UNILATERAL EFFECTS ANALYSIS AND LITIGATION WORKSHOP

UNILATERAL EFFECTS ANALYSIS AND LITIGATION WORKSHOP

HYPOTHETICAL MERGER FACT PATTERN

Background

This is a consolidated preliminary/permanent injunction suit in federal district court to

enjoin or otherwise redress a planned $2.6 billion corporate acquisition as violative of

Section 7 of the Clayton Act. The parties to the merger are two companies both of which

sell “superpremium” ice cream to the retail channel: Tressel Co. and The Higbee Corp.

Tressel sells a wide range of food products, including Alice’s ice cream, a superpremium

brand. Higbee sells primarily ice cream products, including Longford’s, Bee’s, and

Seattle superpremium ice cream and, through a separate division with its own P&L

responsibility, premium brands such as Higbee’s and Cold Slab. The only other seller of

superpremium ice cream is Incline Corp. There are numerous sellers of premium and

value ice cream, however. The parties stipulate that the relevant geographic market in

which to examine the competitive effects of the merger is national.

The plaintiff alleges that the merger likely will injure competition in the sale of

superpremium ice cream because the combined firm will be able to exercise market

power without regard to the conduct of other firms, i.e., unilaterally. The defendants

reply that the narrowest market in which to examine the competitive impact of the

acquisition is all ice cream; and that the combined share of the parties in an all ice cream

market is trivial, indicating that the merger will not create a firm with market power.

The parties have asserted generally that the merger will result in production, distribution,

and other cost-savings, but have not submitted evidence to prove the extent to which

efficiency gains will be realized from the merger and may be balanced against any

otherwise likely anticompetitive effects of the merger.

Market Structure

•

Tressel has approximately a 39% share of superpremium ice cream sales, based on

revenues.

•

Higbee’s has approximately a 16% market share of superpremium ice cream sales, based

on revenues.

•

The combined market share of the merged entity would be approximately 55%. As a

result of the acquisition, the HHI for superpremium ice cream would increase from 3802

to 5050 (+1248). If the relevant market were as broad as premium and superpremium ice

cream, the combined share of Tressel and Higbee would be measured in single digits and

the post-merger market HHI would be less than 1000, quite unconcentrated.

•

The parties stipulate that entry into the sale of superpremium ice cream is not likely

during the next two to three years, but the defendants continuously have maintained that

threatened or actual repositioning by premium ice cream sellers is sufficient to preclude

any post-merger enjoyment of market power.

Superpremium, Premium, and Value Ice Cream

•

Superpremium ice cream is more costly to produce than value and premium ice creams.

First, it has more expensive, higher quality ingredients, such as Belgian chocolate or pure

vanilla rather than domestic chocolate and artificial vanilla flavoring. Second, it uses

more of those expensive ingredients. For example, the butterfat content of superpremium

ice cream is 14%-18%, compared with 10% -13% for value and premium ice cream.

Finally, superpremium ice cream has less air content than value and premium ice cream,

giving it greater “body.”

•

Superpremium ice cream sellers feature unique flavors and “inclusions,” mix-ins such as

candy, nut, or fruit pieces. Premium ice cream sellers offer a variety of flavors, but their

offerings are not as broad or exotic and offer relatively few inclusions. Value ice cream

sellers offer just a few standard flavors.

•

Superpremium, premium, and value ice creams are sold through the same retail channels,

including supermarkets, club stores, and convenience stores. However, superpremium

ice cream is distributed to the retail channel by direct-to-store-delivery. Manufacturers of

superpremium ice cream promote direct-to-store delivery as “protecting product

integrity” and “facilitating flavor rotation and the introduction of new and exciting

flavors.” Premium and value ice cream is distributed through warehouse distribution by

supermarkets and wherever else practicable.

•

Superpremium ice cream sells at a significantly higher price than premium and value ice

creams. Table 1 summarizes the price differences at retail.

Table 1

Price Differences Betw een Superpremium and Other Ice Cream Segments

Segment

Price Per Ounce

Price Per Pint

Price Per ½ Gallon

Superpremium

$0.25

$4.01

$16.03

Premium

$0.08

$1.36

$5.45

Value

$0.06

$0.90

$3.60

•

The per ounce margin on superpremium ice cream sales to the retail channel is greater

than the per ounce margin on value and premium ice cream sales to the retail channel.

•

Credible third party and party documents and testimony indicate that:

Page 2

<

Incline, the largest seller of superpremium ice cream, seeks to price its

superpremium ice cream to the retail channel at “roughly 3x” its premium ice

cream price.

<

Tressel seeks to price its superpremium ice cream to the retail channel “at price

parity” with Incline superpremium ice cream, both as to “everyday and

promotional prices.”

<

The typical purchaser of superpremium ice cream is younger, better educated and

more affluent, and has fewer children than purchasers of premium and value ice

cream. Manufacturers of superpremium ice cream play to this demographic in

their marketing efforts. Manufacturers of premium and value ice cream tend

instead to market their ice cream to attract the consumer shopping for a household

of three to five (and more) people.

<

Market data indicates that approximately 32% of customer shopping baskets that

include more than a single ice cream product include a product from at least two

of the three ice cream categories. Consumer Intercept data indicates that

premium and value ice cream is more apt to be purchased for children and for

“undemanding occasions,” such as children’s birthday parties, whereas

superpremium ice cream is purchased as a frequent, indulgent treat “for me.”

<

Higbee began selling superpremium ice cream roughly three years ago, and since

that time has sought to acquire and maintain share by pricing its superpremium

ice cream to the retail channel at an approximate 5% discount to Incline. Higbee

repeatedly has told retail buyers that it will match all Incline promotional pricing,

provided that the matching price reduction is passed on to consumers by the

retailer. Higbee has done so.

<

A Higbee “Strategic Review” found that approximately 42% of Higbee’s first

year superpremium ice cream sales likely replaced premium ice cream sales.

However, in the “Strategic Review” for the following year, Higbee reported that

“having established ourselves in the marketplace, expansion of our superpremium

sales no longer is fueled by experimentation and subsequent switching by

premium ice cream purchasers.”

<

Higbee has, on a relative few occasions, offered leading supermarkets temporary

price reductions on select superpremium ice cream products “to blunt the

introduction” of “uniquely interesting” premium ice cream products, particularly

products with inclusions similar to those in its superpremium products. For

example, Higbee offered temporary price reductions on its “Chocolate Covered

Coffee Beans in Cream” ice cream product in response to the entry of a “Coffee

Beans in Cream” premium ice cream product under the “Alfred’s” brand. Higbee

discontinued this price reduction roughly three months after the introduction of

the Alfred’s product.

Page 3

<

Alfred’s introduced its “Coffee Beans in Cream” premium ice cream product as

part of its “Better and Beanier” initiative. That initiative followed a memo from

Alfred’s CEO to its VPs stating: [t]he superpremiums are showing us that the way

to increased sales and profits is through products showing more flair and

obviously better, natural, ingredients. . . . We need to remain price-competitive

with our leading premium ice cream competitors, but we can take the lead, even

at a slightly higher price, by increasing our appeal to the “me, now, generation.”

<

Some leading supermarkets have at least a few times negotiated increased “pay to

play” and “pay to stay” payments from sellers of superpremium ice cream by

threatening to reallocate shelf space among the superpremium ice cream sellers or

from superpremium ice cream to premium ice cream. Higbee appears to have

paid greater-than-customary fees to some leading supermarkets when first it

sought to get its superpremium ice cream on store shelves. An estimated 30% of

the shelf space Higbee gained in its first two years of selling superpremium ice

cream previously had been allocated to premium and value ice cream.

<

Premium ice cream manufacturers compete aggressively with one another on

price to the retail channel. Retailers’ demand for specific premium ice cream

products will drop sharply based on relative price changes among premium ice

cream products of just a few pennies.

<

Sellers of premium and value ice cream to the retail channel and retailers appear

near uniform in believing that an increase in the price of superpremium ice cream

to the retail channel would not result in any increase in the price of premium and

value ice cream to the retail channel. Similarly, they believe that a moderate

reduction in the price of superpremium ice cream to the retail channel would not

result in any reduction in the price of premium and value ice cream to the retail

channel.

Summary of Plaintiff’s Economic Expert (Dr. Cassandra) Testimony

•

Available data does not permit direct analysis of elasticity of demand for ice cream sales

to the retail channel. However, analysis of elasticity of demand by consumers is a strong

and methodologically excellent proxy. Analysis of retail scanner data demonstrates that:

<

Sales of superpremium ice cream are insensitive to changes in the relative price of

superpremium and value ice cream.

<

Sales of superpremium ice cream vary slightly in response to changes in the

relative price of superpremium and premium ice cream.

<

Sales of specific value ice creams vary substantially in response to changes in the

relative price of other value ice creams.

<

Sales of specific premium ice creams vary substantially in response to changes in

the relative price of other premium ice creams.

Page 4

<

Sales of specific superpremium ice creams vary substantially in response to

changes in the relative price of other superpremium ice creams.

<

Econometric analysis of retail scanner data shows that the merger will result in an

increase in superpremium ice cream prices by the combined Tressel-Higbee of

between 3% and 5%.

•

Analysis of third party and party documents and testimony and the economic evidence

indicate that superpremium ice cream is a relevant market, i.e., a hypothetical profitmaximizing sole seller of superpremium ice cream profitably would raise prices by not

less than 3%, and likely by more than 5%.

•

Defendants’ economic expert, Dr. Pangloss, has improperly used Critical Loss analysis

because his assessment of actual loss fails to account for the changed incentives of the

parties post-merger. The combined Tressel-Higbee will be able to choose among a

variety of price increase strategies that allow it to recapture a substantial portion of the

unit sales diverted as a result of a price increase. For example, the combined firm could

raise the price of The Higbee Corp. superpremium ice creams by a few percent, but

maintain the price of Tressel’s superpremium ice cream. If more than 5.7% of the unit

sales lost as a result of a 3% post-acquisition price increase for Higbee’s superpremium

ice creams were captured as Tressel’s superpremium ice cream sales, the price increase

would be profitable. If more than 9.1% of the unit sales lost as a result of a 5% postacquisition price increase for Higbee’s superpremium ice creams were captured as

Tressel’s superpremium ice cream sales, the price increase would be profitable.

Accordingly, Dr. Pangloss’ Critical Loss analysis must be rejected.

•

The analysis of retail scanner data implicitly indicates that the combined firm would

employ pricing strategies under which actual loss would not exceed critical loss.

Summary of Defendants’ Economic Expert (Dr. Pangloss) Testimony

•

The Higbee Corp.’s operating margin on sales of superpremium ice cream to the retail

channel is 50%.

•

Critical Loss analysis indicates that if as a result of a 3% percent price increase for

Higbee’s superpremium ice cream unit sales dropped 5.7%, the price increase would be

unprofitable.

•

Critical Loss analysis indicates that if as a result of a 5% percent price increase for

Higbee’s superpremium ice cream unit sales dropped 9.1%, the price increase would be

unprofitable.

•

The Higbee’s price reduction in response to the introduction of Alfred’s “Coffee Beans in

Cream” premium ice cream product shows, and Dr. Cassandra acknowledges, that

premium and superpremium ice cream pricing is somewhat interdependent. That fact,

among others, leads me to conclude that the actual loss of Higbee’s unit sales from any

Page 5

non-trivial post-merger price increase would exceed the critical loss, and could not be

sustained.

•

Plaintiff’s economic expert, Dr. Cassandra, wrongly applied retail scanner data to the

analysis of a manufacturer level merger, and so her conclusions are wrong. The proposed

acquisition is among manufacturers selling to the retail channel, and so the analytically

correct data points would have related to those sales.

•

Accordingly, the relevant market in which to examine the competitive effects of this

merger is at least as broad as premium and superpremium ice creams, and the merger will

not result in competitive harm.

Page 6

The Higbee Corp.

Superpremium Ice Cream Products Division

2005 Strategic Review

† Introduced superpremium ice creams (Longford’s, Bee’s,

and Seattle) in late 2003 and 2004

† 5% price discount to market leader

† First year sales = $52 million (est. 9% share)

† Some cannibalization of Premium Ice Cream Products

Division sales (Higbee’s and Cold Slab), as est. 42% of

dollar sales displaced premium ice creams

JX 1

The Higbee Corp.

2007 Strategic Review

Superpremium Ice Cream Products Division

† Maintaining 5% discount to market leader (Incline) to

gain/keep share

† And keeping “promise” to retail trade: match market

leader promo pricing if passed thru to consumer

† Share = est. 15% -- $102 million of $630 million

superpremium sales (vs. Tressel, $246 million (40% )

and Incline, $347 million (55%)

JX 9:3

The Higbee Corp.

2007 Strategic Review

Superpremium Ice Cream Products Division

†

†

2008 revenue/profit projection adjustments

„

we will take periodic temporary price reductions to blunt sampling by our

customers of newly introduced, uniquely interesting premium ice cream

products, particularly those with inclusions similar to our own.

„

example: in 2007 we reduced the price of “Chocolate Covered Coffee Beans in

Cream” for 3 months to blunt sampling of Alfred’s new “Coffee Beans in Cream”

premium ice cream product

At the same time, having established ourselves in the marketplace,

expansion of our superpremium sales no longer is fueled by

experimentation and subsequent switching by premium ice cream

purchasers. Our growth will not cannibalize Premium Ice Cream Products

Division revenue/profit

JX 9:7

The Higbee Corp.

2007 Strategic Review

Superpremium Ice Cream Products Division

† Continuing to compete for retail shelf space across

all ice cream segments

„

supermarkets threaten to reallocate space among

superpremiums and across segments

† est. 30% of shelf space held at end of 2005 previously

had been allocated to premium/value ice cream

„

supermarkets continue to demand “pay to play” and “pay

to stay” fees

† “Good News”: supermarkets are no longer demanding

outsized fees for “new brand introduction”

JX 9:9

From: Salvatore More, VP, Ice Cream Products, Incline Corp.

To: Richard Treat, Manager, Superpremium Ice Cream Products

cc: Norbert O. Pints, Manager Premium and Value Ice Cream Products

Date: November 12, 2007

Subject: 2008 Pricing

Sal, the Incline Executive Group has approved your sales/profit targets for

2008. Please note that we will require that you meet these targets while

maintaining a roughly 3x price compared with Incline’s premium ice cream

products. Thanks and good luck. --Sal

JX 62

Alice’s Ice Cream

2008 Marketing Plan

• Theme: The finest and freshest ice

cream indulgences; now at your

neighborhood supermarket

• Price to retail channel: at price parity

with Incline superpremium ice cream

products, both everyday and as

promoted

JX 109

“Better and Beanier Initiative”

„

Building on the vision of CEO Alfred Praline:

“We all know the problem. Premium ice cream is treated as a commodity. And

so it’s price, price, price. If you can’t get (or keep) your price down, supermarkets

will find lots of others that can and will.”

“We need to break free of the commodity trap. And the superpremiums are

showing us that the way to increased sales and profits is through products

showing more flair and better, natural ingredients. . . . We need to remain pricecompetitive with our leading premium ice cream competitors, but we can take the

lead, even at a slightly higher price, by increasing our appeal to the ‘me, now,

generation.’”

From CEO Alfred Praline’s Jan. 7, 2007 Memo to All VPs

JX 42:3

“Better and Beanier Initiative”

„ Introducing Alfred’s “Coffee Beans in Cream”

ice cream – first “Better and Beanier” product

Premium butterfat content and density

‰ All natural ingredients

‰ Imagination and flair

‰ To be priced at 10% premium to other Alfred’s

premium ice creams

‰

JX 42:4

THE 2007 ICE CREAM INSTITUTE

FACT BOOK

Any ice cream is better than no ice

cream. But ice creams vary in lots of

ways, from Value ice creams that

appeal to budget-minded consumers,

to Premium ice creams for those

willing to pay a bit more for a more

“ice cream parlor” taste, to

Superpremium ice creams for the

most demanding and adventuresome

of consumers.

Value, Premium, and Superpremium

ice creams generally vary as to kinds

and costliness of ingredients, butterfat

content, “overrun” (the amount of air

incorporated in the mix), and

inventiveness of the manufacturer.

JX 43:3

THE 2007 ICE CREAM INSTITUTE

FACT BOOK

For example:

z

Value ice creams tend to use artificial flavors, have butterfat

content of about 10%, be “airy,” and come in just a few flavors.

z

Premium ice creams use more natural flavors, have butterfat

content of anywhere between 10% and 13%, have more body,

and come in a greater variety of flavors, sometimes including

small amounts of candies, nuts, or fruits (sometimes called

“inclusions”).

z

And Superpremium ice creams use the finest, often imported,

natural ingredients, have butterfat content of 14% to 18%,

have great body, and come in a riot of imaginative flavors,

often incorporating a wealth of inclusions.

JX 43:4

THE 2007 ICE CREAM INSTITUTE

FACT BOOK

Unlike Value and Premium ice creams, which are delivered to

supermarket warehouses, Superpremium ice creams are delivered direct

from factory to supermarket. Superpremium ice cream manufacturers

explain that this enables them to ensure freshness and integrity (for

example avoiding partial thawing and refreezing) of their ice creams and

to facilitate flavor rotation and the introduction of new and exciting flavors.

All of these differences are reflected in the prices of Value, Premium, and

Superpremium ice creams. The following table shows the approximate

retail prices of each during a 1-month period in late 2006.

Segment

$/Ounce

$/Pint

$/½ Gallon

Value

Premium

Superpremium

$0.06

$0.08

$0.25

$0.90

$1.36

$4.01

$3.60

$5.45

$16.03

JX 43:5

THE 2007 ICE CREAM INSTITUTE

FACT BOOK

{

Superpremium ice cream purchasers

z

tend to be younger, better

educated, more affluent, and have

fewer children than purchasers of

premium and value ice cream

z

tend to buy superpremium ice

cream as a frequent, indulgent treat

“for me”

z

tend to buy premium ice cream “for

the kids,” and value ice cream for

“undemanding occasions,” like

children’s birthday parties

Source:

The Consulting Group, 2006 Supermarket Consumer Intercept

JX 43:26

THE 2007 ICE CREAM INSTITUTE

FACT BOOK

32% of supermarket shopping

baskets that include more than

a single ice cream product

include a product from at least

two of the three categories –

value, premium, and

superpremium

Source: The Consulting Group

2006 Supermarket Consumer Intercept

JX 43:27

Summary of Findings of Dr. A. Cassandra

•

Available data does not permit direct analysis of elasticity of

demand for ice cream sales to the retail channel. However,

analysis of elasticity of demand by consumers is a strong and

methodologically excellent proxy.

•

Analysis of retail scanner data demonstrates that:

–

Sales of specific value ice creams vary substantially in response to changes in

the relative price of other value ice creams; and similarly, sales of specific

premium ice creams vary substantially in response to changes in the relative

price of other premium ice creams.

–

Sales of specific superpremium ice creams vary substantially in response to

changes in the relative price of other superpremium ice creams.

–

Sales of superpremium ice cream are insensitive to changes in the relative

price of superpremium and value ice cream and vary only slightly in response

to changes in the relative price of superpremium and premium ice cream.

JX 94:4

Summary of Findings of Dr. A. Cassandra

•

Econometric analysis of retail scanner data shows that the merger

will result in an increase in superpremium ice cream prices by the

combined Tressel-Higbee of between 3% and 5%.

•

Analysis of third party and party documents and testimony and the

economic evidence indicate that superpremium ice cream is a

relevant market, i.e., a hypothetical profit-maximizing sole seller of

superpremium ice cream profitably would raise prices by not less

than 3%, and likely by more than 5%.

JX 94:5

Summary of Findings of Dr. A. Cassandra

•

Dr. Pangloss, has improperly used Critical Loss analysis because his

assessment of actual loss fails to account for the changed incentives of the

parties post-merger.

– The combined firm will be able to choose among price increase strategies that

allow it to recapture a substantial portion of the unit sales diverted as a result of a

price increase. For example, the combined firm could raise the price of The

Higbee Corp. superpremium ice cream by a few percent, but maintain the price

of Tressel’s superpremium ice cream.

– If anything more than about 5.7% of the unit sales diverted from The Higbee

Corp. as a result of a 3% price increase were captured as Tressel’s

superpremium ice cream sales, the price increase would be profitable.

Accordingly, Dr. Pangloss’ Critical Loss analysis must be rejected.

•

The analysis of retail scanner data implicitly indicates that the combined firm

would employ pricing strategies under which actual loss would not exceed

critical loss.

JX 94:6

Report of Dr. Ian Pangloss

Summary of Findings

•

Dr. Cassandra wrongly used retail scanner data in the analysis of a manufacturer

level merger, and so her conclusions are wrong.

•

Critical Loss analysis shows:

– a 3% price increase for Higbee’s superpremium ice cream would be unprofitable

if as a result unit sales dropped 5.7% (i.e., the critical loss would be 5.7%)

– a 5% price increase would be unprofitable if unit sales dropped 9.1% (i.e., the

critical loss would be 9.1%).

– Higbee’s response to the introduction of Alfred’s “Coffee Beans in Cream”

premium ice cream shows that premium and superpremium ice cream pricing is

interdependent. That fact, among others, leads me to conclude that the actual

loss of Higbee’s unit sales from a post-merger price increase would exceed the

critical loss, and could not be sustained.

•

Accordingly, the relevant market in which to examine the competitive effects of this

merger is at least as broad as premium and superpremium ice creams, and the

merger will not result in competitive harm.

JX 12:3