Kamina Johnson, Amanda Ziehl, Steve Davies,

1
Assessing Market Margins, Power, and Risk in the
Wholesale and Retail Pork Sectors

• Kamina Johnson, Amanda Ziehl, Steve Davies,
• and Dawn Thilmany
• Colorado State University
• 2004 WAEA Meetings
• July 2004, Honolulu, Hawaii

2
The Issue at Hand.

• Continued debate on whether increasing
  concentration in wholesale and retail meat
  sectors infers market power
• Particularly controversial in pork markets
• Will the new USDA scanner meat data help us to
  assess whether market power is being exercised
• Boneless, center cut pork chops allow for
  disaggregated analysis

3
Objectives

• Assess whether retailers exercise oligopoly power
  in the consumer market and whether retailers use
  oligopsony power against wholesalers
• Evaluate the role of risk in determining
  wholesale-to-retail market margins
• Analyze disaggregated marketing margins for
  specific meat cuts using richer retail data
  (including feature promotions)

4
Project Overview

• Use several different theoretical and
  methodological approaches to analyze new retail
  meat data
• Price transmission, substitution across meats
• Update Schroeter and Azzam (1991) research on
  market power as determined by marketing margin
  analysis
• Focus more on the wholesale to retail margins and
  competitive interactions

5
Presentation Overview

• Previous Research
• Wholesale to Retail Price Spread
• Competition and Market Power
• Data and Methods
• Figure of Prices and Margin
• Introducing the Model
• Empirical Results and Discussion
• Ideas for Future Research

6
Previous Research

• Fisher (2001) found that adjustments to marketing
  margins are made by changing farm prices
• Gardner (1975) model of perfect competition for
  marketing sectorframework used in several
  studies investigating market power
• Holloway (1991) used Gardner (1975) and
  Wholgenant (1989) concluded no departure from
  competitive behavior in the output markets for
  beef, pork, poultry, eggs, dairy, processed and
  fresh fruits and vegetables

7
Competition and Market Power

• Schroeter (1988) used methods developed by
  Applebaum (1979, 1982) found small monopoly/
  monopsony price distortions in slaughter cattle
  and wholesale beef markets
• Reed and Clark (2000) followed Wholgenant (1989)
  concluded competitive conduct at national level
  for purchase of farm product and sales to
  consumers, cannot conclude the same between
  specific stages of food production or local
  markets
• Schroeter and Azzam (1991) found output price
  risk significant component of margins and
  farm-wholesale margins were more consistent with
  competitive performance than in the past 15 years

8
New Retail Scanner Data

• Standard and feature prices, sales volume index,
  and percent of volume featured
• 142 meat products (pork, beef, veal, lamb,
  chicken, turkey)
• Data from supermarkets with annual sales over 2
  million, who voluntarily provide their
  information
• Represents 20 of all retail meat purchases
• Commercial sources combine the data from
  retailers to protect confidentiality
• Data do not include sales from food service or
  restaurants, butcher shops, warehouse clubs, or
  food distributors

9
Retail and Wholesale Prices Boneless Center
Cut Loin Chop
10
The Empirical Model An Overview

• Conceptual and empirical analysis of marketing
  margins in pork sector
• Decomposition of margins
• Marginal cost of processors (wholesalers)
• Oligopoly/oligopsony pricing power
• Input price risk component
• Updates Schroeter and Azzam
• Wholesale to retail (no farm level)
• Higher level of detail in data
• Input price risk component (rather than output
  since retailers may not be price takers)

11
The Empirical Model-System of Equations
following Schroeter Azzam

• Margin equation-includes marginal cost indices,
  conjectural supply and demand elasticities and
  GARCH(1,2)risk factor
• Demand equation-Pork-to-beef price ratio, income,
  conjectural demand and seasonal dummies regressed
  on retail volume
• Wholesale supply-wholesale price (normalized by
  expected price), lagged volume ratio, conjectural
  supply, lagged price paid to farmers and trend
  regressed on retail-to-wholesale volume ratio

12
Methods

• Iterative 3 stage least square
• Instrumented endogenous variables
• Added trend to correct for pattern in residuals
• GARCH (1,2) process used to estimate price
  uncertainty instrument
• Estimated in Eviews 4.0
• N32, January 2001 to September 2003 monthly data
• One observation dropped for lagged terms
• No remaining autocorrelation

13
(No Transcript)
14
Primary Findings on Margins

• Negative input supply conjectural elasticity
• Signals some oligopsony power exercised by
  retailers on the wholesalers (packing industry)
• No oligopoly power exercised by retailers to
  Consumers
• Positive Risk/Price Uncertainty (GARCH
  instrument)
• As a firm is exposed to more risk their margin
  increases
• Positive correlation between margin and Product
  of Input Supply Elasticity and Ratio of expected
  wholesale price to quantity

15
Output Demand Input Supply Equations

• Output Demand Model
• Little explanatory power
• Seasonal indicator variable
• Outlier indicator variable
• Input Supply Model
• Seasonal indicator variable
• Outlier indicator variable
• Trend
• Lag of Ratio of Retail to Wholesale Volume
• Capacity Constraints

16
Retail-to-Wholesale Sales (Supply) Ratio

• Conjectural term suggests oligopsony power (by
  retailers)
• As todays pork prices (relative to expected
  future price) increases, so does retail sales
  (relative to wholesale output)
• Negative correlation with supply ratio lagged one
  month
• Correction expected since most pork has limited
  storage life
• Positive trend across sample period

17
Discussion

• Wholesalers
• High processing capacity (and cost economies of
  scale)
• Retailers
• Appear to have oligopsony power
• This power persists even controlling for input
  price risk

18
Future Research Potential

• Improving model specification
• Behavior of other pork cuts (i.e. ham)
• Would cold storage potentially change the market
  power interaction between wholesalers and
  retailers?
• Behavior of other meats
• Analyze beef steaks and ground beef
• Chicken, turkey, lamb
• Compare results with more traditional retail
  price data
• Determine if feature price data is important for
  retailers to control retail-to-wholesale supply
  ratio and make margins more flexible

19
Appendix
20
The Model

• Mg0g1TRANSg2WAGEg3(TRANSWAGE)1/2
• q1VOL(BRP/g4) q2VOL(PWPF/(g5QPORK))g6VOL
  RISK
• VOLg7 g4(PRP/BRP)g8SEASg9LOG(INC/FCPI)
• g10LOG(POP) g11OUT
• VQg12 g5PWP/PWPF g13SEAS
  g14(WHOLCOST(-1)/PWPF)
• g15VQ(-1) g16TREND g17OUT