Appendix I. Case Studies

• Pages: 337-395

337

APPENDIX I

CASE STUDIES

A. Introduction

This appendix will discuss a number of different case studies that

illustrate some of the econometric methods discussed in the body of the

text. The reader should note that the case studies are based on advocacy

work done in connection with Worldcom/Sprint, Staples and the

SBC/Ameritech matter. Because the cases involve real disputes, it is

important to recognize that they have not been selected for the purpose of

endorsing either the results of the econometric analysis or t he underlying

substantive positions of the parties in particular cases. Rather, these case

studies have been selected for the purpose of showing how econometrics

has been used to gain insight on substantive issues that frequently arise in

antitrust cases. As the text itself suggests, there are contrary views to

those expressed here. The Section of Antitrust Law does not endorse or

reject any of the viewpoints expressed.

The first case study in this appendix is the abandoned merger

between WorldCom and Sprint. This case has been selected because it

provides a useful example of merger simulation in an agency context. It

is particularly useful because three different merger simulation

methodologies were used in the case—the antitrust logit model (ALM),

almost ideal demand system (AIDS), and price-cost margins—each of

which predicted a different post-merger price effect.

The second case study presented is the proposed merger between

Staples and Office Depot, which was abandoned after successful

litigation brought by the FTC. It demonstrates the use of econometrics in

market definition in a litigation context. It also illustrates how

econometric analysis can be useful in providing confirmation of

documentary evidence.

The appendix concludes with an examination of mergers between

large incumbent local exchange carriers (ILECs). It illustrates how

econometrics can be used to test theoretical propositions and, more

generally, how prior mergers can be analyzed to test the prediction that a

prospective merger will adversely affect competition.

338

Econometrics in Antitrust

B. Case Study: United States v. WorldCom

1. Introduction

In challenging the proposed merger of MCI/WorldCom (WorldCom)

and Sprint, the Department of Justice (DOJ) alleged in its complaint that

the combination would likely lead to a substantial lessening of

competition in the market for domestic residential long-distance service

in violation of the Clayton Act, among other allegations.

1

Econometric

analyses performed by economists hi red by the DOJ were key

components of the DOJ’s investigation, and, in the end, provided key

support for many of the allegations in the complaint.

Significantly, the parties, the DOJ, and third parties took very

different methodological approaches in conducting their econometric

analysis of the proposed merger. In parti cular, economists hired by the

DOJ utilized a variant of the AIDS model that was discussed in Chapter

11, as did at least one third party who submitted comments to the Federal

Communications Commission (FCC); the parties used the profit margin

model described in Chapter 12. And all of this was against the backdrop

of an article published several years before the merger that utilized the

ALM to analyze a hypothetical merger of WorldCom and Sprint. As the

following discussion makes clear, the three different methodologies used

different data sources, made different assumptions, and, in the end,

reached different conclusions about the likely competitive effects of the

merger.

2. Description of the Transaction

The investigation of the WorldCom/Sprint merger was large and

complex, in part because the parties competed with each other in many

different markets. Here, the focus will be only on the long-distance

telecommunications market. Both firms have substantial shares of this

line of business, suggesting the potential for anticompetitive effects.

1

. Complaint, United States v. WorldCom Inc., No. 1:00CV01526 (filed

June 27, 2000), available at http://www.usdoj.gov/atr /cases/f5000/

5051.pdf

Case Studies

339

3. Alternative Methodologies

a. AIDS

The economists hired by the DOJ utilized a variant of the AIDS

methodology in analyzing the competitive effects of the transaction in

consumer long-distance. This analysis was extremely complex and

required the analysis of millions of data observations collected from

multiple sources. The following sets forth a description of the

methodology employed.

(1) Issues in Estimation Stage

The core of an investigation of unilateral effects in a differentiated

products market is an assessment of the degree of economic substitution

between the products of the merging parties. In examining a case such as

the WorldCom/Sprint merger, it is clearly important to estimate the

substitution effects on the demand side of the long-distance market.

Competent demand analysis will reflect the market facts for the

particular demand situation.

(2) The Market Facts

There are two principal ways a residential local exchange customer

can place a long-distance call with a particular long-distance carrier. The

predominant way is by designating a Primary Interexchange Carrier

(PIC)

2

or long-distance carrier to the local exchange carrier (LEC).

When a direct-dialed long-distance call is placed from the access line, the

call is completed using the PIC.

3

The customer can change PIC at any

time. Although their LEC may charge a fee to do so, the PIC often pays

the fee. The customer can also access other carriers using “10-10XXX”

dialing, which is called “dial-around service.” Dial-around allows the

customer to reach carriers other than the current PIC.

2. Telecommunications is full of acronyms. Many of the acronyms have

come into their own, linguistically speaking. For example, PIC serves as

shorthand for Primary Interexchange Carrier, a noun of sorts, but it also

serves as a verb, as in to PIC a carrier.

3. Some carriers such as Sprint and W orldCom have their own long-distance

faciltities on which to carry these calls, but other PICs are resellers of

other carriers’ facilities.