 1, 489 0 0 (1) (1) (1) Western 1, 167 0 (1) (1) (0) (0)
Northwest 1, 059 0 (1) (1) (1) (1) California 3, 954 (10) (4) (4) (11) (5)
Arizona 4010 0 0 0 0

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO analysis using the IRCM99.

Table 35: Estimated Change in 1999 National Average Wholesale- Level Prices
Using Different Assumptions Under the NEDC Scenario

Dollars per hundredweight Estimated change to baseline wholesale- level
prices using different assumptions

Dairy commodity

No- compact baseline wholesalelevel price in 1999

10- year regional supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Fluid milk $15.04 $0.04 $0.05 $0.07 $0.07 $0.05 Soft dairy products 27.31
(0.08) (0.02) (0.03) (0.10) (0.02) American cheese 139.24 (0.03) (0.07)
(0.08) (0.09) (0.05) Italian cheese 93.24 (0.03) (0.06) (0.05) (0.08) (0.04)
Other cheese 115.58 (0.21) 0.06 (0.10) (0.12) (0.15)

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 103 GAO- 01- 866 Impacts of Dairy Compacts

Dollars per hundredweight Estimated change to baseline wholesale- level
prices using different assumptions

Dairy commodity

No- compact baseline wholesalelevel price in 1999

10- year regional supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Butter 119.52 (0.12) (0.03) (0.23) (0.06) (0.12) Frozen dairy products 21.40
(0.01) 0 (0.02) (0.01) (0.01) Other manufactured products 27.68 (0.02) 0.04
0.01 (0.03) 0.01

Nonfat dry milk 102.63 0 (0.01) 0 (0.01) (0.01) Note: Numbers in parentheses
are negative. a The combined assumptions include the previous three
assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO analysis using the IRCM99.

Table 36: Estimated Change in 1999 Wholesale- Level Production Using
Different Assumptions Under the NEDC Scenario Pounds in millions

Estimated change to baseline wholesale- level production using different
assumptions Dairy commodity

No- compact baseline wholesalelevel production in

1999 10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Fluid milk 55,043 (23) (46) (36) (69) (22) Soft dairy products 5,984 1 2 3 4
2 American cheese 3,539 0 1 0 1 0 Italian cheese 3,175 0 1 1 2 0 Other
cheese 598 0 1 0 0 0 Butter 1, 274 0 1 1 0 0 Frozen dairy products 12,131 3
2 4 3 2 Other manufactured products 4,270 1 (2) 0 3 (1)

Nonfat dry milk 1,396 1 4 4 3 2 Note: Numbers in parentheses are negative. a
The combined assumptions include the previous three assumptions: (1) 10-
year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO analysis using the IRCM99.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 104 GAO- 01- 866 Impacts of Dairy Compacts

Table 37: Estimated Change in 1999 Wholesale- Level Expenditures Using
Different Assumptions Under the NEDC Scenario Dollars in millions

Estimated change to baseline wholesale- level expenditures using different
assumptions Dairy commodity

No- compact Baseline wholesalelevel expenditures

in 1999 10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Fluid milk $8,220 $18 $20 $33 $28 $24 Soft dairy products 1,626 (4) 0 (1)
(5) (1) American cheese 4,823 (1) (1) (3) (2) (2) Italian cheese 3,028 (1)
(1) (2) (1) (1) Other cheese 792 (1) 0 (1) (1) 0 Butter 1, 495 (2) 1 (2) (1)
(2) Frozen dairy products 2,570 (1) 0 (2) (1) (1) Other manufactured
products 1,199 (1) 1 0 0 0

Nonfat dry milk 1,012 0 0 0 0 0 Note: Numbers in parentheses are negative. a
The combined assumptions include the previous three assumptions: (1) 10-
year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO analysis using the IRCM99.

Tables 38 and 39 summarize the results of our initial estimates and
sensitivity analyses by presenting the range of estimates of the changes
from our no- compact scenario that we obtained from our various analyses.

Table 38: Summary of Estimated Changes in 1999 Farm- Level Indicators Using
Our Initial and Subsequent Sets of Assumptions Under the NEDC Scenario

IRCM99 region Change in farm

level prices (dollars per hundredweight)

Change in farm- level production (pounds

in millions) Change in farm

level revenue (dollars in millions)

Northeast $0.07 to 0. 12 36 to 90 $26 to 46 Appalachia (0.01) to 0 (6) to
(1) (2) to 0 Florida (0.02) to 0 (1) to 0 (1) to 0 Southeast (0.02) to 0.02
(3) to 0 (1) to 1 Mideast (0.02) to 0 (7) to (0) (2) to 0 Upper Midwest
(0.02) to (0.01) (20) to (4) (9) to (4) Central (0.01) to 0 (5) to 0 (2) to
0 Southwest (0.01) to 0 (3) to 0 (1) to 0 Western (0.01) to 0 (3) to (1) (1)
to 0

Summary of the Estimated Impacts of the NEDC

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 105 GAO- 01- 866 Impacts of Dairy Compacts

IRCM99 region Change in farmlevel

prices (dollars per hundredweight)

Change in farm- level production (pounds

in millions) Change in farmlevel

revenue (dollars in millions)

Northwest (0.01) to 0 (2) to 0 (1) to 0 California (0.02) to (0.01) (35) to
(7) (11) to (4) Arizona (0.01) to 0 (1) to 0 0

Note: Numbers in parentheses are negative. Source: GAO analysis using the
IRCM99.

Table 39: Summary of Estimated Changes in 1999 Wholesale- Level Indicators
Using Our Initial and Subsequent Sets of Assumptions Under the NEDC Scenario

Dairy commodity

Change in wholesale- level prices (dollars per

hundredweight) Change in

wholesale- level production (pounds

in millions) Change in

wholesale- level expenditures (dollars in millions)

Fluid milk $0.04 to 0. 07 (69) to (22) $18 to 33 Soft dairy products (0.10)
to (0.02) 1 to 4 (5) to 0 American cheese (0.09) to (0.03) 0 to 1 (3) to (1)
Italian cheese (0.08) to (0.03) 0 to 2 (2) to (1) Other cheese (0.21) to
0.06 0 to 1 (1) to 0 Butter (0.23) to (0.03) 0 to 1 (2) to 1 Frozen dairy
products (0.02) to 0.00 2 to 4 (2) to 0 Other manufactured products (0.03)
to 0.04 (2) to 3 (1) to 1

Nonfat dry milk (0.01) to 0 1 to 4 0 Note: Numbers in parentheses are
negative. Source: GAO analysis using the IRCM99.

To obtain our initial estimates of the impacts of an expanded NEDC in 1999,
we used the same set of assumptions that we used under our nocompact
scenario. We also used the same assumptions that we used in developing our
initial estimates of the impacts of the NEDC. We then performed the same
series of sensitivity analyses as under the NEDC scenario. Tables 40 through
49 present our initial and subsequent estimates of the impacts of an
expanded NEDC in 1999 compared with a no- compact scenario. The Economic

Impacts of an Expanded NEDC in 1999

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 106 GAO- 01- 866 Impacts of Dairy Compacts

Table 40: Initial Estimates of an Expanded NEDC?s Impacts on 1999 Farm-
Level Indicators Farm- level prices (dollars per hundredweight)

Farm- level production (pounds in millions)

Farm- level revenue (dollars in millions)

IRCM99 region No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline No- compact baseline in1999 Change to

baseline

Northeast $15.07 $0.32 29,138 154 $4,391 $117 Appalachia 15.97 (0.04) 6,437
(10) 1,028 (4) Florida 17.14 (0.03) 2,390 (2) 410 (1) Southeast 16.01 (0.03)
3,548 (3) 568 (2) Mideast 14.64 (0.03) 12,238 (9) 1,792 (5) Upper Midwest
13.82 (0.04) 32,803 (27) 4,533 (17) Central 14.22 (0.04) 14,855 (14) 2,112
(8) Southwest 14.49 (0.03) 10,278 (9) 1,489 (4) Western 13.18 (0.03) 8,854
(9) 1, 167 (4) Northwest 14.85 (0.03) 7,134 (6) 1, 059 (3) California 12.98
0 30,459 7 3,954 1 Arizona 13.76 (0.03) 2,914 (2) 401 (1) All noncompact
regions combined a a 131,910 (91) 18,513 (49)

Note: Numbers in parentheses are negative. a Because farm- level prices are
regional, an ?all noncompact regions combined? price is not

meaningful. Source: GAO analysis using the IRCM99.

Table 41: Initial Estimates of an Expanded NEDC?s Impacts on 1999 Wholesale-
Level Indicators Wholesale- level prices (dollars per hundredweight)
Wholesale- level production

(pounds in millions) Wholesale- level expenditures (dollars in millions)

Dairy commodity No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline

Fluid milk $15.04 $0.17 55,043 (88) $8,220 $80 Soft dairy products 27.31
(0.18) 5,984 8 1, 626 (9) American cheese 139.24 (0.32) 3,539 1 4, 823 (10)
Italian cheese 93.24 (0.23) 3,175 2 3, 028 (6) Other cheese 115.58 (0.18)
598 1 792 (1) Butter 119.52 (0.56) 1,274 1 1, 495 (6) Frozen dairy products
21.40 (0.07) 12,131 13 2, 570 (6) Other manufactured products 27.68 0.01
4,270 0 1, 199 0

Initial Estimates of the Expanded NEDC?s Impacts

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 107 GAO- 01- 866 Impacts of Dairy Compacts

Wholesale- level prices (dollars per hundredweight) Wholesale- level
production

(pounds in millions) Wholesale- level expenditures (dollars in millions)

Dairy commodity No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline

Nonfat dry milk 102.63 (0.02) 1,396 0 1, 012 0 Note: Numbers in parentheses
are negative. Source: GAO?s analysis using the IRCM99.

Table 42: Estimated Change in 1999 Farm- Level Prices Using Different
Assumptions Under an Expanded NEDC Scenario Dollars per hundredweight

Estimated change to baseline farm- level prices using different assumptions
IRCM99 region

No- compact baseline farm- level

price in 1999 10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

Transportation costs increase

25 percent

Northeast $15.07 $0.30 $0.30 $0.48 $0.46 $0.32 Appalachia 15.97 (0.04)
(0.03) (0.05) (0.05) (0.02) Florida 17.14 (0.04) (0.03) (0.05) (0.05) (0.02)
Southeast 16.01 (0.04) (0.02) (0.05) (0.04) 0.02 Mideast 14.64 (0.04) (0.03)
(0.05) (0.04) (0.03) Upper Midwest 13.82 (0.04) (0.03) (0.05) (0.05) (0.03)
Central 14.22 (0.03) (0.02) (0.05) (0.03) (0.03) Southwest 14.49 (0.03)
(0.02) (0.04) (0.03) (0.03) Western 13.18 (0.03) (0.03) (0.04) (0.04) (0.02)
Northwest 14.85 (0.03) (0.03) (0.04) (0.03) (0.03) California 12.98 (0.01)
0.02 (0.01) (0.06) (0.02) Arizona 13.76 (0.02) (0.02) (0.04) 0.03 (0.03)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Sensitivity Analyses for the Expanded NEDC Scenario

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 108 GAO- 01- 866 Impacts of Dairy Compacts

Table 43: Estimated Change in 1999 Farm- Level Production Using Different
Assumptions Under an Expanded NEDC Scenario Pounds in millions

Estimated change to baseline farm- level production using different
assumptions IRCM99 region

No- compact baseline farmlevel production

in 1999 10- year regional supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

Transportation costs increase

25 percent Northeast 29,138 248 150 237 383 160 Appalachia 6,437 (17) (8)
(13) (19) (7) Florida 2, 390 (3) (2) (3) (4) (2) Southeast 3,548 (5) (2) (5)
(7) 2 Mideast 12,238 (16) (7) (12) (15) (7) Upper Midwest 32,803 (49) (20)
(37) (54) (21) Central 14,855 (22) (12) (21) (21) (12) Southwest 10,278 (15)
(5) (12) (11) (7) Western 8, 854 (16) (8) (13) (20) (7) Northwest 7,134 (12)
(6) (9) (10) (5) California 30,459 (20) 21 (7) (112) (26) Arizona 2,914 (3)
(2) (3) 3 (2)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM.

Table 44: Estimated Change in 1999 Farm- Level Revenue Using Different
Assumptions Under an Expanded NEDC Scenario Dollars in millions

Estimated change to baseline farm- level revenue using different assumptions
IRCM99 region

No- compact baseline farm- level

revenue in 1999 10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher over order premiums Combined

assumptions a Transportation

costs increase 25 percent

Northeast $4,391 $125 $110 $177 $193 $118 Appalachia 1, 028 (5) (3) (5) (6)
(2) Florida 410 (1) (1) (2) (2) (1) Southeast 568 (2) (1) (3) (3) 1 Mideast
1, 792 (7) (5) (8) (7) (5) Upper Midwest 4, 533 (20) (13) (21) (24) (13)
Central 2, 112 (8) (5) (10) (7) (6) Southwest 1, 489 (5) (3) (6) (5) (4)
Western 1, 167 (5) (4) (5) (6) (3) Northwest 1, 059 (4) (3) (4) (4) (3)
California 3, 954 (6) 9 (4) (33) (10) Arizona 401 (1) (1) (2) 1 (1)

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 109 GAO- 01- 866 Impacts of Dairy Compacts

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 45: Estimated Change in 1999 National Average Wholesale- Level Prices
Using Different Assumptions Under an Expanded NEDC Scenario

Dollars per hundredweight Estimated change to baseline wholesale- level
prices using different assumptions

Dairy commodity No- compact

baseline wholesale- level

price in 1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Fluid milk $15.04 $0.16 $0.18 $0.27 $0.27 $0.19 Soft dairy products 27.31
(0.18) (0.08) (0.15) (0.19) (0.09) American cheese 139.24 (0.30) (0.23)
(0.41) (0.35) (0.18) Italian cheese 93.24 (0.25) (0.19) (0.30) (0.31) (0.13)
Other cheese 115.58 (0.44) (0.05) (0.31) (0.21) (0.44) Butter 119.52 (0.54)
(0.46) (0.88) (0.52) (0.66) Frozen dairy products 21.40 (0.07) (0.05) (0.11)
(0.07) (0.07) Other manufactured products 27.68 (0.01) 0.07 0.04 0.06 0.07
Nonfat dry milk 102.63 (0.02) (0.04) (0.02) (0.06) (0.03)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional
supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 46: Estimated Change in 1999 Wholesale- Level Production Using
Different Assumptions Under an Expanded NEDC Scenario

Pounds in millions Estimated change in baseline wholesale- level production
using different assumptions

Dairy commodity No- compact

baseline wholesale- level

production in 1999

10- year regional supply elasticities

Higher commodity

demand elasticities Higher over order premiums Combined

assumptions a Transportation

costs increase 25 percent

Fluid milk 55,043 (86) (189) (141) (293) (94) Soft dairy products 5,984 8 14
13 20 8 American cheese 3,539 1 4 2 6 1 Italian cheese 3,175 2 5 3 8 1 Other
cheese 598 1 1 1 1 0 Butter 1, 274 1 3 3 2 1

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 110 GAO- 01- 866 Impacts of Dairy Compacts

Pounds in millions Estimated change in baseline wholesale- level production
using different assumptions

Dairy commodity No- compact

baseline wholesale- level

production in 1999

10- year regional supply elasticities

Higher commodity

demand elasticities Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Frozen dairy products 2,131 14 13 20 17 13 Other manufactured products 4,270
1 (2) 0 0 (2) Nonfat dry milk 1,396 9 18 16 23 12

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 47: Estimated Change in 1999 Wholesale- Level Expenditures Using
Different Assumptions Under an Expanded NEDC Scenario

Dollars in millions Estimated change in baseline wholesale- level
expenditures using different assumptions

Dairy commodity No- compact

baseline wholesale- level expenditures in

1999 10- year regional

supply elasticities

Higher commodity

demand elasticities Higher overorder premiums Combined

assumptions a Transportation

costs increase 25 percent

Fluid milk $8,220 $74 $69 $126 $103 $89 Soft dairy products 1,626 (9) (1)
(6) (6) (3) American cheese 4,823 (9) (2) (11) (4) (5) Italian cheese 3,028
(6) (3) (8) (3) (3) Other cheese 792 (3) 1 (1) 0 (2) Butter 1, 495 (6) (2)
(7) (4) (7) Frozen dairy products 2,570 (6) (3) (9) (5) (5) Other
manufactured products 1,199 0 2 1 3 2 Nonfat dry milk 1,012 0 0 0 (1) 0

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional
supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 111 GAO- 01- 866 Impacts of Dairy Compacts

Tables 48 and 49 summarize the results of our initial estimates and
sensitivity analyses by presenting a range of estimates of the changes from
our no- compact scenario that we obtained in our various analyses.

Table 48: Summary of Estimated Changes in Farm- Level Indicators Using Our
Initial and Subsequent Sets of Assumptions Under an Expanded NEDC Scenario

IRCM99 region Change in farmlevel

prices (dollars per hundredweight)

Change in farm- level production (pounds

in millions) Change in farmlevel

revenue (dollars in millions)

Northeast $0.30 to 0. 48 150 to 383 $110 to 193 Appalachia (0.05) to (0.02)
(19) to (7) (6) to (2) Florida (0.05) to (0.02) (4) to (2) (2) to (1)
Southeast (0.05) to 0.02 (7) to 2 (3) to 1 Mideast (0.05) to (0.03) (16) to
(7) (8) to (5) Upper Midwest (0.05) to (0.03) (54) to (20) (24) to (13)
Central (0.05) to (0.02) (22) to (12) (10) to (5) Southwest (0.04) to (0.02)
(15) to (5) (6) to (3) Western (0.04) to (0.02) (20 ) to (7) (6) to (3)
Northwest (0.04) to (0.03) (12) to (5) (4) to (3) California (0.06) to 0.02
(112) to 21 (33) to 9 Arizona (0.04) to 0.03 (3) to 3 (2) to 1

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Table 49: Summary of Estimated Changes in 1999 Wholesale- Level Indicators
Using Our Initial and Subsequent Sets of Assumptions Under an Expanded NEDC
Scenario

Dairy commodity Change in

wholesale- level prices (dollars

per hundredweight)

Change in wholesale- level

production (pounds in

millions) Change in

wholesale- level expenditures (dollars in millions)

Fluid milk $0.16 to 0. 27 (293) to (86) $69 to 126 Soft dairy products
(0.19) to (0.08) 8 to 20 (9) to (1) American cheese (0.41) to (0.18) 1 to 6
(11) to (2) Italian cheese (0.31) to (0.13) 1 to 8 (8) to (3) Other cheese
(0.44) to (0.05) 0 to 1 (3) to 1 Butter (0.88) to (0.46) 1 to 3 (7) to (2)
Frozen dairy products (0.11) to (0.05) 13 to 20 (9) to (3) Other
manufactured products (0.01) to 0.07 (2) to 1 0 to 3 Nonfat dry milk (0.06)
to (0.02) 9 to 23 (1) to 0

Summary of the Estimated Impacts of the Expanded NEDC

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 112 GAO- 01- 866 Impacts of Dairy Compacts

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

As with the other compact scenarios, we developed initial estimates of the
effects of an expanded NEDC in conjunction with a southern compact in 1999
by using a set of key assumptions and conducting subsequent sensitivity
analyses. In addition to conducting the same sensitivity analyses that we
conducted under the previous two compact scenarios, we also varied the
assumption regarding the Class I minimum price, or compact price, in the
compact region. For that analysis, we increased the minimum price from $16.
94 to $18.00 per hundredweight in the southern compact but retained the
$16.94 minimum price in the expanded NEDC. We used this higher minimum
southern compact price because cooperative pay prices in selected cities in
USDA?s Appalachian, Southern, and Central milk marketing orders averaged
about a dollar higher than in the Northeast Marketing Order in 1999. 2 Under
this scenario we conducted an additional analysis in which we assume that
fluid trade movements into compact regions are limited to the amount of milk
that is produced within a 100- mile radius surrounding a compact region. 34
Because this analysis represents a variation of the model, we also conducted
a separate set of sensitivity analyses.

2 Cooperative pay prices in each marketing order are related to the blend
prices for those marketing orders. However, they also reflect the costs of
certain services performed by the cooperative associations as well as other
costs such as membership dues.

3 We modified the model for each of the three compact scenarios to account
for this additional fluid milk trade restriction. The model results under
the NEDC and the expanded NEDC scenarios did not change when this
restriction was added, but results were different for the expanded NEDC in
conjunction with a southern compact scenario. Therefore, we are including
the results of this additional modeling effort only for the expanded NEDC in
conjunction with a southern compact scenario.

4 We performed this analysis because our data are at the regional level as
opposed to the milk plant or dairy farm level. Therefore, our transportation
cost data are average cost data and do not apply to individual shipments. As
a result, our initial results may reflect more movement of milk between
regions than would actually occur. The Economic

Impacts of an Expanded NEDC in Conjunction With a Southern Compact in 1999

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 113 GAO- 01- 866 Impacts of Dairy Compacts

Table 50: Initial Estimates of the Impacts of the Expanded NEDC in
Conjunction With a Southern Compact on 1999 Farm- Level Indicators

Farm- level prices (dollars per hundredweight) Farm- level production

(pounds in millions) Farm- level revenue

(dollars in millions) IRCM99 region No- compact

baseline in 1999 Change to baseline No- compact

baseline in 1999 Change to baseline No- compact

baseline in 1999 Change to baseline

Northeast $15.07 $0.27 29,138 130 $4,391 $99 Appalachia 15.97 1.35 6,437 348
1,028 147 Florida 17.14 0.04 2,390 2 410 1 Southeast 16.01 0.80 3,548 70 568
40 Mideast 14.64 (0.18) 12,238 (49) 1,792 (29) Upper Midwest 13.82 (0.07)
32,803 (52) 4,533 (30) Central 14.22 0.10 14,855 45 2, 112 21 Southwest
14.49 (0.10) 10,278 (27) 1,489 (14) Western 13.18 (0.10) 8,854 (28) 1,167
(13) Northwest 14.85 (0.09) 7,134 (19) 1,059 (9) California 12.98 (0.06)
30,459 (63) 3,954 (26) Arizona 13.76 (0.11) 2,914 (8) 401 (4) All noncompact
regions combined a a 107,070 (246) 14,805 (125)

Note: Numbers in parentheses are negative. a Because farm- level prices are
regional, an ?all noncompact regions combined? price is not

meaningful. Source: GAO analysis using the IRCM99.

Table 51: Initial Estimates of the Impacts of the Expanded NEDC in
Conjunction With a Southern Compact on 1999 WholesaleLevel Indicators

Wholesale- level prices (dollars per hundredweight) Wholesale- level
production

(pounds in millions) Wholesale- level expenditures (dollars in millions)

Dairy commodity No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline

Fluid milk $15.04 $0.47 55,043 (248) $8,220 $218 Soft dairy products 27.31
(0.33) 5,984 32 1,626 (11) American cheese 139.24 (0.93) 3,539 4 4, 823 (27)
Italian cheese 93.24 (0.62) 3,175 6 3, 028 (16) Other cheese 115.58 (0.66)
598 2 792 (2) Butter 119.52 (2.55) 1,274 6 1, 495 (25)

Initial Estimates of the Impacts of the Expanded NEDC in Conjunction With a
Southern Compact

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 114 GAO- 01- 866 Impacts of Dairy Compacts

Wholesale- level prices (dollars per hundredweight) Wholesale- level
production

(pounds in millions) Wholesale- level expenditures (dollars in millions)

Dairy commodity No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline

Frozen dairy products 21.40 (0.30) 12,131 55 2, 570 (24) Other manufactured
products 27.68 0.20 4,270 (8) 1, 199 6 Nonfat dry milk 102.63 (0.09) 1,396
40 1,012 (1)

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Tables 52 through 57 present the results of our sensitivity analyses for
1999 farm and wholesale- level economic variables.

Table 52: Estimated Change in 1999 Farm- Level Prices Using Different
Assumptions Under an Expanded NEDC in Conjunction With a Southern Compact
Scenario

Dollars per hundredweight Estimated change to baseline farm- level prices
using different assumptions

IRCM9 region No- compact

baseline farmlevel prices in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher over order

premiums Combined assumptions a

$18.00 minimum

southern compact price

Transportation costs increase

25 percent

Northeast $15.07 $0.26 $0.26 $0.41 $0.38 $0.22 $0.25 Appalachia 15.97 1.13
1.10 1.50 1.32 1.85 1.24 Florida 17.14 0.04 0.03 (0.01) (0.01) (0.02) 0.16
Southeast 16.01 0.79 0.80 1.19 1.18 1.66 0.89 Mideast 14.64 (0.19) (0.18)
(0.24) (0.23) (0.24) (0.13) Upper Midwest 13.82 (0.08) (0.08) (0.11) (0.13)
(0.10) (0.06) Central 14.22 0.11 0.12 0.14 0.14 0.17 0.27 Southwest 14.49
(0.09) (0.05) (0.12) (0.07) (0.13) (0.17) Western 13.18 (0.09) (0.10) (0.12)
(0.13) (0.14) (0.10) Northwest 14.85 (0.08) (0.09) (0.11) (0.12) (0.13) 0.03
California 12.98 (0.13) (0.18) (0.17) (0.22) (0.18) 0.18 Arizona 13.76
(0.03) 0.06 (0.05) 0.05 (0.05) (0.36)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional
supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Sensitivity Analyses for the Expanded NEDC in Conjunction With a Southern
Compact Scenario

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 115 GAO- 01- 866 Impacts of Dairy Compacts

Table 53: Estimated Change in 1999 Farm- Level Production Using Different
Assumptions Under an Expanded NEDC in Conjunction With a Southern Compact
Scenario

Pounds in millions Estimated change to baseline farm- level production using
different

assumptions IRCM99 region

No- compact baseline farm- level production in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher over order

premiums Combined assumptions a

$18.00 minimum

southern compact price

Transportation costs increase

25 percent

Northeast 29,138 212 126 203 313 106 126 Appalachia 6, 437 429 283 388 504
478 319 Florida 2, 390 4 2 (1) (1) (1) 9 Southeast 3, 548 111 70 104 164 145
78 Mideast 12,238 (79) (48) (63) (100) (65) (35) Upper Midwest 32,803 (90)
(60) (81) (152) (74) (43) Central 14,855 74 51 64 96 80 121 Southwest 10,278
(41) (13) (33) (29) (34) (45) Western 8, 854 (47) (30) (34) (64) (39) (28)
Northwest 7, 134 (31) (20) (23) (42) (26) 5 California 30,459 (245) (183)
(173) (400) (186) 185 Arizona 2, 914 (4) 4 (4) 6 (3) (29)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional
supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 54: Estimated Change in 1999 Farm- Level Revenue Using Different
Assumptions Under an Expanded NEDC in Conjunction With a Southern Compact
Scenario

Dollars in millions Estimated change in baseline farm- level revenue using
different assumptions

IRCM99 region No- compact

baseline farm- level revenue in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

$18.00 minimum

southern compact price

Transportation costs increase

25 percent

Northeast 4,391 108 95 151 159 80 92 Appalachia 1, 028 146 119 164 172 204
135 Florida 410 2 1 0 0 (1) 5 Southeast 568 47 40 60 70 85 45 Mideast 1, 792
(35) (29) (38) (43) (39) (21) Upper Midwest 4, 533 (39) (34) (47) (63) (43)
(26) Central 2, 112 27 25 30 35 37 58 Southwest 1, 489 (15) (7) (17) (11)
(18) (24)

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 116 GAO- 01- 866 Impacts of Dairy Compacts

Dollars in millions Estimated change in baseline farm- level revenue using
different assumptions

IRCM99 region No- compact

baseline farm- level revenue in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

$18.00 minimum

southern compact price

Transportation costs increase

25 percent

Western 1, 167 (14) (13) (15) (20) (17) (13) Northwest 1, 059 (10) (9) (11)
(15) (13) 3 California 3, 954 (71) (78) (74) (118) (79) 80 Arizona 401 (1) 2
(2) 2 (2) (14)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 55: Estimated Change in 1999 National Average Wholesale- Level Prices
Using Different Assumptions Under an Expanded NEDC in Conjunction With a
Southern Compact Scenario

Dollars per hundredweight Estimated change to baseline wholesale- level
prices using different assumptions

Dairy commodity

No- compact baseline wholesalelevel price in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

$18.00 minimum

southern compact price

Transportation costs increase

25 percent

Fluid milk $15.04 $0.44 $0.39 $0.61 $0.57 $0.59 $0.63 Soft dairy products
27.31 (0.36) (0.17) (0.38) (0.24) (0.43) (0.55) American cheese 139.24
(0.91) (1.00) (1.25) (1.24) (1.41) (0.62) Italian cheese 93.24 (0.62) (0.86)
(0.93) (1.11) (1.02) 0.36 Other cheese 115.58 (0.67) (0.56) (1.01) (1.11)
(1.08) 0.69 Butter 119.52 (2.01) 0.21 (2.32) (0.22) (2.38) (6.53) Frozen
dairy products 21.40 (0.26) (0.10) (0.33) (0.16) (0.34) (0.53) Other
manufactured products 27.68 0.06 0.04 0.02 (0.11) 0.07 1.04

Nonfat dry milk 102.63 (0.07) (0.07) (0.08) (0.09) (0.09) (1.22) Note:
Numbers in parentheses are negative. a The combined assumptions include the
previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 117 GAO- 01- 866 Impacts of Dairy Compacts

Table 56: Estimated Change in 1999 Wholesale- Level Production Using
Different Assumptions Under an Expanded NEDC in Conjunction With a Southern
Compact Scenario

Pounds in millions Estimated change in baseline wholesale- level production
using different assumptions

Dairy commodity

No- compact baseline wholesalelevel

production in 1999

10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

$18.00 minimum

southern compact price

Transportation costs increase

25 percent

Fluid milk 55,043 (233) (445) (330) (647) (328) (324) Soft dairy products
5,984 27 27 35 43 37 51 American cheese 3,539 3 16 5 20 6 1 Italian cheese
3,175 5 21 8 28 9 (3) Other cheese 598 2 3 3 4 3 (1) Butter 1,274 4 (1) 6 0
5 17 Frozen dairy products 12,131 51 26 62 41 65 98 Other manufactured
products 4,270 (1) 3 2 15 (2) (56) Nonfat dry milk 1,396 32 22 40 34 42 98

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 57: Estimated Change in 1999 Wholesale- Level Expenditures Using
Different Assumptions Under an Expanded NEDC in Conjunction With a Southern
Compact Scenario

Dollars in millions Estimated change in baseline wholesale- level
expenditures using different assumptions

Dairy commodity

No- compact baseline wholesalelevel

expenditures in 1999

10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

$18.00 minimum

southern compact price

Transportation costs increase

25 percent

Fluid milk $8,220 $205 $144 $282 $211 $271 $293 Soft dairy products 1,626
(14) (3) (13) (2) (16) (19) American cheese 4,823 (27) (12) (36) (15) (41)
(20) Italian cheese 3,028 (15) (9) (23) (11) (25) 9 Other cheese 792 (2) 0
(3) (4) (4) 4

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 118 GAO- 01- 866 Impacts of Dairy Compacts

Dollars in millions Estimated change in baseline wholesale- level
expenditures using different assumptions

Dairy commodity

No- compact baseline wholesalelevel

expenditures in 1999

10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

$18.00 minimum

southern compact price

Transportation costs increase

25 percent

Butter 1, 495 (20) 1 (22) (3) (24) (64) Frozen dairy products 2,570 (20) (6)
(27) (10) (27) (43) Other manufactured products 1,199 2 3 1 (1) 2 28

Nonfat dry milk 1,012 (1) (1) (1) (1) (1) 1 Note: Numbers in parentheses are
negative. a The combined assumptions include the previous three assumptions:
(1) 10- year long- term regional supply elasticities, (2) higher commodity
demand elasticities, and (3) higher over- order premiums. Source: GAO?s
analysis using the IRCM99.

Tables 58 and 59 summarize the results of our initial estimates and
sensitivity analyses by presenting the range of estimates of the changes
from our no- compact scenario that we obtained from our various analyses.

Table 58: Summary of Range of Estimated Changes in Farm- Level Indicators
Using Our Initial and Subsequent Sets of Assumptions Under an Expanded NEDC
in Conjunction With a Southern Compact Scenario

IRCM99 region Change in farm- level

prices (dollars per hundredweight)

Change in farm- level production (pounds

in billions) Change in farm

level revenue (dollars in millions)

Northeast $0.22 to 0. 41 106 to 313 $80 to 159 Appalachia 1. 10 to 1.85 283
to 504 119 to 204 Florida (0.02) to 0.04 (1) to 9 (1) to 5 Southeast 0. 79
to 1.66 70 to 164 40 to 85 Mideast (0.24) to (0.13) (100) to (35) (43) to
(21) Upper Midwest (0.11) to (0.06) (152) to (43) (63) to (26) Central 0. 10
to 0.27 45 to 121 21 to 58 Southwest (0.17) to (0.05) (45) to (13) (24) to
(7) Western (0.14) to (0.09) (64) to (28) (20) to (13) Northwest (0.13) to
0.03 (42) to 5 (15) to 3 California (0.18) to 0.18 (400) to 185 (118) to 80
Arizona (0.36) to 0.06 (29) to 6 (14) to 2

Note: Numbers in parentheses are negative.

Summary of the Estimated Impacts of the Expanded NEDC in Conjunction With a
Southern Compact

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 119 GAO- 01- 866 Impacts of Dairy Compacts

Source: GAO?s analysis using the IRCM99.

Table 59: Summary of Range of Estimated Changes in 1999 Wholesale- Level
Indicators Using Our Initial and Subsequent Sets of Assumptions Under an
Expanded NEDC in Conjunction With a Southern Compact Scenario

Dairy commodity Change in

wholesale- level prices (dollars per

hundredweight) Change in

wholesale- level production (pounds in

millions) Change in

wholesale- level expenditures (dollars in millions)

Fluid milk $0.39 to 0. 63 (647) to (233) $144 to 293 Soft dairy products
(0.55) to (0.17) 27 to 51 (19) to (2) American cheese (1.41) to (0.62) 1 to
20 (41) to (15) Italian cheese (1.11) to 0.36 (3) to 21 (25) to 9 Other
cheese (1.11) to 0.69 (1) to 4 (4) to 4 Butter (6.53) to 0.21 (1) to 17 (64)
to 1 Frozen dairy products (0.53) to (0.10) 26 to 98 (43) to (6) Other
manufactured products (0.11) to 1.04 (56) to 15 (1) to 28 Nonfat dry milk
(1.22) to (0.07) 22 to 98 (1) to 1

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Tables 60 and 61 provide our initial estimates obtained by modifying the
IRCM to include a more restrictive trade assumption about fluid milk.

Table 60: Initial Estimates of 1999 Regional Farm- Level Indicators Using a
More Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC in
Conjunction With a Southern Compact Scenario

Farm- level prices (dollars per hundredweight) Farm- level production

(pounds in millions) Farm- level revenue

(dollars in millions)

IRCM99 region No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline

The Economic Impacts of an Expanded NEDC in Conjunction With a Southern
Compact Using a More Restrictive Fluid Milk Trade Assumption

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 120 GAO- 01- 866 Impacts of Dairy Compacts

Farm- level prices (dollars per hundredweight) Farm- level production

(pounds in millions) Farm- level revenue

(dollars in millions)

IRCM99 region No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline

Northeast $15.07 $0.28 29,138 138 $4,391 $103 Appalachia 15.97 1.35 6,437
349 1,028 147 Florida 17.14 (0.02) 2,390 (2) 410 (1) Southeast 16.01 2.99
3,548 261 568 156 Mideast 14.64 (0.24) 12,238 (64) 1,792 (39) Upper Midwest
13.82 (0.26) 32,803 (195) 4,533 (112) Central 14.22 0.33 14,855 151 2,112 71
Southwest 14.49 (0.12) 10,278 (32) 1,489 (17) Western 13.18 (0.12) 8,854
(34) 1,167 (15) Northwest 14.85 (0.11) 7,134 (22) 1,059 (11) California
12.98 (0.16) 30,459 (168) 3,954 (70) Arizona 13.76 (0.04) 2,914 (2) 401 (1)
All noncompact regions combined a a 107,070 (519) 14,805 (266)

Note: Numbers in parentheses are negative. a Because farm- level prices are
regional, an ?all noncompact regions combined? price is not

meaningful. Source: GAO?s analysis using the IRCM99.

Table 61: Initial Estimates of 1999 Wholesale- Level Indicators Using a More
Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC in
Conjunction With a Southern Compact Scenario

Wholesale- level prices (dollars per hundredweight) Wholesale- level
production

(pounds in millions) Wholesale- level expenditures (dollars in millions)

Dairy commodity No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline No- compact baseline in 1999 Change to

baseline

Fluid milk $15.04 $0.71 55,043 (336) $8,220 $335 Soft dairy products 27.31
(0.48) 5,984 37 1,626 (19) American cheese 139.24 (1.30) 3,539 5 4, 823 (38)
Italian cheese 93.24 (0.95) 3,175 8 3, 028 (23) Other cheese 115.58 (1.19)
598 3 792 (5) Butter 119.52 (2.48) 1,274 6 1, 495 (24) Frozen dairy products
21.40 (0.51) 12,131 65 2, 570 (48) Other manufactured products 27.68 (0.08)
4,270 2 1, 199 (3) Nonfat dry milk 102.63 0.01 1,396 42 1,012 0

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Tables 62 through 67 present the results of our sensitivity analyses using a
more restrictive fluid trade assumption.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 121 GAO- 01- 866 Impacts of Dairy Compacts

Table 62: Change in 1999 Farm- Level Prices Using Different Assumptions and
a More Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC in
Conjunction With a Southern Compact Scenario

Dollars per hundredweight Estimated change in baseline farm- level prices
using different assumptions

IRCM99 region No- compact

baseline farm- level price in 1999

10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a $18.00

minimum southern

compact price

Transportation costs increase

25 percent

Northeast $15.07 $0.28 $0.28 $0.45 $0.31 $0.26 $0.27 Appalachia 15.97 1.18
1.16 1.59 1.39 1.96 1.24 Florida 17.14 (0.01) 0.00 (0.06) (0.02) (0.07) 0.00
Southeast 16.01 3.01 3.16 3.81 3.84 4.51 3.20 Mideast 14.64 (0.24) (0.21)
(0.27) (0.23) (0.28) (0.24) Upper Midwest 13.82 (0.26) (0.24) (0.29) (0.27)
(0.30) (0.27) Central 14.22 0.27 0.23 0.26 0.15 0.29 0.25 Southwest 14.49
(0.10) (0.05) (0.15) (0.07) (0.16) (0.11) Western 13.18 (0.12) (0.13) (0.17)
(0.17) (0.19) (0.12) Northwest 14.85 (0.10) (0.11) (0.15) (0.14) (0.17)
(0.13) California 12.98 (0.19) (0.20) (0.21) (0.27) (0.24) (0.21) Arizona
13.76 (0.01) 0.06 (0.05) 0.06 (0.05) (0.09)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional
supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 63: Change in 1999 Farm- Level Production Using Different Assumptions
and a More Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC in
Conjunction With a Southern Compact Scenario

Pounds in millions Estimated change in baseline farm- level production using
different assumptions

IRCM99 region No- compact

baseline farm- level production in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a $18.00

minimum southern

compact price

Transportation costs increase

25 percent

Northeast 29,138 225 139 220 255 127 135 Appalachia 6, 437 448 300 412 533
507 320 Florida 2, 390 (1) 0 (3) (2) (4) 0 Southeast 3, 548 418 277 333 534
395 281 Mideast 12,238 (101) (56) (71) (101) (75) (63) Upper Midwest 32,803
(314) (180) (214) (326) (223) (195) Central 14,855 186 104 117 103 130 110
Southwest 10,278 (46) (15) (40) (31) (43) (30) Western 8, 854 (61) (37) (48)
(83) (53) (35)

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 122 GAO- 01- 866 Impacts of Dairy Compacts

Pounds in millions Estimated change in baseline farm- level production using
different assumptions

IRCM99 region No- compact

baseline farm- level production in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a $18.00

minimum southern

compact price

Transportation costs increase

25 percent

Northwest 7, 134 (39) (23) (31) (51) (34) (26) California 30,459 (348) (211)
(218) (491) (249) (224) Arizona 2, 914 (2) 5 (4) 7 (4) (7)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional
supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 64: Change in 1999 Farm- Level Revenue Using Different Assumptions and
a More Restrictive Fluid Milk Trade Assumption Under an Expanded NEDC in
Conjunction With a Southern Compact Scenario

Dollars in millions Estimated change in baseline farm- level revenue using
different assumptions

IRCM99 region No- compact

baseline farm- level revenue in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher over- order

premiums Combined assumptions a

$18.00 minimum

southern compact

price Transportation

costs increase 25 percent

Northeast 4,391 116 103 165 130 95 99 Appalachia 1, 028 153 126 175 182 217
135 Florida 410 0 0 (2) (1) (2) 0 Southeast 568 186 165 201 242 241 168
Mideast 1, 792 (44) (34) (43) (43) (45) (38) Upper Midwest 4, 533 (128)
(103) (124) (133) (129) (115) Central 2, 112 67 49 56 37 62 53 Southwest 1,
489 (17) (7) (21) (12) (23) (16) Western 1, 167 (19) (16) (21) (26) (24)
(15) Northwest 1, 059 (13) (11) (15) (17) (17) (13) California 3, 954 (103)
(88) (92) (145) (105) (93) Arizona 401 (1) 2 (2) 3 (2) (4)

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 123 GAO- 01- 866 Impacts of Dairy Compacts

Table 65: Change in 1999 National Average Wholesale- Level Prices Using
Different Assumptions and a More Restrictive Fluid Milk Trade Assumption
Under an Expanded NEDC in Conjunction With a Southern Compact Scenario

Dollars per hundredweight Estimated change in baseline wholesale- level
prices using different assumptions

Dairy commodity No- compact

baseline wholesalelevel price in

1999 10- year

regional supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a $18.00

minimum southern

compact price

Transportation costs increase

25 percent

Fluid milk $15.04 $0.69 $0.66 $0.91 $0.80 $0.93 $0.69 Soft dairy products
27.31 (0.44) (0.24) (0.48) (0.18) (0.52) (0.31) American cheese 139.24
(1.32) (1.27) (1.68) (1.51) (1.86) (1.38) Italian cheese 93.24 (0.99) (1.08)
(1.24) (1.33) (1.36) (1.11) Other cheese 115.58 (1.25) (0.75) (1.41) (1.33)
(1.50) (1.65) Butter 119.52 (2.08) (0.11) (2.80) 0.07 (2.78) (1.08) Frozen
dairy products 21.40 (0.51) (0.38) (0.64) (0.41) (0.65) (0.53) Other
manufactured products 27.68 0.01 (0.05) 0.08 0 0. 06 (0.18) Nonfat dry milk
102.63 (0.01) (0.02) (0.03) (0.05) (0.03) 0

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional
supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM.

Table 66: Change in 1999 Wholesale- Level Production Using Different
Assumptions and a More Restrictive Fluid Milk Trade Assumption Under an
Expanded NEDC in Conjunction With a Southern Compact Scenario

Pounds in millions Estimated change in baseline wholesale- level production
using different

assumptions Dairy commodity

No- compact baseline wholesalelevel

production in 1999

10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumptions a $18.00

minimum southern

compact price

Transportation costs increase

25 percent

Fluid milk 55,043 (323) (646) (433) (778) (440) (313) Soft dairy products
5,984 35 43 45 50 47 29 American cheese 3,539 5 20 7 25 7 6 Italian cheese
3,175 8 26 11 33 12 10 Other cheese 598 3 4 4 5 4 3 Butter 1, 274 5 2 8 0 7
3 Frozen dairy products 12,131 62 41 77 44 79 49 Other manufactured products
4,270 (1) 7 (3) 5 (3) 9 Nonfat dry milk 1,396 40 39 54 45 54 27

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 124 GAO- 01- 866 Impacts of Dairy Compacts

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Table 67: Change in 1999 Wholesale- Level Expenditures Using Different
Assumptions and a More Restrictive Fluid Milk Trade Assumption Under an
Expanded NEDC in Conjunction With a Southern Compact Scenario

Dollars in millions Estimated change in wholesale- level expenditures using
different assumptions

Dairy commodity

No- compact baseline wholesalelevel

expenditures in 1999

10- year regional

supply elasticities

Higher commodity

demand elasticities

Higher overorder premiums Combined

assumption a $18.00

minimum southern

compact price

Transportation costs increase

25 percent

Fluid milk $8,220 $326 $259 $428 $313 $438 $327 Soft dairy products 1,626
(17) (2) (17) 3 (18) (11) American cheese 4,823 (39) (16) (49) (18) (55)
(39) Italian cheese 3,028 (25) (12) (31) (14) (33) (27) Other cheese 792 (5)
(2) (5) (5) (6) (7) Butter 1, 495 (20) 1 (26) 1 (27) (10) Frozen dairy
products 2,570 (48) (37) (61) (40) (62) (54) Other manufactured products
1,199 0 0 3 1 2 (5) Nonfat dry milk 1,012 0 0 0 0 0 0

Note: Numbers in parentheses are negative. a The combined assumptions
include the previous three assumptions: (1) 10- year long- term regional

supply elasticities, (2) higher commodity demand elasticities, and (3)
higher over- order premiums. Source: GAO?s analysis using the IRCM99.

Tables 68 and 69 summarize the results of our initial estimates and
sensitivity analyses by presenting the range of estimates of the changes
from our no- compact scenario that we obtained from our various analyses
using a more restrictive trade assumption.

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 125 GAO- 01- 866 Impacts of Dairy Compacts

Table 68: Summary of Estimated Changes in 1999 Farm- Level Indicators Using
Our Initial and Subsequent Sets of Assumptions and a More Restrictive Fluid
Milk Trade Assumption Under an Expanded NEDC in Conjunction With a Southern
Compact Scenario

IRCM99 region Change in farmlevel

prices (dollars per hundredweight)

Change in farm- level production (pounds in millions)

Change in farmlevel revenue (dollars in millions)

Northeast $0.26 to .45 127 to 255 $95 to 165 Appalachia 1. 16 to 1.96 300 to
533 126 to 217 Florida (0.07) to 0 (4) to 0 (2) to 0 Southeast 2. 99 to 4.51
261 to 534 156 to 242 Mideast (0.28) to (0.21) (101) to (56) (45) to (34)
Upper Midwest (. 30) to (. 24) (326) to (180) (133) to (103) Central 0. 15
to 0.33 103 to 186 37 to 71 Southwest (0.16) to (0.05) (46) to (15) (23) to
(7) Western (0.19) to (0.12) (83) to (34) (26) to (15) Northwest (0.17) to
(0.10) (51) to (22) (17) to (11) California (0.27) to (0.16) (491) to (168)
(145) to (70) Arizona (. 09) to .06 (7) to 7 (4) to 3

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM99.

Table 69: Summary of Estimated Changes in 1999 Wholesale- Level Indicators
Using Our Initial and Subsequent Sets of Assumptions and a More Restrictive
Fluid Milk Trade Assumption Under an Expanded NEDC in Conjunction With a
Southern Compact Scenario

Dairy commodity Change in wholesale

level prices (dollars per hundredweight)

Change in wholesale- level

production (pounds in

millions) Change in

wholesale- level expenditures (dollars in millions)

Fluid milk $0.66 to 0. 93 (778) to (313) $259 to 438 Soft dairy products
(0.52) to (0.18) 29 to 50 (19) to 3 American cheese (1.86) to (1.27) 5 to 25
(55) to (16) Italian cheese (1.36) to (0.95) 8 to 33 (33) to (12) Other
cheese (1.65) to (0.75) 3 to 5 (7) to (2) Butter (2.80) to 0.07 0 to 8 (27)
to 1 Frozen dairy products (0.65) to (0.38) 41 to 79 (62) to (37) Other
manufactured products (0.18) to 0.08 (3) to 9 (5) to 3

Appendix V: Interregional Impacts of Three Compact Alternatives in 1999

Page 126 GAO- 01- 866 Impacts of Dairy Compacts

Dairy commodity Change in wholesalelevel

prices (dollars per hundredweight)

Change in wholesale- level

production (pounds in

millions) Change in

wholesale- level expenditures (dollars in millions)

Nonfat dry milk (0.05) to 0.01 27 to 54 0 Note: Numbers in parentheses are
negative. Source: GAO?s analysis using the IRCM99.

Appendix VI: Interregional Impacts of Three Compact Alternatives in 2000

Page 127 GAO- 01- 866 Impacts of Dairy Compacts

We present our estimates of the three compact scenarios? impacts on 2000
farm- level and wholesale- level indicators when compared with a nocompact
scenario in the following tables. As with our 1999 analysis, we calibrated
the Interregional Dairy Sector Competition Model (IRCM) using 2000 data to
develop a baseline- an IRCM00. However, we did not conduct a series of
sensitivity analyses for the 2000 estimates for several reasons:

 The data for 2000 were preliminary when we conducted these analyses in
July 2001.

 The dairy industry was undergoing a period of adjustment following USDA?s
regulatory changes to its milk marketing order program in January 2000.
Because the IRCM is a spatial equilibrium model, and the dairy markets
appeared to be in disequilibrium in 2000, we questioned whether 2000 could
be used to accurately estimate the impacts of dairy compacts.

 The sensitivity analyses performed for the 1999 estimates indicated that
the IRCM99 model was robust- that is, the estimates that we obtained when we
used different assumptions were similar to the initial estimates that we
obtained using our initial set of assumptions. As a result, we did not think
that another series of sensitivity analyses would provide much additional
information.

The following tables provide baseline estimates of the dairy sector
indicators under our no- compact scenario and the estimates of the impacts
of the three different compact alternatives. As with our 1999 estimates, we
present estimates under both less restrictive and more restrictive fluid
trade assumptions for the expanded NEDC in conjunction with a southern
compact scenario. Our baseline estimates for 2000 also include the effects
that USDA?s milk marketing order regulatory reforms had on farm- level and
wholesale- level dairy sector indicators. As a result, the baseline
estimates for 2000 are not comparable to those for 1999.

We used the same set of assumptions to develop our 2000 estimates as we did
to develop our 1999 baseline and initial estimates:

 No more than 40 percent of any one region?s milk may be shipped to another
region without being subject to the receiving region?s pricing requirements.
This assumption is used to simulate USDA milk marketing order regulations
regarding minimum pricing requirements for milk shipped between marketing
orders. This threshold was chosen as a proxy for the requirement that an
adjustment be made if a plurality of an order's packaged milk was sold in
another region.

 Supply elasticities are medium- term (that is, 5 years). Appendix VI:
Interregional Impacts of Three

Compact Alternatives in 2000

Appendix VI: Interregional Impacts of Three Compact Alternatives in 2000

Page 128 GAO- 01- 866 Impacts of Dairy Compacts

 Demand elasticities are medium- term.

 The Class I minimum price in the compact region, or the compact price, is
$16.94 per hundredweight of milk.

 Market- driven over- order premiums are zero in the compact region. This
initial assumption represents a lower bound, and assumes that all
marketdriven over- order premiums in NEDC states are replaced by Compact
over- order producer price payments.

 A handler must pay the compact over- order producer price into the compact
pool for milk shipped into the compact region in order to receive the
compact price.

Using these assumptions, tables 70 through 75 compare our no- compact
baseline scenario with our estimated changes in farm- level and
wholesalelevel indicators across different compact scenarios for 2000.

Table 70: Range of Estimated Impacts on 2000 Farm- Level Prices Across
Different Compact Scenarios Dollars per hundredweight

Farm- level prices IRCM00 region No- compact

baseline in 2000 Change to baseline with the NEDC

Change to baseline with an expanded NEDC

Change to baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and 100- mile
import constraint

Northeast $13.39 $0.12 $0.47 $0.41 $0.36 Appalachia 14.35 (0.01) (0.04) 0.46
0.42 Florida 15.64 (0.02) (0.04) 0.03 0.03 Southeast 13.52 (0.02) (0.04)
1.35 2.25 Mideast 12.84 (0.01) (0.04) (0.12) (0.14) Upper Midwest 11.82
(0.01) (0.04) (0.04) (0.43) Central 12.14 (0.01) (0.03) 0.08 0.77 Southwest
12.95 0 (0.03) (0.08) (0.10) Western 10.82 0 0. 01 (0.03) (0.17) Northwest
12.83 0 (0.03) (0.09) (0.12) California 11.05 (0.01) (0.05) (0.10) (0.10)
Arizona 11.93 0 (0.03) (0.09) (0.10)

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM00.

Appendix VI: Interregional Impacts of Three Compact Alternatives in 2000

Page 129 GAO- 01- 866 Impacts of Dairy Compacts

Table 71: Range of Estimated Impacts on 2000 Farm- Level Production Across
Different Compact Scenarios Pounds in millions

Farm- level production IRCM00 region No- compact

baseline in 2000 Change to baseline with the NEDC

Change to baseline with an expanded NEDC

Change to baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Northeast 29,165 66 262 231 203 Appalachia 6, 493 (5) (12) 132 122 Florida
2, 461 0 (2) 2 2 Southeast 3, 505 (1) (4) 139 232 Mideast 12,704 (5) (13)
(39) (45) Upper Midwest 33,104 (15) (34) (41) (375) Central 15,753 (5) (19)
43 420 Southwest 10,920 (3) (11) (29) (33) Western 9, 709 (3) 1 (14) (67)
Northwest 7, 233 (1) (8) (23) (30) California 32,093 (7) (56) (129) (127)
Arizona 3, 006 0 (3) (9) (9)

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM00.

Table 72: Range of Estimated Impacts on 2000 Farm- Level Revenue Across
Different Compact Scenarios Dollars in millions

Farm- level revenue IRCM00 region No- compact

baseline in 2000 Change to baseline with the NEDC

Change to baseline with an expanded NEDC

Change to baseline with an expanded NEDC

and a southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Northeast $3,905 $44 $173 $152 $133 Appalachia 931 (1) (4) 50 46 Florida 384
0 (1) 2 2 Southeast 474 (1) (2) 68 115 Mideast 1, 631 (2) (7) (20) (24)
Upper Midwest 3, 913 (5) (17) (18) (185) Central 1, 912 (2) (7) 18 176
Southwest 1, 414 0 (4) (12) (15) Western 1, 050 0 1 (4) (23) Northwest 928 0
(3) (9) (12) California 3, 546 (4) (22) (46) (46) Arizona 359 0 (1) (4) (4)

Note: Numbers in parentheses are negative.

Appendix VI: Interregional Impacts of Three Compact Alternatives in 2000

Page 130 GAO- 01- 866 Impacts of Dairy Compacts

Source: GAO?s analysis using the IRCM00.

Table 73: Range of Estimated Impacts on 2000 National Average Wholesale-
Level Prices Across Different Compact Scenarios Dollars per hundredweight

Wholesale- level prices Dairy commodity No- compact

baseline in 2000 Change to

baseline with the NEDC

Change to baseline with

an expanded NEDC

Change to baseline with

an expanded NEDC and a

southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Fluid milk $13.55 $0.07 $0.26 $0.47 $0.58 Soft dairy products 27.44 (0.03)
(0.14) (0.29) (0.28) American cheese 140.13 (0.08) (0.27) (0.88) (2.20)
Italian cheese 99.58 (0.08) (0.23) (0.70) (1.71) Other cheese 126.23 (0.09)
(0.28) (0.84) (1.77) Butter 123.04 (0.17) (0.99) (2.22) (1.51) Frozen dairy
products 21.61 (0.03) (0.12) (0.24) (0.17) Other manufactured products 32.49
(0.04) 0.01 0.03 0.66 Nonfat dry milk 102.1 (0.02) (0.02) (0.17) (0.34)

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM00.

Table 74: Range of Estimated Impacts on 2000 Wholesale- Level Production
Across Different Compact Scenarios Pounds in millions

Wholesale- level production Dairy commodity No- compact

baseline in 2000 Change to

baseline with the NEDC

Change to baseline with

an expanded NEDC

Change to baseline with

an expanded NEDC and a

southern compact

Change to baseline with an expanded NEDC, a southern compact, and a 100-
mile import constraint

Fluid milk 55,549 (37) (155) (313) (371) Soft dairy products 6,134 2 13 31
29 American cheese 3,610 0 1 3 8 Italian cheese 3,300 0 2 6 15 Other cheese
613 0 1 3 7 Butter 1, 251 0 2 5 3 Frozen dairy products 12,015 4 21 43 32
Other manufactured products 4, 115 2 1 1 (23) Nonfat dry milk 1,507 3 17 36
41

Note: Numbers in parentheses are negative.

Appendix VI: Interregional Impacts of Three Compact Alternatives in 2000

Page 131 GAO- 01- 866 Impacts of Dairy Compacts

Source: GAO?s analysis using the IRCM00.

Table 75: Range of Estimated Impacts on 2000 Wholesale- Level Expenditures
Across Different Compact Scenarios Dollars in millions

Wholesale- level expenditures Dairy commodity No- compact

baseline in 2000 Change to

baseline with the NEDC

Change to baseline with an expanded NEDC

Change to baseline

with an expanded NEDC and a southern

compact Change to baseline

with an expanded NEDC, a southern compact, and a 100- mile import constraint

Fluid milk $ 7,467 $34 $122 $216 $268 Soft dairy products 1,672 (1) (5) (9)
(9) American cheese 4,962 (3) (8) (26) (67) Italian cheese 3,315 (3) (7)
(19) (44) Other cheese 995 (1) (1) (4) (7) Butter 1,502 (2) (10 (21 (15)
Frozen dairy products 2,574 (2) (9) (19) (13) Other manufactured products
1,347 (1) 1 2 20 Nonfat dry milk 721 0 0 0 1

Note: Numbers in parentheses are negative. Source: GAO?s analysis using the
IRCM00.

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 132 GAO- 01- 866 Impacts of Dairy Compacts

We reviewed and analyzed five studies that provide estimates of the
interregional economic impacts of various compact alternatives. These
studies used a variety of economic models, assumptions about model
parameters such as demand elasticities, and data sets. The compact
alternatives that they examined also varied. Despite these differences, the
results of these studies on the impacts of relatively small compacts, such
as the NEDC, on dairy farmers in noncompact regions were generally
comparable to ours. In addition, these studies agree with ours that, as
compacts expand in size, the economic impacts on dairy farmers in noncompact
regions increase.

A study issued in 1999 by USDA estimated the interregional impacts that the
NEDC would have on noncompact regions. 1 USDA used a model derived, in part,
from a national dairy sector model developed by USDA?s Economic Research
Service and dairy sector data from 1999 to forecast the impact of the NEDC
from 2000 through 2005. 2 The parameters used in the analysis were not
directly estimated, but instead, were drawn from the Economic Research
Service model. The Service?s national dairy sector model is an annual, time-
series dairy model that is estimated as a system of equations using a three-
stage, least squares regression analysis. USDA modified the Service?s model
to allow for a multiregional analysis. The modified model used to estimate
the interregional impacts of the NEDC consisted of five sections: (1) milk
supply, (2) dairy product supply, (3) dairy product demand, (4) market
equilibrium conditions, and (5) regional market utilization and pricing. 3
The major features of USDA?s analysis and model included the following:

 The model used an iterative process to solve a system of simultaneous
dairy demand and supply equations.

 The model used 36 regions, including the former 32 federal milk marketing
orders; California; and three other nonfederally regulated regions.

 The model did not provide for milk movements between regions. 1 USDA
Agricultural Marketing Service analysis provided to the Subcommittee on
Livestock and Horticulture, Committee on Agriculture, U. S. House of
Representatives, June 14, 1999. 2 The NEDC comprises Connecticut, Maine,
Massachusetts, New Hampshire, Rhode Island, and Vermont. 3 For further
discussion of the USDA regional dairy model, see Federal Milk Marketing
Order Reform: New England, Final Decision, Regulatory Impact Analysis,
USDA?s Agricultural Marketing Service, Mar. 1999. Appendix VII: Studies of
the Interregional

Economic Impacts of Various Dairy Compact Alternatives

USDA (1999)- A Study Using an Annual, Time- Series Dairy Sector Model

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 133 GAO- 01- 866 Impacts of Dairy Compacts

 USDA adjusted parameters developed for the Economic Research Service model
to reflect regional differences in the dairy industry.

 USDA used 1999 data to project the NEDC?s impacts in each of 6 years- 2000
through 2005- and the average annual impact over the same period.

USDA used two different scenarios to estimate the NEDC?s impacts: (1) the
NEDC price of $16. 94 per hundredweight of milk would remain constant during
the years 2000 through 2005 and (2) the NEDC price of $16.94 per
hundredweight would remain in effect only in 2000, after which time the NEDC
price would change to the USDA federal milk marketing order Class I price
for Boston plus $1.30 per hundredweight.

Table 76 summarizes USDA?s estimates of the 6- year average annual impact of
the NEDC in (1) noncompact regions of the country affected the most by the
NEDC, (2) all 32 marketing orders combined, (3) California, and (4) the
country as a whole. These estimated impacts are expressed as changes to
average production levels, farm- level prices, and farm- level revenue from
levels that would be expected in the absence of the NEDC.

Table 76: USDA Estimates of the NEDC?s Interregional Impacts on Average
Annual Production Levels and Farm- Level Prices and Revenue, 2000- 05

Amount without the

NEDC Change due to the NEDC using a

$16.94 over- order compact price in 2000- 05

Change due to the NEDC using a $16.94 over- order compact price

in 2000 and a Boston Class I price plus $1. 30 in 2001- 05 Region Baseline
Amount Percentage Amount Percentage New York- New Jersey

Production (pounds in millions) 12,028 (1.6) (0.01) (2.4) (0.02) Farm- level
price (dollars per hundredweight) $15.15 0 0 (0.01) (0.07) Farm- level
revenue (dollars in millions) $1,822 (0.5) (0.03) (1.6) (0.09)

Southeast

Production (pounds in millions) 5, 690 (0.5) (0.01) (0.8) (0.01) Farm- level
price (dollars per hundredweight) $16.56 0 0 (0.01) (0.06) Farm- level
revenue (dollars in millions) $943 0 0 (0.4) (0.04)

Southern Michigan

Production (pounds in millions) 4, 606 (0.4) (0.01) (0.7) (0.02) Farm- level
price (dollars per hundredweight) $15.24 0 0 (0.01) (0.07) Farm- level
revenue (dollars in millions) $703 0 0 (0.4) (0.06)

Chicago

Production (pounds in millions) 14,777 (1.6) (0.07) (2.1) (0.01)

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 134 GAO- 01- 866 Impacts of Dairy Compacts

Amount without the

NEDC Change due to the NEDC using a

$16.94 over- order compact price in 2000- 05

Change due to the NEDC using a $16.94 over- order compact price

in 2000 and a Boston Class I price plus $1. 30 in 2001- 05 Region Baseline
Amount Percentage Amount Percentage

Farm- level price (dollars per hundredweight) $14.60 (0.01) (0.01) (0.07)
(0.07) Farm- level revenue (dollars in millions) $2,154 (1.3) (0.06) (1.7)
(0.08)

Southwest Plains

Production (pounds in millions) 3, 059 (0.3) (0.01) (0.4) (0.01) Farm- level
price (dollars per hundredweight) $15.08 0 0 (0.01) (0.07) Farm- level
revenue (dollars in millions) $461 (0.1) (0.02) (0.3) (0.07)

Texas

Production (pounds in millions) 6, 353 (0.6) (0.01) (0.9) (0.01) Farm- level
price (dollars per hundredweight) $15.16 0 0 (0.01) (0.07) Farm- level
revenue (dollars in millions) $963 (0.1) (0.01) (0.6) (0.06)

Great Basin

Production (pounds in millions) 2, 635 (0.2) (0.01) (0.3) (0.01) Farm- level
price (dollars per hundredweight) $14.25 (0.01) (0.07) (0.01) (0.07) Farm-
level revenue (dollars in millions) $377 (0.2) (0.05) (0.3) (0.08)

Pacific Northwest

Production (pounds in millions) 7, 619 (0.5) (0.01) (0.8) (0.01) Farm- level
price (dollars per hundredweight) $14.37 0 0 (0.1) (0.07) Farm- level
revenue (dollars in millions) $1,098 1 (0.01) (0.6) (0.05)

All federal order markets

Production (pounds in millions) 112,739 20.5 0. 02 30.7 0. 03 Farm- level
price (dollars per hundredweight) $15.04 0.01 0.07 0.01 0.07 Farm- level
revenue (dollars in millions) $16,944 12 0. 07 19.5 0. 12

State of California

Production (pounds in millions) 32,666 (2.2) (0.01) (3.4) (0.01) Farm- level
price (dollars per hundredweight) $14.25 0 0 (0.01) (0.07) Farm- level
revenue (dollars in millions) $4,667 (0.9) (0.02) (3.3) (0.07)

United States

Production (pounds in millions) 165,157 16.3 0. 01 24.8 0. 02 Farm- level
price (dollars per hundredweight) $14.73 0 0 0.01 0.07 Farm- level revenue
(dollars in millions) $24,369 8. 6 0.04 13.2 0. 05

Note: Numbers in parentheses are negative. Source: USDA study.

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 135 GAO- 01- 866 Impacts of Dairy Compacts

USDA also reported on the interregional impacts in each of its thenexisting
32 milk marketing orders and in all noncompact regions combined. For
example, under the first scenario, for which it used a $16.94 NEDC price for
each of the 6 years, USDA estimated that the average allmilk price in all
noncompact regions would decline by less than 1 cent per hundredweight
during the years 2000 through 2005. 4 With regard to its individual
marketing orders, such as the Upper Midwest Marketing Order, USDA estimated
that the all- milk price would decline by 1 cent per hundredweight in 2000,
2001, 2004, and 2005.

A study issued by Cox, Cropp, and Hughes in 1999 estimated the impacts of an
expanded NEDC and an expanded NEDC in conjunction with a southern compact on
noncompact regions. 5 Cox et al. used a spatial market equilibrium model and
dairy sector data for 1997 to estimate the potential impacts of compacts in
that year. For purposes of the analysis, the expanded NEDC included the six
New England states, New Jersey, New York, Maryland, and Pennsylvania. The
southern compact consisted of 10 states. 6

The spatial market equilibrium model also incorporated two additional
modeling features: It simulated the processing of dairy commodities in a
vertical marketing sector and used price wedges, or mark- ups, for each
dairy commodity to simulate USDA?s classified pricing system and compacts. 7
To generate a competitive spatial market equilibrium, the model maximized
producer and consumer surpluses in each region, minus transportation costs,
for the different commodities, subject to certain trade- flow and other
constraints. In addition, the model allowed for

4 The all- milk price is the weighted average farm price of milk that meets
the more stringent federal standards for milk to be used as fluid milk
(although some is used for manufacturing purposes), and milk that meets less
stringent federal standards and can only be used for manufactured dairy
products.

5 Tom Cox, Bob Cropp, and Will Hughes, ?Interregional Analysis of Interstate
Dairy Compacts,? Department of Agricultural and Applied Economics, College
of Agricultural and Life Sciences, University of Wisconsin- Madison,
Marketing and Policy Briefing Paper No. 69, July 1999.

6 The southern compact included Alabama, Arkansas, Florida, Georgia, Kansas,
Kentucky, Louisiana, Mississippi, Missouri, and North Carolina. 7 The
vertical market portion of the model transformed raw milk into dairy
products consistent with how milk components (protein, fat, and
carbohydrates) are allocated and priced. Cox et al. (1999)- A

Study Using a Spatial Market Equilibrium Model

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 136 GAO- 01- 866 Impacts of Dairy Compacts

classified pricing so that raw milk used for fluid milk attracted a higher
price than raw milk used for manufactured dairy commodities. Using an
iterative technique, the model solved for regional farm- level milk prices
and production, wholesale- level dairy prices and production, and
interregional trade flows. This model- the IRCM97- is an earlier version of
the model that we used for our analysis of interregional impacts.

The model assumed intermediate- run (3- to 5- year) supply and demand
functions for 12 geographic regions of the country (the current 11 USDA milk
marketing orders and California), representing different milk and dairy
product supply and demand regions. In addition, Cox et al. used component
yield data from other researchers. 8 Features of their analysis follow:

 The model was an interregional, spatial equilibrium model.

 The model prohibited fluid milk trade between compact and noncompact
regions.

 Twelve demand relationships for dairy products were developed for the 12
regions in the model; these demand relationships were based on consumer
demand for nine distinct dairy products using national estimates of per
capita wholesale demand.

 To link prices among the 12 regions, the model used 1995 transportation
cost estimates provided by researchers at Cornell University.

 The model used 1997 price and production data obtained from USDA for
developing a 1997 base year.

 The model added a price wedge of $2 per hundredweight to the 1997 Class I
differentials in each compact region.

 All scenarios assumed no Commodity Credit Corporation milk price supports
because these were set to expire in 2000.

Under the model assumptions used, the researchers estimated the impacts of
dairy compacts on farm- level prices and revenue, and commodity prices and
expenditures. Tables 77 and 78 provide the study?s results.

8 R. Selinsky, T. L. Cox, and E. V. Jesse, ?Estimation of U. S. Dairy
Product Component Yields,? College of Agricultural and Life Sciences,
University of Wisconsin- Madison, Staff Paper No. 355, Sept. 1992.

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 137 GAO- 01- 866 Impacts of Dairy Compacts

Table 77: Cox et al. Estimates of the Interregional Impacts of Compacts on
Farm- Level Prices and Revenue in 1997 Expanded NEDC Expanded NEDC in
conjunction with a southern

compact IRCM97 region

Change in farm- level price (dollars per

hundredweight) Change in farm- level

revenue (dollars in millions)

Change in farm- level price (dollars per

hundredweight) Change in farm- level

revenue (dollars in millions)

Northeast $0.66 $237 $0.53 $190 Appalachia (0.09) (10) 0.86 99 Florida
(0.08) (3) 1. 48 53 Southeast (0.08) (4) 1. 32 74 Mideast (0.10) (16) (0.26)
(42) Upper Midwest (0.10) (42) (0.22) (92) Central (0.07) (15) (0.24) (52)
Southwest (0.05) (7) 0. 58 79 Western (0.08) (8) (0.21) (22) Northwest
(0.08) (8) (0.14) (14) California (0.08) (31) (0.28) (110) Arizona (0.05)
(2) (0.22) (8) United States 0.06 91 0.09 156

Note: Numbers in parentheses are negative. Source: Cox et al., 1999.

Table 78: Cox et al. Estimates of the Interregional Impacts of Compacts on
Commodity Prices and Expenditures in 1997 Expanded NEDC Expanded NEDC in
conjunction with a southern

compact Dairy commodity

Change in commodity price (dollars per

hundredweight) Change in commodity

expenditures (dollars in millions)

Change in commodity price (dollars per

hundredweight) Change in commodity

expenditures (dollars in millions)

Fluid milk $0.32 $148 $0.70 $326 Soft dairy products (0.10) (3) (0.30) (8)
American cheese (0.70) (19) (1.74) (48) Italian cheese (0.70) (15) (1.14)
(32) Other cheese (0.52) (4) (0.72) (5) Butter 0. 67 6 0.51 7 Frozen dairy
products (0.06) (5) (0.12) (17) Other manufactured products (0.30) (9)
(0.52) (28) Nonfat dry milk (1.15) (6) (1.76) (21)

Note: Numbers in parentheses are negative. Source: Cox et al., 1999.

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 138 GAO- 01- 866 Impacts of Dairy Compacts

In 2000, Kenneth Bailey, an agricultural economist at Pennsylvania State
University, issued a study that estimated the interregional economic impacts
of a large compact on noncompact states. 9 Bailey used a static equilibrium
model, similar in structure to the constant elasticity functional form
policy models developed by Gardner. 10 , 11 In his 2000 study, Bailey relied
on 1997 data to forecast the impact of compacts in 2000. The compact
simulation included those states in USDA?s Appalachian, Florida, Northeast,
and Southeast milk marketing orders. 12 The model was multiregional to
reflect milk supply, allocation, and class prices in federal milk marketing
orders. However, overall supply and demand for dairy products were modeled
at the national level, as opposed to on a marketing order basis. The model
also relied on medium- run supply and demand elasticities as reported in
agricultural economics literature. The data sources for the model included
USDA?s Agricultural Marketing Service, federal milk marketing order
administrators, and the California Department of Agriculture.

The model incorporated several significant aspects of USDA?s milk marketing
order reform that were adopted in January 2000, including component pricing.
13 As specified in milk marketing order reform, the model based the Class I
price on the higher of the Class III or IV price. The model posited 13
regions: 11 federal marketing orders, California, and an

9 Kenneth W. Bailey, ?Evaluating the Economic Impacts of Regional Milk
Pricing Authorities: The Case of Dairy Compacts,? Agricultural and Resource
Economics Review,

Vol. 29, No. 2 (Oct. 2000), pp. 208- 19. 10 A constant elasticity functional
form policy model is one in which the elasticity of supply or demand is
assumed to be constant over the relevant range of the function. 11 Bruce L.
Gardner, The Economics of Agricultural Policies (New York, N. Y.: MacMillan
Publishing Company, 1987). 12 In Bailey?s model, the northern dairy compact
consisted of the following states or portions thereof: Connecticut,
Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New
York, Pennsylvania, Rhode Island, and Vermont; and the District of Columbia.
The Mid- Atlantic compact consisted of the following states or portions
thereof: Georgia, Indiana, Kentucky, North Carolina, South Carolina,
Tennessee, Virginia, and West Virginia. The southern dairy compact consisted
of the following states or portions thereof: Alabama, Florida, Georgia,
Kentucky, Louisiana, Mississippi, Missouri, and Tennessee.

13 Even though the model incorporated changes included in USDA?s January
2000 milk marketing order reform, the model used ?Option 1B? Class I fluid
milk pricing differentials that were initially proposed by USDA, as opposed
to ?Option 1A? milk pricing differentials, which were adopted in the Final
Rule. The use of Option 1B as opposed to Option 1A resulted in different
prices for the various classes of milk. Bailey (2000)- A Study

Using a Regional Economic Simulation Model

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 139 GAO- 01- 866 Impacts of Dairy Compacts

?unregulated? region representing all areas of the country that did not fall
under federal marketing orders or California?s milk pricing plan.

Bailey analyzed the impacts of a large multiregional dairy compact
accounting for about 27 percent of all milk marketed in the model. This
compact scenario was evaluated relative to a no- compact baseline. Bailey
also conducted additional analyses by varying retail fluid milk demand
elasticities (from -0.32 to -0.23) and by using a fixed- percentage farm-
toretail milk markup instead of a fixed- dollar markup. The major features
of Bailey?s analysis and model follow:

 The model did not allow for trade between regions.

 The model estimated supply, price, and demand for fluid milk and three
dairy commodities: butter, cheese, and nonfat dry milk.

 Both a $1 and a $2 fixed amount per hundredweight were used to model an
effective compact over- order producer price payment.

 The assumption about the amount of the market- driven over- order premium
was varied to reflect either the full amount of the market- driven over-
order premium as that in the no- compact baseline or half the amount in the
no- compact baseline.

 The model estimated demand for fluid milk at the retail level and demand
for manufactured dairy products at the wholesale level. In addition, the
model used various farm- to- retail markup assumptions.

Table 79 summarizes Bailey?s estimates of the multiregion compact on farm-
level milk prices, revenue, and production in 2000.

Table 79: Bailey?s Estimates of the Interregional Impacts of a Multiregional
Compact on Farm- Level Prices, Revenue, and Production in 2000

Impact in all noncompact regions Farm- level milk price (dollars per
hundredweight) Farm- level revenue

(dollars in millions) Milk production (pounds in millions)

Modeling assumptions Baseline and

change to baseline Percentage

change Baseline and

change to baseline Percentage

change Baseline and

change to baseline Percentage

change

No- compact baseline $12.94 $15,402.40 $119,042.20 Compact using a $1
compact producer price payment; full over- order premium (0.07) (0.06)
(114.3) (0.7) (212.3) (0.2)

Compact using a $1 compact producer price payment; one- half over- order
premium (0.04) (0.03) (56.4) (0.4) (106.2) (0.1)

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 140 GAO- 01- 866 Impacts of Dairy Compacts

Impact in all noncompact regions Farm- level milk price (dollars per
hundredweight) Farm- level revenue

(dollars in millions) Milk production (pounds in millions)

Modeling assumptions Baseline and

change to baseline Percentage

change Baseline and

change to baseline Percentage

change Baseline and

change to baseline Percentage

change

Compact using a $2 compact producer price payment; full over- order premium
(0.14) (1.1) (217.8) (1.4) (406.6) (0.3)

Compact using a $2 compact producer price payment; one- half over- order
premium (0.10) (0.8) (163.0) (1.1) (305.6) (0.3)

Note: Numbers in parentheses are negative. Source: Bailey, Oct. 2000.

In addition to estimating the multiregional compact?s impact on farm- level
milk prices, revenue, and production, Bailey also estimated changes within
the compact region, and the impact that these changes would have on all
noncompact regions. Bailey estimated that milk production within the compact
region would increase by 0.4 to 1.4 percent, causing lower wholesale prices
for butter (by 0.3 to 1.0 percent), cheese (by 0.5 to 1.7 percent), and
nonfat dry milk (by 0.3 to 0.8 percent) in all federal milk marketing
orders.

A study conducted in 2001 by Balagtas and Sumner estimated the interregional
effects of the NEDC and an expanded NEDC on noncompact regions. 14 Balagtas
and Sumner used an annual, national- level supply and demand simulation
model to estimate the effects of the Compact on the U. S. dairy sector,
based on 1999 data. 15 For purposes of this analysis, the expanded NEDC
included the NEDC states and New Jersey and New York.

The model simulated class and blend prices that would be announced by USDA?s
milk marketing orders in the absence of any compact. The model's parameters
were established by using milk marketing order and Compact

14 Joseph V. Balagtas and Daniel A. Sumner, ?The Effect of the Northeast
Dairy Compact on Producers and Consumers, with Implications of Compact
Contagion,? Department of Agricultural and Resource Economics, University of
California, Davis, 2001.

15 The model built on a price discrimination framework developed by Ippolito
and Masson, which assumed linear supply and demand schedules for milk across
regions. See R. A. Ippolito and R. T. Masson, ?The Social Cost of Government
Regulation of Milk,? Journal of Law and Economics, Vol. 21 (Apr. 1978), pp.
33- 65. Balagtas and Sumner

(2001)- A Study Using a Price Discrimination Framework

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 141 GAO- 01- 866 Impacts of Dairy Compacts

commission data and intermediate- run (3- to 6- year) supply and demand
elasticities. The model also used national- level, as opposed to
regionallevel, supply elasticities.

The major features of the model follow:

 The model included two aggregate milk categories- fluid milk and
manufactured dairy products- as opposed to four milk classes.

 The model estimated the interregional impacts of the two different
compacts in four noncompact regions: California, Wisconsin, Minnesota, and a
combined rest of the United States that excluded the NEDC states.

 The model calculated the elasticity of demand for manufactured dairy
products in the New England region.

Table 80 summarizes the results of Balagtas and Sumner?s estimates of the
interregional impacts of the NEDC and an expanded NEDC on farm- level milk
prices and production in noncompact regions.

Table 80: Balagtas and Sumner?s Estimates of the Interregional Impacts of
the NEDC and an Expanded NEDC on Farm- Level Prices and Production in 1999

Noncompact region Baseline

amount with no NEDC

Change to baseline due

to the NEDC Percentage

change due to the NEDC

Change to baseline due to an

expanded NEDC Percentage

change due to an expanded NEDC California

Farm- level price (dollars per hundredweight) $13.47 $( 0.02) (0.16) $(
0.05) (0.40) Production (pounds in million hundredweight) 304.57 (0.49)
(0.16) (1.16) (0.40)

Wisconsin

Farm- level price (dollars per hundredweight) 13.88 (0.02) (0.16) (0.05)
(0.37) Production (pounds in million hundredweight) 228.34 (0.35) (0.16)
(0.84) (0.37)

Minnesota

Farm- level price (dollars per hundredweight) 14.01 (0.02) (0.15) (0.05)
(0.36) Production (pounds in million hundredweight) 93.87 (0.14) (0.15)
(0.34) (0.36)

All noncompact regions

Farm- level price (dollars per hundredweight) 14.36 (0.02) (0.15) (0.05)
(0.35) Production (pounds in million hundredweight) 1,562.63 (2.34) (0.15)
(5.37) (0.34)

Note: Numbers in parentheses are negative. Source: Balagtas and Sumner,
2001.

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 142 GAO- 01- 866 Impacts of Dairy Compacts

Balagtas and Sumner also estimated the interregional impacts on the price of
milk used for manufactured dairy commodities and farm- level revenue. They
estimated that in 1999, the price of milk used for manufactured dairy
commodities would fall in noncompact regions by about 2 cents per
hundredweight, which would have translated into producer surplus losses for
noncompact producers of about $34 million in 1999.

A study issued by Allen Rosenfeld in 2001 estimated the interregional
impacts of the NEDC and a larger compact on noncompact states. 16 Rosenfeld
used a classified pricing model and dairy sector data from 2000 to predict
potential impacts in that year. For purposes of the analysis, the larger
compact included a total of 29 states located in the Northeast and the
South. 17 These 29 states accounted for about 62 percent of Class I milk
consumption in 2000.

The model used supply and demand analysis within the context of traditional
modeling of classified pricing. Rosenfeld used supply and demand
elasticities from the dairy economics literature to estimate the increase in
milk production and the decrease in milk consumption in the compact regions
caused by a higher compact minimum price for milk used for and sold as fluid
milk. Rosenfeld then used these elasticities to estimate the decrease in the
price of milk used for manufacturing purposes, which he then used to
estimate the decline in dairy farm revenue in noncompact states.

The major features of Rosenfeld?s analysis and model follow:

 The model used a compact Class I over- order premium for both compact
scenarios of 18.5 cents per gallon of milk.

 The model used a supply elasticity of 0.227 to estimate the impact of a
higher blend price within the compact region on milk production. The model
subsequently used a higher elasticity (0.35) in a sensitivity analysis.

16 Allen Rosenfeld, ?The Impacts of the Proposed Expansion of Dairy Compacts
on Dairy Farm Revenue in Noncompact States,? M& R Strategic Services,
Washington, D. C., May 2001.

17 The expanded scenario included the six NEDC states in conjunction with
Alabama, Arkansas, Delaware, Florida, Georgia, Kansas, Kentucky, Louisiana,
Maryland, Mississippi, Missouri, Nebraska, New Jersey, New York, North
Carolina, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Texas,
Virginia, and West Virginia. Rosenfeld (2001)- A

Study Using an Economic Model of Classified Pricing

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 143 GAO- 01- 866 Impacts of Dairy Compacts

 The model used a wholesale demand elasticity of -0.144 for Class I milk
and a wholesale demand elasticity of -0.261 for all other classes of milk to
estimate the impact of price changes on consumption.

 The model did not allow for decreased production by farmers in noncompact
states in response to lower prices.

 The study did not discuss interregional trade. Table 81 summarizes
Rosenfeld?s estimates of the interregional impacts of the NEDC and a larger
compact on dairy farm revenue in noncompact states for 2000.

Table 81: Rosenfeld?s Estimates of the Interregional Impacts of Compacts on
FarmLevel Revenue in 2000

Dollars in thousands Reduction in farm- level revenue under different
compact

scenarios Noncompact state NEDC scenario Expanded 29- state

compact scenario

Arizona $735 $9,510 California 9, 170 118,610 Colorado 475 6,145 Idaho 1,920
24,805 Illinois 520 6,690 Indiana 445 5,730 Iowa 970 12,565 Michigan 1,070
13,825 Minnesota 2, 785 36,020 Montana 55 705 Nevada 120 1,565 New Mexico
1,010 13,075 North Dakota 125 1,615 Oregon 415 5,375 South Dakota 440 5,710
Utah 450 5,795 Washington 1,370 17,735 Wisconsin 6, 825 88,250 Wyoming 20
230

Total $28,920 $373,645

Source: Rosenfeld, 2001.

Rosenfeld noted that his use of 2000 data may have overstated the estimates
of the impact of the NEDC and an expanded compact because, in 2000, the
difference between the Class 1 milk marketing order price and the NEDC price
of $16.94 was larger than in other years since the NEDC

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 144 GAO- 01- 866 Impacts of Dairy Compacts

has been in effect. To account for this, he performed a sensitivity analysis
by adjusting the compact Class I price to simulate the average amount of the
NEDC over- order payment during the first 44 months of NEDC operations. The
result was that the 29- state compact had a smaller impact on dairy farm
revenue in noncompact states ($ 228 million as opposed to $374 million).

The five studies we reviewed and analyzed used a variety of models,
assumptions, and data sets. They also varied in terms of the dairy compact
alternatives they examined. To allow a comparison, table 82 summarizes the
key features of these studies.

Table 82: Summary of Characteristics of Studies That Estimate the
Interregional Economic Impacts of Dairy Compact Alternatives

Study Study characteristic USDA Cox et al. Bailey Balagtas and

Sumner Rosenfeld

Type of model A 36- region timeseries dairy sector model that uses a
multistage regression analysis technique

An interregional spatial equilibrium hedonic model incorporating a vertical
milk marketing sector

A static equilibrium, 13- region, economic simulation model using a constant
elasticity functional form

An annual nationallevel supply and demand simulation model based on Ippolito
and Masson?s price discrimination model

A supply and demand simulation model based on price discrimination inherent
in classified pricing

Compact scenario NEDC An expanded NEDC and an expanded NEDC plus a southern
compact

A region encompassing the Northeast, Appalachian, Southeast, and Florida
milk marketing orders

The NEDC and an expanded NEDC The NEDC and a

29- state compact Baseline year Forecasted for 2000

through 2005 based on 1999 data

1997 Forecasted to 2000 based on 1997 data 1999 2000

Level of model aggregation Market equilibrium is

determined at the national level for seven dairy commodities

Supply and demand are modeled at the regional level for nine dairy
commodities

Supply and demand are modeled at the national level for four dairy
commodities

Supply and demand are modeled at the national level for milk used for fluid
milk and manufactured products

Supply and demand are modeled at the state level for raw milk used for fluid
milk and manufactured products Trade assumption No trade between

regions or compact and noncompact regions

No trade between regions or compact and noncompact regions

No trade between regions or compact and noncompact regions

Not a regional model Not a regional

model

Comparison of Studies Reviewed

Appendix VII: Studies of the Interregional Economic Impacts of Various Dairy
Compact Alternatives

Page 145 GAO- 01- 866 Impacts of Dairy Compacts

Study Study characteristic USDA Cox et al. Bailey Balagtas and

Sumner Rosenfeld

Compact price assumption Two price scenarios:

either $16. 94 per hundredweight for the 6 forecasted years or $16.94 per
hundredweight in 2000 followed by $1.30 per hundredweight above the USDA
Class I price in the subsequent years

Price wedge of $2 over- order producer price payment per hundredweight of
milk

Price wedge of either $1 or $2 overorder producer price payment per
hundredweight of milk

Price wedge of $1.20 over- order producer price payment per hundredweight of
milk

A compact Class I over- order premium of 18.5 cents per gallon for both
compact scenarios.

Other key assumptions Parameterized

supply response Intermediate supply and demand

elasticities Intermediate supply

and demand elasticities; sensitivity analyses include (1) both 100- percent
baseline and 50- percent over- order premiums, (2) lower retail fluid demand
elasticity, and (3) an alternative farm- toretail mark- up assumption

Intermediate supply and demand elasticities; supply elasticity of 1. 0;
fluid demand elasticity of -0.2; national manufacturing demand elasticity of
-0.2; an estimated demand elasticity for the Northeast of -39; and assumes
no over- order premiums

Supply elasticity of 0.227 and demand elasticities of -0.144 for Class I
milk and -0.261 for milk used for manufacturing purposes

Source: GAO analysis of cited studies.

Appendix VIII: Comments From the Executive Director, NEDC

Page 146 GAO- 01- 866 Impacts of Dairy Compacts

Appendix VIII: Comments From the Executive Director, NEDC

Note: GAO comments supplementing those in the report text appear at the end
of this appendix.

Appendix VIII: Comments From the Executive Director, NEDC

Page 147 GAO- 01- 866 Impacts of Dairy Compacts

See comment 3. See comment 2. Now on p. 2. See comment 1.

Appendix VIII: Comments From the Executive Director, NEDC

Page 148 GAO- 01- 866 Impacts of Dairy Compacts

See comment 5. See comment 3.

See comment 3. See comment 3.

See comment 4.

Appendix VIII: Comments From the Executive Director, NEDC

Page 149 GAO- 01- 866 Impacts of Dairy Compacts

See comment 6. See comments 2 and 4.

See comment 2.

Appendix VIII: Comments From the Executive Director, NEDC

Page 150 GAO- 01- 866 Impacts of Dairy Compacts

See comment 11. See comment 10.

See comment 9. Now on pp. 77 to 78. See comment 8. See comment 7.

Now on p. 76.

Appendix VIII: Comments From the Executive Director, NEDC

Page 151 GAO- 01- 866 Impacts of Dairy Compacts

Page 152 GAO- 01- 866 Impacts of Dairy Compacts

The following are GAO's comments on the NEDC's written response dated
September 7, 2001, to our draft report.

1. We recognize that the National Agricultural Statistics Service revised
its estimates of New England's milk production for 2000. However, our
estimates of the impacts of the NEDC are based primarily on 1999 data. While
we also used preliminary data for 2000 to estimate the impacts for that
year, we note that these data are preliminary and as a result, we did not
conduct sensitivity analyses for that year. Further, a 2- percent adjustment
in New England's milk production would be unlikely to affect our estimates
of the NEDC's impact on noncompact regions in the country.

2. We recognize that before USDA's milk marketing order regulatory reforms
took effect in January 2000, New York was in a different milk marketing
order than were the New England states. However, we do not model New York as
being in a different milk marketing order. Rather, the dairy model that we
used to estimate the impact of compacts aggregates states on a regional
basis. As discussed in appendix II, the states included in the model's
Northeast region include the six NEDC states, Delaware, the District of
Columbia, Maryland, New Jersey, New York, and Pennsylvania. Because the
model is a regional model, the NEDC is represented as being part of the
Northeast region. As a result, the model is unable to estimate the impact of
compact states on noncompact states within the same region.

3. We disagree that the report misrepresents the regulation's design and
function. We state that all raw milk used for and sold as fluid milk in the
six NEDC states is subject to the NEDC's regulations, not just milk produced
by dairy farmers in the six NEDC states. We have revised the report to
explicitly state that farmers in New York may receive NEDC payments.
However, the impact of the NEDC on dairy farmers in New York was beyond the
scope of work that we agreed to perform for Senator Kohl.

4. We concur that New York dairy farmers have received NEDC payments, based
on data developed by the NEDC commission. The report, however, does not
state that $50 million has been provided to New York dairy farmers. We have
revised the report to include data published by the NEDC that indicates that
about 1,300 New York dairy GAO Comments

Page 153 GAO- 01- 866 Impacts of Dairy Compacts

farmers have received NEDC payments that have averaged, on an annual basis,
$9,812 since the NEDC began in July 1997. 1

5. We disagree that the model omits from consideration the regulatory
treatment of plants located outside New England that sell packaged milk in
New England- partially regulated plants. The model provides for shipments of
packaged milk between compact and noncompact regions. A discussion of how
this is modeled is included in appendix II.

6. USDA officials and other dairy economists who have analyzed the NEDC's
impact on the premium structure in New England told us that the NEDC has had
the effect of eroding much of the market- driven over- order premium that
processors had been paying prior to the NEDC's establishment, as opposed to
premiums that cover services provided by cooperatives or handlers. This
effect would vary, however, depending upon market forces. Such market forces
could fluctuate from month to month and year to year. Much of the data
needed to determine the specific impact that the NEDC has had on the premium
structure in New England are proprietary in nature, and thus we do not have
access to this data.

7. The report states that the trends regarding farm attrition rates and
production in the NEDC states and the rest of the country were similar, both
before and after the compact. This is not to say that the percentage change
in the NEDC states was identical to the percentage change in the rest of the
country. Regarding attrition, the number of farms has been steadily
declining both within the NEDC states and in the rest of the country, but at
a slightly different rate.

8. We disagree that the report concludes that the NEDC had no impact on farm
attrition. We note that the percentage reduction declined slightly following
the NEDC's establishment but conclude that it is difficult to determine the
extent to which the NEDC, relative to other factors, may have changed farm
attrition.

9. We disagree that the herd size discussion is less helpful. We believe
that providing a longer- term perspective that includes national trends
provides a useful context for discussing dairy policy options.

1 Testimony by Daniel Smith, Executive Director, Northeast Dairy Compact
Commission, before the U. S. Senate Judiciary Committee, July 25, 2001.

Page 154 GAO- 01- 866 Impacts of Dairy Compacts

10. We agree that the NEDC's impact on retail milk prices is a complex
question. Our report states that many factors affect retail prices, in
addition to the farm- level price of milk. We revised the report to
recognize that retail milk prices, after increasing by about 20 cents per
gallon in July 1997, subsequently fell by as much as 5 to 7 cents per gallon
for several months.

11. We relied on USDA's Food and Nutrition Service's analysis of the NEDC's
impact on the Food Stamp Program. On the basis of its analysis, USDA was
unable to determine if the NEDC increased program benefit levels. According
to USDA, if the NEDC's impact on retail milk prices in the NEDC states had
caused a $1 increase in national benefit levels, this would have resulted in
an additional $60 million in federal funding per year. Aside from the issue
of whether the NEDC has increased federal costs, USDA indicated that the
NEDC has increased Food Stamp Program participant costs.

Page 155 GAO- 01- 866 Impacts of Dairy Compacts

James R. Jones, Jr (202) 512- 9839 In addition to those named above, M.
Shawn Arbogast, Venkareddy Chennareddy, Jay R. Cherlow, Nancy L. Crothers,
Oliver H. Easterwood, Barbara J. El Osta, and Marcia B. McWreath made key
contributions to this report. Appendix IX: GAO Contact and Staff

Acknowledgments GAO Contact Staff Acknowledgments

Page 156 GAO- 01- 866 Impacts of Dairy Compacts

Dairy Industry: Information on Milk Prices and Changing Market Structure
(GAO- 01- 561, June 15, 2001).

Fluid Milk: Farm and Retail Prices and the Factors That Influence Them

(GAO- 01- 730T, May 14, 2001).

Dairy Products: Imports, Domestic Production, and Regulation of
Ultrafiltered Milk (GAO- 01- 326, Mar. 5, 2001).

Dairy Industry: Information on Prices for Fluid Milk and the Factors That
Influence Them (GAO/ RCED- 99- 4, Oct. 8, 1998).

Dairy Industry: Information on Marketing Channels and Prices for Fluid Milk
(GAO/ RCED- 98- 70, Mar. 16, 1998). Related GAO Products

(150192)

Page 157 GAO- 01- 866 Impacts of Dairy Compacts

The first copy of each GAO report is free. Additional copies of reports are
$2 each. A check or money order should be made out to the Superintendent of
Documents. VISA and MasterCard credit cards are also accepted.

Orders for 100 or more copies to be mailed to a single address are
discounted 25 percent.

Orders by mail:

U. S. General Accounting Office P. O. Box 37050 Washington, DC 20013

Orders by visiting:

Room 1100 700 4 th St., NW (corner of 4 th and G Sts. NW) Washington, DC
20013

Orders by phone:

(202) 512- 6000 fax: (202) 512- 6061 TDD (202) 512- 2537

Each day, GAO issues a list of newly available reports and testimony. To
receive facsimile copies of the daily list or any list from the past 30
days, please call (202) 512- 6000 using a touchtone phone. A recorded menu
will provide information on how to obtain these lists.

Orders by Internet

For information on how to access GAO reports on the Internet, send an e-
mail message with ?info? in the body to:

Info@ www. gao. gov or visit GAO?s World Wide Web home page at: http:// www.
gao. gov

Contact one:

 Web site: http:// www. gao. gov/ fraudnet/ fraudnet. htm

 E- mail: fraudnet@ gao. gov

 1- 800- 424- 5454 (automated answering system) Ordering Information

To Report Fraud, Waste, and Abuse in Federal Programs
*** End of document. ***